From 7d23ceb886b1c816ccad89e208dc727be04d212c Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Wed, 22 Apr 2026 14:46:40 -0400 Subject: [PATCH 01/11] Add Cuthbert EnKF There is now an EnKF in cuthbert, as of v0.0.10. This commit adds integrations to this filter, including: - Upgrading the required cuthbert version - Updating the corresponding filter_config information - Make EnKF the default discrete-time filters - Add EnKF integration code - Update tests accordingly - Add to documentation (including discrete-time LTI profile likelihood notebook) --- .../public/inference/filter_configs.md | 6 +- ...discrete_time_lti_profile_likelihood.ipynb | 74 ++++--- docs/faq.md | 6 +- docs/index.md | 2 +- dynestyx/inference/filter_configs.py | 12 +- dynestyx/inference/filters.py | 12 +- .../integrations/cuthbert/discrete.py | 160 +++++++++++++-- dynestyx/inference/integrations/utils.py | 4 + mkdocs.yml | 1 + pyproject.toml | 4 +- tests/fixtures.py | 11 + tests/test_filters.py | 193 +++++++++++++++++- 12 files changed, 416 insertions(+), 69 deletions(-) diff --git a/docs/api_reference/public/inference/filter_configs.md b/docs/api_reference/public/inference/filter_configs.md index bc885b85..2c4d83be 100644 --- a/docs/api_reference/public/inference/filter_configs.md +++ b/docs/api_reference/public/inference/filter_configs.md @@ -7,9 +7,9 @@ The single `Filter()` handler is directed to the appropriate filtering algorithm | Config class | Time domain | When it fits best | |----------------------------|---------------------|-------------------| | `KFConfig` | Discrete | Linear-Gaussian dynamics and linear-Gaussian observations (exact & optimal). | -| `EKFConfig` | Discrete | Nonlinear, differentiable Gaussian dynamics, nonlinear but differentiable Gaussian observations (approximate). *(default)*. | +| `EnKFConfig` | Discrete | Nonlinear or expensive models with Gaussian observations; cuthbert-backed and a good general-purpose default. *(default)* | +| `EKFConfig` | Discrete | Nonlinear, differentiable Gaussian dynamics, nonlinear but differentiable Gaussian observations (approximate). | | `UKFConfig` | Discrete | Nonlinear, differentiable Gaussian dynamics, nonlinear but differentiable Gaussian observations (approximate). Generally more accurate, but slower than `EKFConfig`. | -| `EnKFConfig` | Discrete | High-dimensional or expensive models with lower-dimensional structure and Gaussian observations (approximate). | | `PFConfig` | Discrete | Applicable for arbitrary state-space models, but quite expensive and noisy estimates (asymptotically exact in the limit of infinite particles, approximate in practice). | | `HMMConfig` | Discrete (HMM) | Finite discrete latent state space (exact & optimal). | | `ContinuousTimeKFConfig` | Continuous-discrete | Linear-Gaussian SDE + linear-Gaussian observations (exact and optimal). | @@ -48,4 +48,4 @@ The single `Filter()` handler is directed to the appropriate filtering algorithm ::: dynestyx.inference.filter_configs options: members: - - HMMConfig \ No newline at end of file + - HMMConfig diff --git a/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb b/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb index 2fb38488..ed2c2d71 100644 --- a/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb +++ b/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb @@ -4,11 +4,12 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "# Profile likelihood for discrete-time LTI: KF vs EKF vs UKF vs PF\n", + "# Profile likelihood for discrete-time LTI: KF vs EnKF vs EKF vs UKF vs PF\n", "\n", "This **deep dive** builds a **profile likelihood** and **profile score** for a single parameter (one entry of the $A$ matrix) in a discrete-time LTI system. All other parameters are **known** (no biases). Controls are simulated i.i.d. from $\\mathcal{N}(0,1)$. We compare the profile log-likelihood computed with:\n", "\n", "- **KF** (exact Kalman filter, cuthbert & dynamax)\n", + "- **EnKF** (ensemble Kalman filter, cuthbert)\n", "- **EKF** (Taylor-linearized Kalman filter, cuthbert & dynamax)\n", "- **UKF** (Sigma-point Kalman filter, dynamax)\n", "- **PF** (bootstrap particle filter, cuthbert)\n", @@ -49,7 +50,7 @@ " LinearGaussianObservation,\n", " LinearGaussianStateEvolution,\n", ")\n", - "from dynestyx.inference.filters import EKFConfig, KFConfig, PFConfig, UKFConfig\n", + "from dynestyx.inference.filters import EKFConfig, EnKFConfig, KFConfig, PFConfig, UKFConfig\n", "\n", "# Dimensions: 2 states, 1 control, 1 output\n", "state_dim = 2\n", @@ -192,7 +193,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "**Note:** The PF profile is stochastic (one random key per run). Re-run the PF cell with a different key to see variability. KF and taylor_kf are deterministic for this LTI model." + "**Note:** The EnKF and PF profiles are stochastic approximations. `EnKFConfig()` uses a fixed common-random-number seed by default, so its profile is reproducible unless you set `crn_seed=None`; re-run the PF cell with a different key to see variability. KF, EKF, and UKF are deterministic for this LTI model." ] }, { @@ -210,6 +211,7 @@ " \"\"\"Build data-conditioned model with given filter type.\"\"\"\n", " config = {\n", " \"kf\": KFConfig(filter_source=filter_source),\n", + " \"enkf\": EnKFConfig(n_particles=100, filter_source=filter_source),\n", " \"pf\": PFConfig(n_particles=1_000, filter_source=filter_source),\n", " \"ekf\": EKFConfig(filter_source=filter_source),\n", " \"ukf\": UKFConfig(filter_source=filter_source),\n", @@ -258,12 +260,13 @@ "text": [ "Filter | max profile ll | time (s)\n", "------------|-----------------|--------\n", - "KF (Cuthbert) | -912.7115 | 2.244\n", - "KF (Dynamax) | -912.7115 | 1.381\n", - "EKF (Cuthbert) | -912.7115 | 2.741\n", - "EKF (Dynamax) | -912.7115 | 1.231\n", - "UKF (Dynamax) | -912.7115 | 1.293\n", - "PF (Cuthbert) | -912.7143 | 11.901\n" + "KF (Cuthbert) | -912.7115 | 1.805\n", + "KF (Dynamax) | -912.7115 | 0.915\n", + "EKF (Cuthbert) | -912.7115 | 2.249\n", + "EKF (Dynamax) | -912.7115 | 0.694\n", + "EnKF (Cuthbert) | -916.0390 | 1.562\n", + "UKF (Dynamax) | -912.7115 | 0.784\n", + "PF (Cuthbert) | -912.7143 | 10.546\n" ] } ], @@ -274,16 +277,19 @@ "profile_ekf_cb, t_ekf_cb = profile_log_likelihood(\"ekf\", \"cuthbert\", alpha_grid)\n", "profile_ekf_dynamax, t_ekf_dynamax = profile_log_likelihood(\"ekf\", \"cd_dynamax\", alpha_grid)\n", "\n", + "profile_enkf_cb, t_enkf_cb = profile_log_likelihood(\"enkf\", \"cuthbert\", alpha_grid)\n", + "\n", "profile_ukf_dynamax, t_ukf_dynamax = profile_log_likelihood(\"ukf\", \"cd_dynamax\", alpha_grid)\n", "\n", "profile_pf_cb, t_pf_cb = profile_log_likelihood(\"pf\", \"cuthbert\", alpha_grid)\n", "\n", - "timings = {\"KF (Cuthbert)\": t_kf_cb, \"KF (Dynamax)\": t_kf_dynamax, \"EKF (Cuthbert)\": t_ekf_cb, \"EKF (Dynamax)\": t_ekf_dynamax, \"UKF (Dynamax)\": t_ukf_dynamax, \"PF (Cuthbert)\": t_pf_cb}\n", + "timings = {\"KF (Cuthbert)\": t_kf_cb, \"KF (Dynamax)\": t_kf_dynamax, \"EKF (Cuthbert)\": t_ekf_cb, \"EKF (Dynamax)\": t_ekf_dynamax, \"EnKF (Cuthbert)\": t_enkf_cb, \"UKF (Dynamax)\": t_ukf_dynamax, \"PF (Cuthbert)\": t_pf_cb}\n", "profiles = {\n", " \"KF (Cuthbert)\": profile_kf_cb,\n", " \"KF (Dynamax)\": profile_kf_dynamax,\n", " \"EKF (Cuthbert)\": profile_ekf_cb,\n", " \"EKF (Dynamax)\": profile_ekf_dynamax,\n", + " \"EnKF (Cuthbert)\": profile_enkf_cb,\n", " \"UKF (Dynamax)\": profile_ukf_dynamax,\n", " \"PF (Cuthbert)\": profile_pf_cb,\n", "}\n", @@ -302,7 +308,7 @@ "outputs": [ { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -330,7 +336,7 @@ "source": [ "## Plot profile log-likelihoods\n", "\n", - "All three filters should peak near the true $\\alpha = 0.4$. KF is exact for this LTI; EKF is a linearization; PF is a Monte Carlo approximation." + "All filters should peak near the true $\\alpha = 0.4$. KF is exact for this LTI; EKF and UKF are Gaussian approximations; EnKF and PF are Monte Carlo approximations." ] }, { @@ -340,7 +346,7 @@ "outputs": [ { "data": { - "image/png": "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", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxYAAAHqCAYAAACZcdjsAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjgsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvwVt1zgAAAAlwSFlzAAAPYQAAD2EBqD+naQAA685JREFUeJzsnQV8leX7xq+dWHePwajR3SFdohiAAaioIPqzCwv9G2BgdyAqtoKooKLSKBKCdPdoxlj3dur/uZ75Hs8OZ2Od95fPy855z9v5XM9dbjabzQZBEARBEARBEIRyoCvPzIIgCIIgCIIgCCIsBEEQBEEQBEGoEMRiIQiCIAiCIAhCuRFhIQiCIAiCIAhCuRFhIQiCIAiCIAhCuRFhIQiCIAiCIAhCuRFhIQiCIAiCIAhCuRFhIQiCIAiCIAhCuRFhIQiCIAiCIAhCuRFhISjMZjMeeeQRNGrUCDqdDqNHj1bj3dzc8Mwzz9iP0meffabGHT16tEKO3B9//KGWx78aN998M5o0aWL/znVxmldffbVC1um8fFf7yc8cl5iYiJrCoEGD1CAIglCZWK1WtG/fHs8//7wcaKFCcfXOF8rPY489hl69eqEmIMKiBqM14rXB09MTLVu2xN13342zZ89W6LrmzJmDV155BVdffTU+//xzPPDAAxW6fKF6KUqc2Ww2/O9//yskrLQHv6th/PjxqE4oCovaNt4fZYFijY0oZ1asWAFvb2907doVycnJahwFaVHrz83NRXVR3DnjMHfuXPu03IfLLrvsvGV8+eWX0Ov1GDlypH1filpeZGQkqpMLCX+eT0cRXlevf39/f3Tq1AmvvfYa8vLyzjs+roZZs2ZdcD3ffvstTpw4od41Gv/884/63q5dO/j4+CAmJgbXXnstDhw4UGH79/7776v3Xlk5ffq02vdt27ahpjBv3jzccMMNaNGihTr+taFzaN26deo4pqamoqZR047nb7/9VqhTsjLJzs5W63Ilyu6//35s374dP//8M6obQ3VvgHBhZsyYgaZNm6qX/Zo1a/DBBx+oi3nXrl2q4VMRrFy5EtHR0XjjjTcKjc/JyYHBULWXyUcffaR6zKqS6tjP6oaNqjvvvBOzZ8/Gk08+ed7D8d5770WPHj0KjXO29FQHHh4e+Pjjj88bz0ZxRcH74fLLL0erVq2wfPlyBAcH23/r3Lkzpk6det487u7uqG5cnTPSp0+fYuf7+uuvVaN12LBhWLhwYSGRNnz4cNx4442Fpvfy8kJtpy5c/2z4/fDDD3jooYdUw99RQBK+K3x9fQuNK0mvJjuZKKICAgLs41566SWsXbsW11xzDTp27Ij4+Hi8++67Snj//fffLsV5WYRFaGiouhbLKiymT5+uzhPv05oAz8HmzZvVtZSUlITaAIUFjyPPQ2BgIGoSNe14si323nvvVYm4yM7OVueFOAsqdvZceeWVqvPkiiuuQHVSv1pStZRLLrkE3bt3V5+nTJmCkJAQvP766/jpp58wYcIEl/NkZWWpXqWSkpCQ4PIBUtZe4PJgNBqrfJ3VsZ/VzT333KN6L5944gklXp3p37+/smDVNCgA2WNVWfz5559KVNA66CwqCAV4Za6/PJTlnLExetNNN2HIkCHqmeJ8L/A41NT9LQ915fqnOKJYYE8u3wsNGjSw/8btZ0O9NGzdulX1fNIK4siDDz6Ib775ppCAHjduHDp06IAXX3wRX331Vbn2q65CSyCfGXQxrgjxVdNgJ2B+fn6VvUPr+vEsD7QgUvgfOXIEzZo1Q3UhrlC1EDYASFxcnPrLXgX2Sh0+fBiXXnop/Pz8cP3119sFBntXGTvBni72wFLRsrfO0UVg1apV2L17t91crpnanGMPiuL3339XL2KKGa5/1KhRankVFQPhDLf/tttuUy+5H3/80T6eL7du3bqpHlU2CNnrRpP+hShqP9kjqPXasPdu0qRJqtfAOT7l2WefRfPmzdUx5rY//vjjhVwTHHvk6ErA6dgAuOuuu1yam9mLyuVxP3r27Im//vrL5XYfP34c+/btQ2m57777VC/LtGnT8Nxzz6Ei4HXF43js2LHzfuN6eK5SUlLU94MHD+Kqq65SvSx8ITVs2FCdq7S0tAp1I2QPKxtEYWFh6tocM2YMzp07V+R8PM68dmNjY5WooIgvL3RbZGNQ62lyZP/+/Wo72fNLTCaTmo5mfh4Xrr9fv35YtmwZKoPvvvtONVLZ+0UTekU0DjZt2qT2iS6VzixZskT9tmjRIvU9IyNDmfB5z/CeCA8PVxaSLVu2oDKpS9c/G1ha72VFxL7RYsVtHTBgQKHxffv2Pc8qx+uUz7O9e/decLm0cPD5yX3luY6KilI9rNo28xrgO4PCXnsPaftFV0RaZShi+K6jCxg73CiANPjO0ixMXI+2DEfXqg0bNihXPz7Lae0fOHCgekY4UtHXpBa3WFboqcB3EwU+rxUet7Fjx6r3vcaF3vMaPB50Z+M5ZqOc0/L8LV682D4N1/Xwww+rz/SU0I6jdp60ZdDKqb3LtPkpSnleeH54noYOHaqsWRVJeY8nO1FvueUWREREqONJV0LnZ1VRcSBae0m7ptg24HOEOLobOrtf0hOkcePG6n3Oa47eJiWJn3RsC3F5fI8RviO0dTm2W2hxJuwgqk7EYlEL0R4ojo0eNm4vvvhi1QjhhcyHJh8qNIlRNPBGommYL3Y+NE6dOqUudl6o7AFgkF5mZiZmzpypltemTZsSbw/nZ48n109zORveNFdyW/igqWj3AYvFgsmTJ6seugULFqiGIOE+0KWBqp2WHTYg33nnHfWC5HaUxaTLZfHhyuPCFwtdEPii4X5qcF18MLF3kA93vrw4PV+23D4NPgD4QODNf8cdd6hGJY8TXRj4ctMsNZ988ony++aLnC849j7wPFIo8aHqCF1U+CJ2foEUB+Nn3n77bTz66KN44YUXipyOL1hnH3ZuQ1EPdR4rJgBgY1V7MWlw3IgRIxAUFKR6t3itUHix15iNK16PbGxSZDm6XxSFK996Nnr4QnOEy+c6n376afVgfvPNN9VLkdeOMzwHFOY834yvKKqnl41/5/XzfivKLZEvML5MeAy4HY5wO+jCxV4m7RrhtcNrioIyPT1dNdR57bFxcyFcnTPtWaG98DToQsMOCN4fv/zyS5HuTWzYOC+TnQdsULiC1lX2lnF/+Vxw3l+eD55/cvvtt+P7779X56Rt27bKtYHunrx36GJTGdSF678k7wSixQZp8FrjNlzIDYaNzpJYjvncoXBmA/NCUEhROHCf+U5gA4+CmZ0j/M57k7+xQUorknbvED4D2RjmfcL7k+v88MMP1X21Z88e1UnDdxYtT0899ZTqdGJHF+FzVHNvZKOXHU+8D3keP/30U9VRxw4F3m/VdU0W965jPBSfRxSeFMS8Lnnc2Dhl51NJ3vOOcF/YGUdLF+9j3gs8NzwPvH4oWhg3wzgbzqs9B7VGrXYseU3zGPF3TRTymPMZzPuA1w/PERvMfEfVhMBiujxzew4dOqS2ndfS/PnzVQOe9x6Pb2nge5rudzwfbAe54osvvlDnjJ2IfJa+9dZb6prbuXOn/fouCTz+bC+w7cAOMp4nQrdEDT47eE3wXVatcbI2ocby6aefsrVoW758ue3cuXO2EydO2ObOnWsLCQmxeXl52U6ePKmmu+mmm9R0jz32WKH5Fy5cqMY/99xzhcZfffXVNjc3N9uhQ4fs4wYOHGhr167dedvA+Z9++unztikuLk59z8jIsAUGBtpuvfXWQvPFx8fbAgICzhvvzKpVq9Ty+FeD+9O4cWP7d66L07zyyis2k8lkGzdunNr/JUuW2Kc5evSoTa/X255//vlCy9+5c6fNYDAUGu+8fFf7yc8cN3ny5ELTjRkzRh1/jW3btqnppkyZUmi6hx56SI1fuXKl+p6QkGBzd3e3jRgxwmaxWOzTvfvuu2q6OXPmqO/5+fm28PBwW+fOnW15eXn26WbPnq2m43lyhN9Lchtrx5D7zb8PP/zwBc+Jq0E770XRp08fW7du3QqN27hxo5r3iy++UN+3bt2qvs+fP99WWrRr3dVw8cUXn3edDhs2zGa1Wu3jH3jgAXWdpKamFjqGwcHBNj8/P3UP8FwVhXb8nAfHa8cVH374oZqO16Mjbdu2tQ0ZMsT+vVOnTrZRo0aV+rgUd844nDlzptA+NGjQQN0XgwYNsmVlZRW53KKWx+NbHNOmTbMZjUZbcnKyfRyvZz4rHO8pPiPuuuuuUu+vdn/yuegKnkfHe6UuXf8+Pj5qvznwGf7CCy+o53nHjh3POz7Og/NzzxUNGza0XXXVVSXani+//FIt95NPPil2upSUFPszvDicz5tGbm5uoecm4bnw8PCwzZgxwz7un3/+cXl98hnQokUL9YxwfB5kZ2fbmjZtahs+fHi5r8mSUNT+FQXfC9yf119//bzftP0ozXue0/E95Dhu+/btavw777xjH8fzVNT1zvE6nc62e/fuQuNHjx6tln348GH7uNOnT6vn6oABA4p951fV8XzzzTfVur/66iv7OL5zed/6+vra0tPTi91G7TnieH3xWnH1Do77d1rHthrZsGGDGs93kQb3wdV+OLdVeM9f6H3DNkabNm1s1Ym4QtUC2MNNtcreavZasEeHPeH0M3SEStY5qIg9VAxCdIS96nw+0H2pvFCpU+kz1oO9e9rA9bKHgr0oFQV7+thjxZ497ht7ADXYA0NfT/YaOm4HewNpri/rdrD3yhH2yLAHiz3JhNtB6G7jiBbc++uvv6q/dKvh9tMC4djjeeutt6oeHm069k6zJ4/rdXQ7YI+Kq55MmmpLY63QsonRrH4h2PPH8+s4XCgbEH2uGVjnaKZnLzV7t+n2QLT9YK+as1tZSaD52nm7ONDP2xn2XDr21PP8sRfQ2V2FrgTsVWIPkrPVwxle187rdg5udoa9S3SHcrSUsMeRva08Zhq0qrHnj64yZcHVOePgHCfCnmxaOemWcqFAbJ435+VpFoei4D7RsuPoprh06VL1rHDeX1r42OtXFdSF65/XKt8HHOiyR7dLBuc7WkcdrVKO20/3lQvB59uFrBqELpjsheW6nS1TzvAa4/OMzyvNHaw08Phpz03ev9xGvgfp8lMSFyVmieI9dd1116l5tfcDjyXddVavXm1PGFLV12Rx8PzRIkBLjjPac62073m2J9irrcEebz7zaBUqKbQU0ZqjwXPC+5tp6h19++m2xWNOK4n2zqxOeKx4DzvGptKywmNHjw1aViqa0aNHF2qr0TLGd4jWdqhoeO9Wd5p8cYWqBdCHjy9CNkzY8OHD1Nkcz9/YSHCEjSeaiGnudERzc3LlC1xatAaQFvfhzIUaaaWBLiK8+fmgdPZH5HbwIUoRUZEB4Uyp6Ij2wuXLkfvGY8hzwRe8I3x48QWlHWPtL8+dI3zZ8kHsPJ3zfnD7KyIYi+4ffKDRhMvtKy44lf7Mms9mSaHwo8hiY4oNHp4Tmpo1v1tC8zOnYaApGzps7NOUT1//kriB8CVa0u0q7vw5wvNHccDjw5cOt7moLFN80Zf2uHAeNmDoPsB4HMJjxPtWM2kTunKwAcr7ne4o9AefOHFiIXN3cZT0nHFbeGxoWqfooHm+KPhcKe3+0m+5devWah/pnkH4mcfB8Vnx8ssvq0YpO03ookJXNJ6HirjWnV2/6sr1T2FN1zWtwc3lOT/7NejmVtrgbXKhzgrGS9AFldtLt6ELZWTjdtJ9lI1dvsN69+6tXHx4rkuSupiNfl6jjFFjbCEbsholiYPS3lPFCSDGt/D5UJnXZGmhQOU7o7iMhaV9zzs/Ewn3uzSCj9ecI3Q7pkh2fr9p28Hzx1jHkrjMVSY8Fny3OrefKrJN5IyrNgmf73wXVAa8d109+6oSsVjUAqhw+YJjY5o3gCsfX8cenapE6+Whf6GrntKKDCJiLykDcPngd64ZwO3gzcQgMlfbQV/PslDUC9NVUFxtgL18FGZs9NG/nr1MFQlfcGwoaQ9NBu7Rd9exl5ow48yOHTtU44t+r+wx4kvn5MmTFbo9JT1/hH7BHNjzS0tSaSxBJYHWRvouazn2eYzYwHds+LEhyMYE68pQWDCmh37drtLrlhcGjHOb6GNdGakSec5pKWTvGeMJGBxOX27HRhItjOwpZSwUrx2mOeV1cCFrqhZkzmvHFWzkuApErwvXvyasOXBdRYmKssKGenGNTDbAKZRofeLz1jELVXHQWsvrnx1EPDeMh+P7jPFvF4KxMBRjvD+YoIPWHj7XecxKkppcm4bXl6v3AwctLW9Zr8naQmmeiUVRF9JNF0dR73NHQVtT15eSklKmzoSKRIRFHYZZCGjOpYuHI1oWIf5eXjSTKgOatZed41CRxWvYy8UAPgYXsmeQrhyO28EHI3tSXG0H560MeAz50nJ2XaHLBV+82jHW/jJg2xG6R7EHznk65+XRrUTLAlYRDQc2qLTsIuvXr0dFwkYUs7VwX9lzy6Bmpm911SP8f//3f8oNgcGTDDQsSfGuyoS9qgycZlAns9BUtEmcFioeE4oLNrJcFVyjBYFZbbQiZbRWVEbDnx0RDCxkI5FJBSgwKvo64D1Kdw42yugK4Wp/eR0ykJT3Nq9xXp8Xqvhc1P2kiQoet6Keb3L9Fw9FV1HPGnbo8F7mtUuXVEd3mJLA5zStFnz+0BWQzz/HtLZFNbBoFRk8eLBKbMFriG6wfK47Z9Qran7tPUWrkav3AwdHq3ZZrsnKgNvNa5zP/6p8z5e2o4xueXzOu7ofuR181jgnHqkOeCz4bnUWo87HSrNsO19friwaFzpWB124tfL+cUxqw/W5yg7pvL6SnBder6VJvlMZiLCow9CES8WrpbLUYKYHXqBsUFSEFYEPa/YouXr4FZfasyzwBcC8++wpo4uI9oBgA5k9MWwgOfe88HtlFdLhMSbMaOII3RyIlrGK281GJRtvjtvHFyV7ALXpmFGHD2k2sPnS1WB6O1cPnrKmm6XPJ3vpaAHiupmhoqJgrzTPBRvGdAOhy4NjTRU2MB1FoSYy+PJxlaK3qqF1iy4yPIcVlYqU0PWG9wt7s3kN83qg2HDE+TplLyrdtCrruLAxxUbbRRddpHqUi8psUhb4cuN5pZDiwMaaYwpTPpuc06uyg4K9xBfaX1p6ePzoyuXcSGCqZl5fxT3f5PovGsZMsNHvfA54vigW2RHB+/pCRRedxZ6zlZmNZrrvOK6HzwlXzzk+T5yf69wGdkY4oj1nnJdBlyaujxkT6U5b1HuqPNdkZcBnKS1+zu9woh2PynjPF3Uci4Lnh2KPHgqOKY/ZwcbaJ8wQWZFu0WWFx4pufI6xbnxW0DrFZy1jRzSBwX1ip5cjdMUr7bFauHBhoet048aNKobH8bzw2uR73LG9xM4551TIWubBotbFa5cWby0TWnUhMRZ1GPYssZeHqft4s9PvmT1FvPnZiHAM4CorfFjw5c5GPl022JvEhjEbvAxIZoPF1UOxPLAxxh5l+r1y/WwIcl/YCGS+eO4rp+FLi+qdri0M4q3oHmjCY0p/XDZmeLPzwcQHB9PPcht4/AmPCbeNwod+8/SpZu8OH1TMva4VvGJDj/tBH3D6ovNFzn3g/rry8S1LullH30+6FNCqxAYvA+wqwo+YL2LuNxvm7EVzdgNhqkKm+qPVib6mfLCzQcsHOV+kF4LTF1WMi2n4SlMY0hUUOPR950Oa7hq0ILD3siLgseC55nnnMXdOgcweYJ4PNoS4Xgbza6kvSwItP84NOEKrR1FxGnxZ8V7ltcs0zvSbr6jKrdxfBkHT9YWxFo7umrw26MZDEcf7iC92Jjlg+mXn4myurjEulxYvihVuL/eD1kwKWjZyXFnJ6sL1X9kwxodxQHyuOCbIoKWB7mw8rgz+d74HiyuiyB5aikG6GfEapzscn8tseDpasXjd833CZyAFNY8ln4PsnGD8ES15bDSxI4T3qPP54nuA9xQ7Zvj857OAgbK0ZNOdkI05ujVxORSXbPDRXY/vEcatlPSaZBA6zzHT1l7ImsjGqdZAZcORAeNahwWvXed6Ic7Pd1oV6QbG9wpd3zg/t4nPJJ6rynjP8zwQLpPnh+8lrqe4Zyv3iZ1VFBHcNp5jvpspyOi+XBxVdTzZDuA2MRkKkyzQasDnKxvw7BzU4lT4DOT9ScFBccZjSAsdE6sUdazozsjniF6vL3RN8zrmMWFyHR4LrocWMLrdavC5y+cF5+dzkuvhNcxr1THonS5ovH8ojPjs4DuCLrNaoUBeF2wLaIkiqo1qzUklFIuWMpMp9EqSgtAVTAfLtGZML8n0j0y5x1Ryjin3ypNuVoNp2ZjKj6n6PD09bc2bN7fdfPPNtk2bNlVoullH3n//fTWeqV01fvjhB1u/fv3U8eDQunVrlQ5u//79RS7f1X4Wlc7S1f4zBe706dNV2kIe40aNGql0m0yR6AzTy3KbOF1ERITtjjvuUKkYneG+cXlMp9i9e3fb6tWrXaakK226WVfpHv/66y+VEo/rO3XqlP2clCUdpsZHH32klsFUgzk5OYV+O3LkiEo5ymuE1wpTvQ4ePFilVS5PulnH81LUvePqeivq2s/MzLT17t1bpVb8+uuv1TheN2VJB6vBdIY81s4pDzWYMrJnz54qLSun47XCVMlMiViedLOO13ZR+8AU0bGxseqcaMeH85Yn9ebBgwft27BmzZpCvzH9LNO+MsUurxPer/zMa7+k8BjyHHFe3is8XrwXne+9unT9F/WsL0063gvB1LW33HKLy2dNUUNxJCYmquuI54fbz/dEr169bN9999151yCvTR43x/TaPJ9Tp061RUVFqXN10UUX2davX+/ymfjTTz+pNM5Mp+ycGpSpfseOHatShvN64b1w7bXX2lasWFGqa/KXX35Ry541a9YFj2VRqX9LkqZaS4n7xBNP2N8vkZGRKpWsY1rXkr7ni7qfeRx4bTny7LPP2qKjo9Xzz/HZWtwzYcuWLaodwNSt3t7e6rpet27dBZ/BVXk8z549a5s0aZItNDRUpcft0KGDy/TZvHeYdpn7ERQUZPvf//5n27Vr13nXlNlstt1zzz22sLAwld4X/94Ljs+c1157TbULeM31799fpfh19Sxr1qyZ2iamm2c6fVdtFR5PprTmdM77zFT8bP9UN278r3qljSAIgiAINQVaUJhKlpbnshQWreuwt5mWMRZaK6pQpFC/j+fRo0eVpYzB/5XhLeEMXby4PrrZVrfFQmIsBEEQBEGww4xZTEvKVOfC+dB9im6SdaURXN3I8Sw/dLFiTFt1iwoiMRaCIAiCINhhLAwDuAXXMOZCqDjkeJYfVwViqwuxWAiCIAiCIAiCUG4kxkIQBEEQBEEQhHIjFgtBEARBEARBEMqNCAtBEARBEARBEMqNBG+XE1Z9PX36tCqsUpJy64IgCIIgCIJQW2BlChaQZBV6x0KnrhBhUU4oKho1alTexQiCIAiCIAhCjeXEiROqOn1xiLAoJ1oJeB5sf3//8i5OEARBEASh0snPz8drr72mPk+dOhXu7u5y1AWXpKenq050rc1bJ4XFli1b8Oijj6r8x3q9HldddRVef/11+Pr62qe59957sXbtWpWPu02bNti2bdt5y9mxY4eqMMrlhIWF4Z577lFVIEuK5v5EUSHCQhAEQRCE2iIsPD091We2X0RYCBeiJC7/utrqfjRs2DDExsZiw4YNWLx4MXbv3o2bb775vGknT56McePGFanARowYgcaNG2Pz5s2q9PozzzyD2bNnV8FeCIIgCIIgCELdoVZaLBYtWgSj0Yj33nvPHkQya9YsdOzYEYcOHVKCg7z99tvq77lz55Rlwpmvv/5aKfY5c+Yopd6uXTtl1aDl47bbbqvivRIEQRAEQRCE2kuttFjk5eUpIeAYme7l5aX+rlmzpsTLWb9+PQYMGFDI/HfxxRdj//79SElJKXLdtHQ4DoIgCIIgCIJQ36mVFoshQ4bgwQcfVK5L9913H7KysvDYY4+p386cOVPi5cTHx6Np06aFxkVERNh/CwoKOm+emTNnYvr06aXeZovFApPJVOr5BKE8OAtwQRAEQRCEeiEsKA5eeumlYqfZu3evcln6/PPPlbiYNm2aCt5moDZFQWU3org+rtc5Ur643L8UKampqZW6XYLgCt4PFM8SlCcIgiA4wrbTwIED7Z8Foc4JC6Y7cxWA7UizZs3U3+uuu04NZ8+ehY+Pj4pUZ2yE9ntJiIyMVPM7on3nb67w8PBQQ0nRREV4eDi8vb2liJ5Q5cUbacWLiYmRa08QBEGwQzExaNAgOSJC3RUWTPfKoTRorksMwGbatOHDh5d43j59+uCJJ55QLkoMBifLli1Dq1atXLpBlcX9SRMVISEh5V6eIJQW3k8UF2az2X6NC4IgCIIgVAa11vn63XffVbUsDhw4oLJD3X333Sr+ITAw0D4NM0QxyxOtBjk5OeozB2aCIrR40EXklltuUelq582bh7feequQq1N50GIqaKkQhOpAc4GiyBUEQRAER1fthIQENfCzINQ5i0Vp2LhxI55++mlkZmaidevW+PDDDzFx4sRC00yZMgV//vmn/XuXLl3U37i4ODRp0gQBAQFYunSpKpDXrVs3hIaG4qmnnqrwVLMlKSgiCJWBXHuCIAhCUZ2fH3zwgT1+VGLxhHotLL744osLTvPHH39ccBrWvvjrr78qaKsEQRAEQRAEoX5Sa12hhNoLLUsvvPBCla2P1qk333yzyN+PHj2qevbpJldVsKDj5ZdfXmXrEwRBEARBqGxEWAguYXau0aNHFxr3/fffqwD51157zT4NG+TOA2NbimL79u347bffVHpgRzjPpEmT0LBhQ5V1iylSJ0yYgE2bNpX4DH322WeFYmxqCjwmCxcuLDRu8uTJKkZIrGWCIAiCINQVRFgIJeLjjz/G9ddfr/wxmRZYY+TIkSqdqePgXHTQkXfeeQfXXHMNfH197eMoHhjjwkB8xsrs2bMHCxYsULEzjuuqbWhJAlxBX1YmD3j77berdJsEQRAEQRAqCxEWwgV5+eWXcc8992Du3LnKquAIrQus+eE4FFVoh5mJaPVwdAFiJgpaPlq0aKF670eNGoXmzZujc+fOKjj/p59+ssfLsOffsdAgXZc4jq5M/J3blpaWZrecPPPMM/Zps7OzlZXAz89P1XSYPXv2edu3b98+9O3bV1ll2rdvXyjwn+zatQuXXHKJEkVMc0yXrsTERPvvzAfO7GT333+/SgRw8cUXKzcsMmbMGLVN2nfC4/Dzzz+rjGWCIAiCIAi1nVobvF1bYUM6x1Q9qT+9jPpSZwl69NFH8f7772PRokUYOnRouda/Y8cO1fDv3r17IXHAVL/ffPONy6rpJXVtoiBgHAWzeu3fv1+Nc7SK0H3r2WefxeOPP67EzR133KEqjrJmicbDDz+sltG2bVtVbJENf2YQYw0SCpohQ4aoTGNvvPGGEgM8Ntdeey1WrlxpXwYrwnPZa9euVd+Dg4NVHZNPP/1UWXccRRePA+tLbNiwQYoUCYIgCNXGuZNxyEk7h8z0ZGSnJiMnMwW5GenIz86EKScLlpx8eOX7wwh3WHVWWA1W2Aw22NiK9NBB726A3sMTBncPGDy94OHlh8gmLRHdsjO8fP57Fwt1HxEWVQxFRdunlqA62DPjYni7l/yU//7778pisGLFCtWodgUFh2MDnj368+fPdzntsWPHVMOaDW2NgwcPqr90eyoPdC1i+mAKJ1dV0y+99FLceeed6jMFAcXBqlWrCgkLWhuuuuoq9ZkuX4sXL8Ynn3yCRx55RNVNYbpix6BzFmVs1KiRcuFq2bKlGkfLCy08rgSS83axvgm3mcdFEARBECqCuL2bcWD9MqSeOorclGSYMjNhzcmHzWSBzQzYrG6w2XSw6Axwj2oMm5sbvp32ENzwXy0LHfQI9ohCuFcMGnl2RKhHNAw6hyKr5n8HkglYbRaYrPkw2/JhtuYj35qHo0v+xvrsr5GSfRw6qwk6WOHGQWeDmx5wM+qh8zTC6O+LgJhmaNl7GFp2GwSDFHOt1YiwEIpNxUtXH7ok9ezZs5CA0Bg8eLA9Dzbx8fEpcnns5afrlKPVpKqK8nBfNDTxwaJAzpXYNQwGg7Io7N271x50TiHi6hgcPnzYLiwYK1IavLy8lJuWIAiCIFyItKQE7Fy9CGf2bUXm6ZMwpWXAlmuB1cwmuwEmnQEWvbP1n989Cz7SaO7grex+7lTBFG56hHg0QLhHIyUm+FnvKCQA5JmzkGNOg0HnAYPO3f5Xm99D7wUPeNmnD/NsiNaBvWCy5iE+5yjOZB/GmZwjyLVkFUxg+nfIyEHiqd04vH43llrfgNFigh4WuLnboPfxgGdYOCKbd0bbjsMRHhsLfZAn3HRSH6ymIsKiGtyRaDmornWXhujoaOU2RPFANx5aMBij4AiFRGxsbImWx7gDNqIZ1KwV4tEa5Ixv0AoYukJzk3IUIlpl85JgdOoBobiwWq0lnp+FGOka9dJLL533W1RUVImElSuSk5MRFhZWqnkEQRCEuovZZML2P3/BofVLkXHiOCxpObCa3GB2MyJfb+ALzGFqT4BfHV5xejcDWvl2ga8hQEkKFXeo46CDTqeHm8EAvdEdBndPuHv5wNPgB0OqDnDy0tb5GuHRNAAezQoGQ7j3ee7UNqsNtnwLrHkW2P4drHlmWDJMyNh1BvkHU2HM90Ajn1ZqIGmmM0jIPoCErANIyToJm61g32xGb/h6BsPPGAQ/YzB81d8g+FmC4X7YE9bDqYjHJlis+cg0JyHXPQu+LRoguntneDTwgyHYC256ERzVjQiLKoY3ZWnckaqbxo0bqyBmTVzQPchZXJQUBmQTZn3SPvMvYxoYAzFu3Ljz4iwY20A3Iq3xzaxTQUFB6rNz3QmKFQaIl5W///4bAwYMUJ8Z+7B582blHkW6du2KH374QQVf05pRGihqXG0XLR25ubnFCipBEAShbpKfm4N1P83BsY1/IS8hEdYcC6xWA/J1RgerA//6FBIOOqsVRou5oFdfb4XOQweDnw+8IyMR3aQ7miZ0AJI0P6WiseXbkGnKRQ7y4GvzhN7Po0BE/CsmDGFeF4zLVILF0wCd5/nvRZ8u4Up4mE5lInd/MnL2p8B0MgMBxigEBEShRcBA6LwNMIR6wZycC2tm8Z2FOeYMuOu8oNe5I8A9CgEceRBIPXhA/W6xmZGDVCDEHZFd2sG7bSiMDXxKHVsqlI/a08IVqg3GETDrEsUFMx1RXPj7+5d6ORQHbKCvWbPGLix4wzOwediwYejfvz+eeOIJFW9BC8Evv/yCpUuXKmFDqwi3g5menn/+eRXXoNXT0GCjn/MxJqRTp04qhoFDSXnvvfdUjESbNm1UDEZKSorKJEXuuusufPTRR6q2BmMuGJTN2hvMlMVUvEVlwtK2i9t00UUXKVcwTRgxC1azZs1UFixBEASh7nLuVBzW//gxzu3eAXNKJiwmPfL17rDYO9M8C7kpudlscDebYHAzw80dMAb6IqBZc7ToOwJtew5zGYeQsy8ZyfP2w5Zjhs7HCJ/eUQU9+MpioV64BYOuQBCYLCbMWfGZmvehW++Dd1RBnGJFwvW4N/JTg/+wxrBk5iP3QApy9yUj90AqrNlm5B/PsE+v8zPCEOKlxAYH479/9cGeOLJ7PTb8Ng/mo4nwyguAjzES3p4R8HcPgb8xRLll+SIUSAYyV5xQQ44lHblhuWh22QB4twyBm0GSoVY2IiyEEsHCdY7iYsmSsgWgM6vSF198YbcEEMZvsJYFBcOtt96q4jroXqRletJ6/b/99luVcYnxEj169MBzzz2namJocPrbb79dWT6SkpJUbIhjytkL8eKLL6qBlhAKGaaCpfsWadCggcr0xMDvESNGIC8vT1lzaMVxlc3KEQqgBx98UAkTupcxPS7h/nB/BUEQhLrDqSN7sPbbd5F26DAsGfmwWI3IMxhVkLRR5wEPQzS83L0QoPeCp5sHPG06pOQfRrZ7GrwiwtGgY3d0H3k9AkL+S3RSHLQKpK84joyVx8H4azbig29oA0OAx4VrLa0o+GwMPd/NqTLQ+7rDp2uEGmwWG/JPpMOSnl8gJkI8XVo+NGK79FODIwknDmPdD7ORuGcX3DPc4W2IhJdHJEK8ohHp1RReen94Jfsj5Yt9SLDmIsl6DF6dotDmqhHQeZ8vzoTy42arqujZOkp6errK7MM0qs69+HRzYbpSFoxjbQShIICbmZjmzZtXKFi6vsEUu8y0RcsLr5/KQq5BQRCEiodNJ8YT5KflYMui+UjasQu6DMDg5ged0Reeeh+4M5hZ5wV3nee/nz3hpkwHLnADvDqEwm9wDNyjSh6rZ802KStF7v4U9Z1WisDLmpWoZ57CYubMmerztGnT7LGPdSHAfdWXryFxy3YE2Zog1LctGvjEwsvwnxu31WZFSu4xJHucQZOrh6Bpzx7Vus21ua3rjFgshCqFWZBosXAsLFcfYawIj0NligpBEASh/JjT8pB/JA15cWnIPHQW+clZMNo8VCYkEoMWiPFqAYeESMXi5q5TveWML+Bf1XMfl4acHYlq8GwbAv/BjZTloTjyT2Ui6as9sKTkAQYdgsbEwqdbRL0/5bT0jL7/pUJuaCs+fhGIy0G4sRXC/Voj0D0cIV5NEYKmwI+52PXNPCTq4tBs/AjE9Oxa749heRBhIVQ5rFBd32FMiSAIglDzMKfkIk8TEvvOQpf53290FvKAd8EHZidkzQZLNkyWLOTbsmDxtMK/cQyi2raF3i4eCgSEJiZcWRPyT2ci448TyNmZiNw9SWrwaBmkBAaDqZ3J2nQWKQsPAWarij8IuaEN3BtIITpXhEU3xfinP7R/P7Tjb6z/+jP4Jwch3KstQrxiEOjZAIFoAPyYhT3fzsc5fRxaXH8pGnRpXyHXVH1ChIUgCIIgCPXaIpF3MEWJicz9CdD9W2aB6P51m0nNP4uE3BNIzD0JU24yLLZUWP2sCOvcGQPG31PieIiioCgIua4NTAnZSmBkb0tA3oEUnDuQAvem/vAfEgOP2ECmPkLqL4eRtSFezefZOhjB17aUeIFSENuxtxo0VnzyJvI3nEK4RzuEeTeFv0ck/BEJ67wU7PnieyQZjqHF9aMQ0amVZJgqASIsBEEQBEGod1jzLTg1fxNsO3NUpWlHIZGSF4+E3OM4l3sCGZlxgD4TntGh6HzdzejUf1SlbZMx3BvB17ZSGZQy/jyhLBP5celI/GQXjA0LLBKmk5nKYsJp/AY3kmJx5WToLfcDtxTUD1nxyeswbzmLCPf2CPNpBn+PCPgjAua557D/iz1I9DmJHg9OgUdQyTNO1jdEWAiCIAiCUC/IycrEsjkzYd2Rgrb+l8DbGAg36JGcd0ZVh6aQyMyIAwxZ8IwOQ+frb6pUIVEUhmBPBI1poSwVGatPImtjfIGgoDuWlwEh41vBs1VwudbBjIbdu3e3f67vMIXvxbc/qj5TZCz78EVYtiUhwrM9wn2aw9c9DL6mMJx9cSPO5u9DxOjuaDSgm1gxnJCsUOVEskIJNRnJCiUIQn1n8/IfsX3B58g/mw69IRwdwy5GtE8L9VuWOQ07zy5GUv5ueDaMQJexN6HjRZegpsH6D5lrTsOcnIOAkU2V8BCqBoqMxe/NgNvOLDTy64Vgzwb23zLyE5Adk4Mu/xsPnVfd7auXrFCCIAiCINRLmG508ftPIXX3AZjzDcg1ukMHHVoGD0G7wItg0BlhtVlwNO8fGAeE4+qx76Kmw/oPASObVPdm1FtLxmX3P6s+H9+/DeveeAsN0BkN/dvDzz0cfvHAiadX4ywOodXkUfBv1aBeWzHqrrwSBEEQBKFesH/zaqz7/E3knUpBrpsnLHrGTHgDRiDMoyG6hwyHv0dBgLW+kTcirm6NmIj6naGQtTgyM1gF3ApPr38L5DmVNnOudKbTu8Hooa+3DeeYVp0RM+tTZcVYOPNB+Bz1RqPAnip9bQO0QcZnR3Ayfw30vYLRetxw1EdEWAiCIAiCUCNhjQebxarSqtrMBZ/d9DpVC+KPBbNweOnPMKdZkG3wBNjYNRQUlzOazfAzGtC+0RWI0rdR43Q+RgRc2hTeXcPrTMOYVbdzs03ITs9HTno+sjP4t+B7fo4ZpjyLGvJz//ushlwL8vPykRC2Ri0nJL4v3GwFAewXxA1KXHBw9zQ4fC74a+Q4Tz28/d3hG+gBnwAP+PBvkAeM7iVcRy2wYlz91Dvq84bfv8Ge7z5GQ69eiPZtgwD3BsBWYNWqZxFyQ3d07FfzXOsqE4mxKCcSY1F6Jk6ciDZt2uDxxx8v7+GvU7BoYNu2bbFlyxY0bNiwQpYpMRaCINRkbCYLUn+LQ97BVNhMViUcKCCg/loBpx5zZ+jSZLaZYbGaYLXmw+pmgs7LHYHRjWA+mwtbrlk1hH16RiLg4ia1Ki2r1WpDZkouMhJzkZ6Ug/TEXGQm5yrxoAmJnAzu9wUOUhHY3CxIilxXemFRDty9DEpk+Aa62wWHX4gngiJ9EBzlA0/f2nN+nElJOIVfn3sYDXLao2VIfzVuf+rfOJT2OwZMfQptug+qFzEWIiwq8WDX5kbdzTffjNTUVCxcuNA+7vvvv8cNN9yA559/HlOnTlXTfP755+fNe/DgQcTGxrpc7vbt2zFkyBAcO3YMvr6+9oJ5f/75p/rs7u6O0NBQdO3aFZMmTcLYsWNRn3jooYeQkpKCTz75pEKWV5uvQUEQ6jbm5FwkfbkHpjMOhSMu4LpDIcGK1yW1OBgb+CBwdCw8YopvDFUXtCSkxGcjPZHCoUA8qL9JuchMyi2xaPDwMcDbz11ZCbz83dVnD29aEwqsB5pVQfvs7mEA9Ba8M+sNNf8jDz+q3r8Kp0Pr+NVitv1nAcl1YQ1R48zIy7EgOy0PWal5yErLR2ZqHsx5lgvuB7c9OMobwZE+CIoqEBv86+VnrDVWJrPJhLXT30Zzc0/1/WDaZmw/txReHrkY+tAMxHbph9qGBG8LFc7HH3+Mu+66C7NmzVINfo2RI0fi008/LTRtWFhYkct55513cM0119hFhcatt96KGTNmwGw24+TJk1iwYAHGjx+vxMvs2bPrzRnlse3WrRteeeUVBAeXL5WgIAhCTSX3QAqSvt0HW45ZuSgFjm4OQ5CnqkqdnHgKq796HZlxp5BjM8CkY104C2ywwmCxwMMtFz4xURh408OIbtwWVlo68i3KwmHLtyorCK0fbu56eLYMqhF1HuiWlByfheTTWUg5k4Xkf4fM5Lxi52NMA3v0/UO94B/iqT4r8fCviNA+611U877gNuXn2z9z/pIsQ6cvcIPieksDRWF+rqVAaPw7UGxkp+YhLTEHKWeykZGcq6wwpzjsTy00v6ePEUFR3giP8Udk8wBExQYoi0dNdZMa+NxUpK87ibSfj6BFQDclhjclLcGi55+Ht6cJQx+biebte6AuIjEWwgV5+eWX8fTTT2Pu3LkYM2ZMod88PDwQGRlZoqNosViU1ePrr78+7zdvb2/7cugG1Lt3b7Ru3RqTJ0/Gtddei2HDhilLB12F3n33vwwe586dQ3R0NH7//XcMHToUTZo0wW233YZDhw5h/vz5CAoKwv/93/+pcRqPPvqoEi4UMFzn9ddfj6eeegpGY4EJ9plnnlGWmnvvvVd9Tk5Oxo033qhE0WuvvYbXX38dVqsV9913H5544gn7cjmeIuvIkSNKFFx++eXq2GkiivuyadMm/PPPP+q48aHeq1cvdOjQAV988YWapl27dmjQoIHavltuuUWuTkEQ6hSMCWBl6fRlx5Sbk7GRH0Kub4OzyYex/PVnkHPsHHJ1XrBodRXcTDCazPAx5CGgXUtces8L8A8KKbTMmuS1b7VYlQUi4VgGkk5nIuX0vwIiJa/YXvrAcAoHL/iFeqq/AWEUEV7KVUhXA4RReaG1wcPLoAZaIVxBy0fq2WwlvnjMNAFG601ulglnDqWpYfvKE2p6/1BPRDYLQFTzAEQ2D0RwA58adaz8+zaE3sOIlO8PoLl/Zxhtbvg7eQkyTHoseuYpeHmbMPKJV1VAeF1ChEVVwxQLpmxUC0bvguC2UsBG+Pvvv49Fixaphnt52LFjh3IZ0wryXIibbrpJuVz9+OOPSlhMmTIFd999t2rcs2FOvvrqKyUsKDo0+Puzzz6rYjgoZO644w4MHDgQrVq1Ur/7+fnhs88+Uw34nTt3KmsJxz3yyCP2ZRw+fFiJlcWLF6vPV199tRIMLVu2VG5b69atU0KB20VxoBUYevvtt5XbEae988471TJ5/Ah/69SpEx577DG88cYbSpTQ3cxRKJGePXvir7/+EmEhCEKdwpprRvJ3B5C7J0l917X1xpLtLyP7lnjk6DxhpZj4N/ja3WyCuzEfIV06YtTdz8PLp7CVu6aIiOQz2Th3PB3njmUg4XgGkk5mwmyyupzeO4BuPv+597AhrOIKfGpvXEFFwkDw8Mb+anDElG9Bany2Empn49Jx5nAakk5l/us2losDG8/+O78eEf8KjajYQPW3LJacisSnWwTcDG5InrcfMQGd4OHnhQ2H5iPHaEBGnh4LnpgGnwAbJrzxLXz8g1AXEGFR1VBUvPBfcZUq5fHTgLvrngJXsGH9008/YcWKFYUa7o5QcDi6NV1yySXKUuAKxlXo9XqEhxek/LsQbKizIX/06FH1nfEWFBbcJloxCAUC3aUcfS8vvfRS1ajXhBEb8atWrbILC1owNGjhYFwDrTGOwoIWiTlz5ijBQSvJ4MGDsX//fvz2229qu7isl156SS1XExb3339/oeU+99xzuP322+3CgseJQogih8t988031fzOsTkUPFu3bi3RMRIEQagNmM5mIenLvTAn5sAKC3Ym/IKDh/bCoqeY8FbTeJhMMHqaENW7Fy69Y4ZyKalJlhZaIuLj0nDueIYaEk8WpGp1hnEMYY38ENrQ1y4eKCREQJQNZpIKi/FTQ+veUXbXMp6L+MNpSmhQcNDV6sSeZDVo5yGmbTCadAxF4/Yh8PItnftWReHdKVxlMqPrXwRa4urLX8bv259HxvEU5BndkZYJfHbztWhwUSeMmfoqajsiLIQi6dixo8pURDco9qI7x0UQNrg/+OAD+3cfn6KFS05OjrI0lCYAi36Z2vQMPmZGKTb4KSyYPWnXrl34+eefz9tuDc5Ld6eEhAT7uHnz5inrAS0RmZmZKq7DuXFPYcDGv0ZERIQSRRQVjuMcl7t8+XLMnDkT+/btU4FOXC6Dp7Ozs5WrF+nTp48SMrSoUPT063d+EJeXl5eaRxAEoS6Q+s8JpP1wEHoYkW1Kx9qEBUjOjwf0un/FhBmNBg7CiMmP1RgxYc63IOFYQe84h/gjacjLMhcpIsIa+yH838ZvYLh3jYjruBB8n9GKrn2uTTC7VEzbEDUQBrnTiqEJjVP7U1TmrMNbzqmBEeiRTQPQpGOIEhoUe1UZDO7VPhQhE9si6as9yN+fhlGdnoTnU1GY9/hNyEy0INfogSMb92HWNYMxYOrjaNu79tbAEGFRHe5ItBxU17pLAV2M6EpE8cAgbVowHBvbmpAoKgOUM8z2xAYzYwvs2ScuEJPBDFM9evwX4ER3qM6dO6v4CMYz0JLSuHHjQvNpsRIafHjQAkHWr1+vYiqmT5+Oiy++WGX0orWC7lMXWkZxy6VV5bLLLlNuV8yaxRiLNWvWKHcm7q8mLDj92rVrlUhhHIgrGNNRXAC8IAhCbciMs+id/0PQgUA0D+ynRMXZnGNYn/AzrPnp8DXkIaJ3D1x29/M1QkxkpeUVNEqPFPSC0yJhtRTOyGQw6hDehK46mpDwR0CYV60QEa4wGAwYPXo06gKMrVACr5EfOgxqqCxMdE07uiMRR3cmIvFEphKHHP5eeETFZzTuEIqmHULRoGVglbhMebUORuhN7ZD0xR7k7k+B7UcbJr39E/bv+AurX5qOTJ0PsnQ+WPbq61gf8iaue31ejXQBvBAiLKoaKuRSuCNVN2y0M6ZAExeMOXAWFyWFgoDs2bPH/rk4mMqWqVevuuoq+zgGOjNG46OPPsI333xzXnzChWBsBPfJMeiaLlrlZfPmzUo0UKBoPT/ffffdedMx2xMtGjymFDYUR45ZtgitMEzBKwiCUNtYs2AO9vz4DbzMDdA2fDgiApuo8ftT1uNw8hKEdWqJKx6YU+0NJvZmnzqQgpP7U3BqXwrSzuW4jIlQgcHKbz8QoTG+0NN1S6jxUOxFNPFXQ68rmqmMU8d2JiJuR5KyZjA2Y+eqk2pgqt6WPSPRpk8UQhv5Vqolw7NFEEIntUPiZ3uQdygViZ/uQrPhndF05gKsnT8bKXuOwmRwhy5fhyV3PIGQdi3Qddg1LNiixJJnq2AYAmtmNiwNERbCBWnUqBH++OMPJS7YGKa4uFCBFFewF571KdiT7ywsaMmIj48vlG6WsRG0AHC9jmhB3LSWOGepuhAtWrTA8ePHlZWClpBff/1Vrau80GpjMplU5ihmg6JVgql5HWHcBLNP0Qp00UUXqSxSzCzFmItmzZrZjwNFygsvvFDubRIEQagKDu/6B3+8Ox36ZHeEB3bFRWF3IcA9VP1mtuZhv34Ner9wJ4aG/BfHVtXk5ZhxWhMS+1OQdMqpdoYbENLAtyCV6b8DU7vWltoJZYGuxnxvEVrk6/K++gV7ov3AhmpgvY0Te5OVJePoziSV4lYTGSHRvmjTNwote0aoNL6VgUezQITe0h6Jc3YhPy4d52bvVONboQ8Q3afwxClAyvwD9q8hk9qJsBDqBkwB6ygulixZUqblUBQwtSqFgSO0QHCgi1RISIiq5cBYCFfCYcKECSpQmn9LW/TtiiuuwAMPPKDWn5eXh1GjRuHJJ59UaWXLA/1UKRQY0D1t2jQMGDBAxVswTS1hrAWLCzLQnMKDMAUuhQ3jRlavXq3coxiYHhMTg/79C6p2CoIg1ETSU5Lw88sPIO/QOYT4dUYXvwkIbxZj/91iM0Pf3BvRV3ZFk4hh1RIjwdSkFBIczh1LV0kZHQlp6IuGrYLUwLoIHrWoKndFQFHB9xThe6skLsp1AdbhaNY5TA2MzaDI2LfuDI5sP6fiNNbMP4h1PxxC4w4haN0nSv2taEuVR2N/hN3aAam/HIE12wTo3Apc6vQFf7MyUpB26jjMdLm2WZmCDHoPIMjQBjUdqbxdTupq5e3KggHczKhE0cBA5rLAeIbmzZurehC0gNQlWL+D9TOuu+66ClmeXIOCIFRk3MTij57DmT/XIsijHaIDOqCBd3Po3Qz/FUELNiFyaDsVrKrzrFqnCNZAOLY7Ccd3J+HUgdTzMjYFRngj+l8hEd0qsNqyBNUUGP9XH4VFUbBWxsF/zmLf+jOqDokGq3637FXgKkWLRlXy/Yv34uzGfcg1Fpwbf28Tbv20bB275UEqbws1FmY8osWC2abK0ruSlJSk0sWyAV7XRAWPCVPq0hIjCIJQU9i/eTX++mAmzCluaBE+DENjHoW7/r/OshxjJsKHtIVP10gYqrAaMusb0K3p+O5kJSjSneIkfIM8CiwSrSkkguAbJB18QtEwHTADvznQckGBsX9DPHIyTNi+/IQaolsGosuIxohpF1wlrmNXP/a2sg7Oe/A65KXZMGTq8zX+FEqMhVDllDUwmXELdMVibQvGKdQ1mDXLsZaGIAhCdZGTlYmFrzyAtJ2HkK3zhr9HJAY0vhL+7gXpPfOQBf/ejRHQKwbuRVRSrjSrxC4Hq4T5P6uETu+mXJpi2oWgcbsQVUOiLscNCJUHLRMXXd0Cvcc0x4ndydi7/gyObk9U1xwHXltdRsSgRfeISs8o5R8UoqwUWekptaKIXq0VFqxhwDoAdIehbzozB9HHXau1sH37drz44osqUJg9waxLwGJlDJZ1hHEDDz74IHbv3q2ClNkbTj94oWYKEpraBUEQhMph3U+fYdf8L5GXY0C+wagqYTdnDEXwUOh1Btg8gdBr28KzdXCVpFllJpyzR9NxZNs5xG1PVMLC2SoR075ASNAywerNglBRMLaCdS84MLPUjpUnsPuv00g+nYUVn+3Fhp+OoNPQRmjbr0GlX3s+tUBUkFp5B54+fRrDhg3DuHHjVLpR+n4xmJeCQOvJZmYdVnhmpWMKBqYZZbAsRYgWOMz4BwbvUnB8/fXXqsI0g4ujoqJUgLIgCIIg1HUS449j0YsPIud4CrKNXnRaVa0DL5sRPSIuRpRPOzWdZ6sgBF3TEvpKjk0wmyw4uS8FcRQT/2btKWyVCFRCgtWUg6K8xSohVFlmKVoxul/aBLtWn8L2lSeRmZKHtd8fwj+/HkX7gdHoOLghfKrQHbAmUiuFxaJFi1RqtPfee89eM4CpPVlxmUXHmPpz8uTJheZhOk8WR/vxxx/twoLzMLBaK47Wpk0bZeFgmlMRFoIgCEJdZuXcd3Fw4Y/INXvArNcDSlQA3qYcRDZpix4hE4AMi8pUEzCyKXwvalBpVgoGzrLOwJHtiTi+JxnmPIv9N3dPvRIRTTuFKeuEh1etbLoIdQRmD+s2somyVBzYcBZblx1XlrQti49h2/LjaN0rEp2HxyAosvbULKtIauXdyTShzF7gWIKeQcGEwqCoStDM3MSKyBoUGrR8OEJBQetHcevmoEFriSAIgiDUBlISTuGnmfcj51jyv9YJb0APuJtN8PCyoP21N6K93yCkLzumRIU+xBMhE1rDvWHZCqNeqEgdXZwObU7A6YOpyu1JwyfQA007haohumVQlVRGrm+wDdW2bVv7Z6F0GIx65QLFuhdxOxKxdekxxB9Jx561Z7B33Rm06dcAPS9rWu8sGLVSWAwZMkTFRbCKMWMmsrKy8Nhjj6nfzpw543IeukIxxSnrBmiwIFtERESh6fidYoFpUTWx4ghTs02fPr3C90kQBEEQKovV38/G3h/mIjffA2bDv9YJmw0+lhz4tW6EsY+/B3erO5Ln7Uf6umNqHq9OYQgaE1uhaWOVmNiagENbEnD6QGqh2hIh0T7KKkExERbjJy5OlYzBYMA111xT2aup89CKp9XFOHMoFZuXHMOxnUnY89dpHNh4Fl2GNVIWjPoS/1Oj9pLigAXGimPv3r1o164dPv/8cyUumHuZcRPM/U9R4Ep179q1C1deeSWefvppjBgxolzbyPVxvRoUIYzhEARBEISaRFpSAha+cC+y4879Z50w0DphhoeXGZ0n3oqeF49T0+YeSMHZ73bBmmmCm1GHwCubw7tbRIU07osTE+GN/dC8a7gaAsLO78wThNpEVGwgLosNVBa4dT8ewtm4dBV/wYDvnpc3VdYNXQUX26tp1ChhMXXq1AtmZGKsBGEBMQ5nz56Fj09BSjlmhdJ+19izZw+GDh2qAreZ8cmRyMhINb8j/M5Cd66sFcTDw0MNgiAIglAT2bz8R2z+9D3k5hphMhjs1glvcw78WjXEmGlvwVPng/xjGUj9PQ75x9KRf7TArdcY6YPg61rDGO5drm3IycjH4S0J/7k5OYuJbuGI7RoO/1ARE0Ldo0GLQFz1SDd1/f+98DDSE3Pxx9f7VcB33zHNVTXvupoKuUYJi7CwMDWUBs2Vac6cOaq69fDhw+2/MYUs3aZuuukmPP/8+UVFWPn5t99+KzRu2bJlZa4ILVRcNVD6fbKQXt++fSv9sLKSN4P4t27dis6dO7uc5rPPPlOxN6mpqagqxo8fjx49eijBLQiCcKGq2D++8gCSN+9Gtt4bNreCzE5GsxmenmZ0ufwWtGkyCPnH05Hx4SGkJOWetwyf3lEIHNUUbkZ9mQ52fq5ZpYSl+8eJvcmFYiZETNQ8pPJ25eLm5qbqXDTrFKaySP3zWxxSzmTh1/d3qEJ7fa+KRXhjf9Q1apSwKA1MM8tGJ+tWUAw8/PDDqm5FYGCg3f2JooLB2HRdYjwFoduUJl6YZpbLYVEyZpFauXIlvvvuu0JxGPUVWo7obuYMj+fixYvVZ9YGYWNbC3ZnjQmeh9mzZ+Pnn39WdSc4zbFjBf66GtHR0Th58mSR69aydTmLilWrVqm4mg0bNqgYGC77kksuUeeXyyzpflEcLFy4EDUF1lJh4b+UlBT79UtoYRswYIBKgRwQEFCt2ygIQvVjTs5F+tKjyItLBwxucDPoYLbmIfn0UZjz8xGDLmgY1REWmxlullzo/dzRpHlfWM/kwrbNgtRthwotzxDhDY/G/nCP8YNHkwAYymA9sFisOLEnWYmJuO3nYM7/r2Ad4yRiu4tlQqjf6I06lUGqdZ9IbP79GHasOqmK7M2fuQktekSgz5jmKpVtXaHWCouNGzeqmInMzEy0bt0aH374ISZOnGj/nfUszp07p+pYcNBo3Lix6qEmbLxSRDzwwAN466230LBhQ3z88ceSavZfRo4ciU8//bTQcS/KDcxiseDWW29VqYApALp162b/bcaMGeo3DYq7oqA4odjjPI7w/N55553K+vTDDz8oUXH8+HFl1WC6YLrB1UZMJlORv7Vv3x7NmzdX1+9dd91VpdslCELNwZprRsYfJ5Cx5hRgPr9IaLCxEWB0Pa/laJb66+auVwLCvbF/gZho5AddGdO28jkdfzgNB/45q1w9cjP/e475h3mhZc8ItOwRUW/TbQpCUWlqaaVoPygaG34+olLVHvznLI7uSFTiov2A6CopOlnZ1FphwQZlcTzzzDNquBDsVacLjHA+FBGMQ7kQTL87YcIEbNq0CX/99RdatWpV6Hc/P78SLUcrbHj48GFVuFCD1g0G53NgjRENigv26GvuSTzftERs27bNPs2bb76pBopJ/q5ZYTTfRoogLoccOXJEiUxaRFq0aKEsJ85ucVw+rTInTpzAwIEDlRB1DN7/6aefVNYwxvY0aNBACaEnnnhCZd/Q1vv+++/j999/VwUZmZFD26agoIKqmpyHrlfk8ssvx9y5c0VYCEI9xGaxIWtTPNKXHoM1q6DxnoJT2HtiObL1eujdDNC76eFpdYOnnzvajrgaIaENYTNZ1QCzFfogD7g3DoAxwrvcjRbm6t/39xllnchwcKXy8jMql4+WPSMR3kSyOQlCcfiHeGH4pHboPDQGq+ceQPyRNPX34KazGHxD61ovyGutsKitsKcnx5xTLev2MnhVeLAQLUYUAWz8r127ttwZsihMWrZsqcSIxvz585UvKF3WXOHoPlQcDz30kMoqxkxemiWGdU1YyZ1QALz66qtKVPAzxRILLmqiIDs7W8XqUNSyjgotKIyD4H5r237jjTfi7bffRv/+/ZVAYtIAQuuaBgUO3fYoeGi9ueKKK3DVVVdh//795yUO6Nmzp1onxZskDRCE+gOzNKX+egTms9nqe6Y5ETvPLsfx/H9dS00FqWIDO8biskffgrtn5QRB5+WYcWjTWexbH68aQBpGDz2adQlT1omGrYLqfKYbQahowmL8MPahrtj55ymsX3gYZw6lYd5z/6DHZU1Uelp9Lb2nRFhUMRQVvb7phepgw3Ub4G0seaYPujUxhsWRxx9/XA0azz77rBIBbLAXFXj/6KOPFsrI9cILLyjrgysYj8GefkcOHjyoGtxRUVEoD9wXNtrZSHdlQaHw0CwltDowrTGFBV3tNLclumn16lVw/mhpYLV2uuVRAHAepkymxYEwQxmPDwWRo7BgNrNJkybZv8fFxam/4eHh54kkHguKKsYI0Y1PEIS6jelsFtJ+i0Pu/hT1Pd+Sjd3Ja3AwcztssMJgscDTPQ8dJtyMvpffWCnbYLXacGpfCvauP6MK2Flo/VAWVyCmXQha9Y5Ek46hMLqXLchbEIQCaEXsOLghmnQMwZ9f71dV5/9eeES5GA6Z2EaJj9qGCAuhSBhQ/MEHHxQa51i5nLAuyPLly5VYcHRTcoSuQ45phENDQ4tcJ4Oymd3LEVp5qiItW8eOHe2fNRGTkJBgFxa0XDBLkwbHUwhQVFFYbN++XVkvHDOQMfYkNzdXWTu8vQtEXffu3Uu8TZr1gvMLglB3sWTmI335cWRtPANYAavNjINpW7A7bR1M1jx4mvLhGeaOS6e9gagmBc+kiiYlPgv7/o7H/r/jkZWaZx8fFOWjAk9b9Yqsd1WEBaGq3KMuu6cT9m+Ix5r5B5F4IhPzX9yELiNi0GNUE1Xlu7YgwqIa3JFoOaiudZcG1geJjY0tdhrWCLnnnntUAUKr1aqC4J2hkLjQchyn3blzZ6FxdI1KS0tTVdWLs1qwOCJFSEmDo50xGv+LftSEDPepNG5htFqMHTv2vN8cxRKPa0lJTk5Wf0ubhlkQhNqBJcuEzLWnkLHmJJBf8Pw6mbUf25P/RKY5Bd6mbIR1bI5rnngfBodnVEXBFLGHNiVg77rTiD9SUMuCeHgbVAB2675RUgW7jsJ3Jl1/tc9C9eLm5obWvaMQ0zYEf807oKwWWxYfw5Gt5zB4Yms0iC2Z23d1I8KiGi6c0rgj1QZotfjll19UrAAb9owxKCtdunRRVhJHK8XVV1+tXIxefvlll1YRBm/TcsDGN12GHOd1DOQmjI2gFaEsmM1mFaBO6wRhTATXTXco0rVrVzWupCLKcZuIq+1i2mRmKyvOyiMIQu3Dkp6HjNWnkLH+JNwsBc+r5Lx4bEteiaTsY/DW5aDzlaMx9Lr7KnzdfEYmHMvAnjWnVVYaU57F7pYR0y5YNW6adgxVaTKFugut8HTNFWoW3v7uuPjW9mjR4xz+/Ha/Spqw4NUt6DAwGr3HNIe7Z81uutfsrROqFcYiaPU/HB9Erhq5w4YNUzEZzGLEXn7GIpTV/Yo9/yxuyHSrhAHhFBR33323CrxmgDQzOTFgnIHUjJ1gyllm+GKKYQoQihHW22D2JcZnaHC+JUuWKAEQEhJSqvoQtGjQOkPhxOPA7endu7ddaDz11FO47LLLEBMTo9bPHiC6R1EcPPfcc0Uul7ETFEI8fpdeeqlyf9JiWxgQTuEmCELdwJyUg4zVJ5Gx8TR0Nh3c4KYExZ7U9TiXvhue/sDlT05HbMfeFb7uvGwT9m84iz1rTyPpZKZ9fEC4F9pe1EDFToirkyDUDJp1DlOF9Nb9cAh71p5RQd5+oV7oMjwGNRkRFkKRsGHu7HrEVLL79u1zOT0LErIuCBvXWj2K0sLG/pgxY/D1119j5syZ9vHMwESXKGZt4u9agTyuiwXyCC0HTOXKeA8GTTPTEgOyWbBPg/U0WJCOcQ4UMI7pZi8EYyQYiM4enlOnTqnMT5988kmh4oEUB6zB8dJLLykhwjgMFrgrDhb30wK/GdRN4cR0s4zNYHpbrSChIAi1PCh75XHkbE+AG3TqX0LuCexNXY+0tL3waRqMyW/NhY9/QdrpioLPYmaboXXi0JYEeyC23qBD865haNuvARq0CKySODZBEEpf+2LwxDaI7RGBnatOqkDvmo6bzdkpXSgV7EFnrzdjABx7xgkbhsz4w0J8zgHJQtHs2LEDw4cPV+lanbNS1SfoErZgwQIsXbq0zMuQa1AQqpf8kxlIWXYEpv3/xS+cyT6iLBQ5mQcQ2qsDRj/4aoXHT7BoHbM67V17Ginx/yV/CG7go8QEA7E9fSo+ZkOoPTDjIDvrCDvhNLdcQShNW9cZsVgINQ5mZ2KPP0VZhw4dUF+hxeOdd96p7s0QBKEM5J/IwNmFu+B2yqy+sw/vZPYB7Eteh3xTHFqMvQpDxs+q0GPLdZyNS8eu1adUQLbFXGCdMHjo0aJ7uBIUEU38xTohlCnBiSCUBBEWQo3EMT1tfeVCLlSCINRMC8Wp7zbDmKAHnYusNiuOZ+7BgeR1MOvPoO9dj6B934qNm2LwNYOwd/55UqWp1Aht5Iv2A6JVVWx3L3ndC4JQ+ciTRhAEQRDKSf6pTBz9ag28U3xghF4JiqOZu3Ao8U/YArNxxavvICy6aYUe5+QzWco6wboT+Tlme+xEbPdwtB8YLdYJQRCqHBEWgiAIglBG8k9n4uAnyxCQFQpv+ChBcSxzN46cWwW3JgZMeHkOvHwqLlbMYrEiblsidv15EqcOpNrH+4d5oX3/aLTpGwVPX4mdEAShehBhIQiCIAilJOdEGg58+CtCzI0QgFAV33Asaw+OnFsBz45huP7Nbyo0IDs7PR971pzCrj9PISstX41jIqcmHUOVu1OjNsGqDoUgCEJ1IsJCEARBEIrBZrXBmpGvalDkxqfj8E8rEe7WHCFopATF8ay9iEtajqCB7XHj7fMr9FieO56BHStP4MCms7CaC5I4evm7o12/BioY2y9YMg4KglBzEGEhCIIg1HsoHixpeUo8mJNyC/4mFvy1JOfC9m/9B0JRQU5k7sWR1OWIuWo4bhgzv0LdnY5sPafy1p85nPbfehv7oeOQRojtFq5iKQShPLB2CQu0ap8FoSIQYSEIgiCgvhevS/xsNywpeUVOw9iJLHMqMk2pyMxPxJmcTWg35QZc3/+7CtuOnIx87F5zusDdKbVgW3Q6NzTvFo6OQxoismlAha1LEJjSXDIwChWNCAtBEASh3mJKyMa5j3bCmmkC9G4wBHvCEOKFfPc87Fr3I1LSkpFiTUe2OR16iwle3iYMfWQmBrefWmHbkHgyE9tXnsDBjWfttSe8/IxoNyBaBWT7BHpU2LoEQRAqExEWQo2sBtq2bVt88cUX6Nu3b3VvTo1i8eLFeOyxx7BlyxbodOIKIQjlwZSYYxcVxigfhN3aASdP7MbPMx9AboYOJkPBK9JoNsM/ABj11JuIatK6Qg66is3Yk4xty47j5L4U+/iwGD90GtIQsd0ioDfKPS4IQu1CnlqCS2gepc+l8zBy5Ej7NE2aNMGbb75Z6EX50EMPqXLvf/zxh30a52U0bNiw2KM+a9YsNG3atJCocJzfx8cHLVq0UNu4efPmenUGefxpvv7666+re1MEoVbD2InE2TtUULYx0hsZvbLw8W2XYMET05CR465EhYfJhMBAG657bzamzFlcIaLCbLJgz9rT+HbGRix6Z7sSFXRvb941HGMf7oZrpnVHq95RIiqEKunEe+WVV9TAz4JQEYjFQii2Efvpp58WGufh4dokb7FYcOutt2LRokVYtWoVunXrZv9txowZ6jcNvV5f5DopTt599101jzPcFm5Tbm4uDhw4gNmzZ6NXr16YM2cObrzxxnpzJimo3n77bUycOLG6N0UQaiXm5Fycm70TlvR8WPxtWLbpGaStscCi9wT0gIcpHz7RPrhqxufwDwqpsPiJnX8yXexJ5GSY1Dijpx5tL2qAjoMbwj/Uq0LWIwilITs7Ww6YUKGIxUIoEoqIyMjIQkNQUNB50+Xl5eGaa67B8uXL8ddffxUSFcTPz6/QMsLCwopcJy0Qhw8fxqhRo877LTAwUM1PK8iIESPw/fff4/rrr8fdd9+NlJQUZGVlKWsJxzuycOFCZeXIyMjA0aNHldXjxx9/xODBg+Ht7Y1OnTph/fr19umTkpIwYcIEREdHq987dOiAb7/9ttAyBw0ahHvuuQf333+/OiYRERH46KOP1DZMmjRJ7XNsbCx+//33QuLrlltuUdYYLy8vtGrVCm+99Zb9dwqmdu3a4bbbbrOP47HgsiieNC6//HJs2rRJ/SYIQukwp+bi3Ec7VAaoTNM5/LbjHSSbbLDodfA05SG0oS9u+fJHTHpnQYWIClbHXvXVPnw+bR3+WRSnRIVvkAf6XhWLm2ZehH7XtBBRIQhCnUGERRXDHnlrdna1DFx3RZOZmalEwJ49e7B27VrVWC4PFCYtW7ZUjemS8MADDyjBsGzZMiUexo8ff56Vhd+vvvrqQst84oknlNvWtm3b1PooJMxms72BT3H066+/YteuXaqhT+vAxo0bCy33888/R2hoqBpPkXHHHXcogUUXLsZAUPxwPq1HyGq1Kjew+fPnq+P11FNP4fHHH8d33xVklfH09FQuTlzuTz/9pITIDTfcgOHDh2Py5Mn29cbExCghw2MlCELJMafl4dTbG1T2p4z8ZKw48x2ybTnwMuUhvGkg/jfvN9z02txyV8rms/b0wRQsem87vp2+AXvWnFZB2UwXO+KWdrjhuT7oMjwGHl7iNCAIQt1CnmpVjC0nB/u7Fu7RrypabdkMN2/vEk9PtyZf38IvWDaEOWg8++yzqsG+d+/eIi0Rjz76KP7v//7P/v2FF17Avffe63LaY8eOoUGDBiXextatC3yeaYkgU6ZMUQ37M2fOICoqCgkJCfjtt9+UNcURigrNKjJ9+nRlKTh06JBaHi0V/F2DomHJkiVKAPTs2dM+npYObb+mTZuGF198UQkNze2LwuGDDz7Ajh070Lt3bxUbwXVp0HJBSwmXe+2116pxnTt3xnPPPaf2gyKJx4PnwRkeI/4mCELJ2LliMdx/y4WPMQSZphSsip8Lt9wkRLSKwvgZn1RIlWzWwojbnogtS4/hbFx6wUg3oGnHUHQeHoOo5gFSL0AQhDqNCAuhSOgqxIaxI8HBwYW+s1eejXaKhTfeeMPlch5++OFCubLZ+C6KnJwc1XNfUjQrjFbchw1/igT2+jN70ldffaUKAA0YMKDQfB07drR/pgAhFCEUFrQUcH/Y4D916pQKaqO7F92iiloG40ZCQkKU25QGrQracjXee+895dZ0/Phxta9cNsWEI1OnTlXuW4w1oSsVl+sMXanEN1YQLsz2v37F1lkfomfEZPi4U1SkYv2Jz+Af64fx03+rEEFhMVmxf0M8ti47jtSzBRZKFrBr3ScSnYfFIDCi5B06giAItRkRFlWMG33rt2yutnWXBroWMU6gOIYOHap69K+88krl6uMYM+AoJC60HMdpd+7cWeJtpKVE6/3XYG8/G/AUFnSDYsyDc1VRWg80tN+4/YQZMrgfzHhFocDjwFgK56wZjsvQllPccufOnassIa+99hr69OmjLD1c14YNGwoth0KEwekUKwcPHiyUiUsjOTm52FgVQajv7Fq3FGvffhEWBGNgg0nwdw9BtikNuzyXYuLX8ypEUOTlmLF79SlsX3EC2ekFzwcPbwPaD4hWFbK9/d0rYE8EQRBqDyIsqhiVMrUU7ki1AVotfvnlF1xxxRXKgsCMRWWlS5cuykrC5TiLAVew8c+A7WHDhtnHMS7hkUceUdvBWIabbrqpVNvAWBEKJS5HEwZs6LO2Rnngcummdeedd9rHuQrAZjwFBQ0DvelWxX1r06aN/XfGgHA+HitBEAqzf/Nq/PnaM8gyecJgCMHgqAkIcA9FHrIQ88hAtIy4rNyHLDMlDztWnsCuv07BlGtR4xiQ3WloI7Tt1wDunvJqFWo+fMdqrscled8KQkko19PPZDIhPj5euWSw99TZTUao3dD9h+fXEYPB4NKViY1fxgIwYxEb4nTjKav7FQPCd+/ejfbt2xf6LTU1VW0Pt4sN/Q8//FC5DLGQHjNGaTBL09ixY5ULFkXPhepmOMMaGcwstW7dOrWs119/HWfPni23sOByua2M16CF5csvv8Q///xTyNpCSwvjLhiX0ahRIxVAzsxXf//9N9zdC3o/+ZkZu2j1EAShgMO7/sGKmY8hO88DFr03/D1C0T98NHzdQ6DzMyLmfwNhLGdK19SEbGxdcgz7/o6H1VLghhkU5YOuI2LQokeEcn8ShNoCLeyOqeAFoVqEBTPw0G+dbh3MhkP3EK13mQ04NuSYRadHjx4VsoFC9VZ51uIPNJj1ad++fS6nHzJkiGoIX3bZZfZ6FKWF8QRjxoxR2ZFmzpxZ6De6NBHGYDDAul+/fuoa7Nq163nLYW//N998UyibUklhQPaRI0dw8cUXq7gKXs+jR49GWloaysP//vc/bN26FePGjVP3CzNR0XqhpaTlcaUY+uSTT5SoIO+//76K5XjyySfx0ksvqXFMfUux4RzzIQj1kRMHduL35x5ATrYRZr2XqkPRzKM5uja4EnoYoQ/0QOjk9uUSFUmnMrF58TEc2nQWWnK9qNgAdB3RGI3bh8BNJ729giAIxM1Wihyk7Ll9/vnn0bx5c9UzzUBZmtEYSEqfb6bmZApM9iKzcNk777yjemnrMunp6QgICFCNTrrkOEKXlbi4ONUjXZqA5PoOe+uZYpXuPs5ZqUoKrQFMRXv69Gl7T39dIDExUYk71rFwtHQUhVyDQl3lzNF9WPTMvcjO1MP8b9FNL5MJnWJHobGuwE3QIzYQwRNaQ+9TtngKZnba9PtRHN2RaB9HIdFtZGNExf5nJRUEQaivbd1yWSzotrF69WqVdccVFBrsIZ41a5YKmqXIqOvCQqh42EPP3nmKMscsSyWBbnlMNcvUr7QQ1CVRoaXVpRWjJKJCEOoiifHHsfCJ25CdCpgM7vZK2UExERja+kGYj2aq6fwGNoT/iCZw05fOmsC+tlMHUrH596M4uS+lYKQb0LxLGLqNbIKwmJLV2BGEmg7d2el+S+66667zEpIIQlkolbBwrj5cFPT/vv3228u0QYJAHNPTloaXX35ZWdWYXpa1Jeoa3bt3V4Mg1Dey0lMw79EbkZVgQr7BoN5eHiYTfKK9MfauD5G94JQSFW7uOgRd0xLeHcJKLSiO7UpSgiL+SEENCro4teoZga4jGyMo0qeS9kwQqgde85qLb2UU0BXqJ5K6QqhTPPPMM2oQBKFukJ+bg28fvxEZR9OQZ3RnBgm4m03wCXfH1S98A/0hE1K+PARYbDCEeiFkYhsYI3xKVdTuyPZz2PTbUSSeKLB2MAi7zUVRqjq2fzkDvgVBEOoTZRYWDGadMWNGoSJhgiAIglARmE0mzJt+G1L3nESu0QMwusNgtsAn0IbRz3+EkNCGSP3lMNI3FGSu82wbguBrW0JXwlSvFBSHt1JQxCHpVJYaZ/DQqxoUnYc1gk+Ah5xIQRCEqhIWl156Ka666ipVNXj69On2VJysKHzJJZeodKGCIAiCUFq+f+k+nNuwC9lGL8DoAb3FAm8vEy559g00atkB5rQ8nPtwB/JPZKj4B//hjeE3qFGJsjNZKSg2J+Cf344i5UyBoDB66tFxcEN0HhoDT1/xMxcEQSgrZU663a1bNyUmFixYoKwWLPzFWga9e/dWWaIqmy1btqjMQaxfwBSlTAnK+gcaSUlJqmIxs1Yx5oPpO++++24V2e7IH3/8odKVchpWh/7ss88qfdsFQRCE8/l11nR8MHYIjm05rESFzmqFry4bl0ybhtu+XK5ERd6xdCS8s1WJCjcvA0Jvbgf/ITEXFBUUFAc2xmPujA1Y+sluJSrcvQzoPqoJbny+L3pf2VxEhSAIQnVZLCZOnKiqATOgm0XTmIP/tddeU9lqli5disqEKUQpYlgPgLUSKBbuv/9+FfDLwmZEp9Op6snPPfecKt536NAhlfWAaXFZ34Aw69CoUaNUoDnrJqxYsQJTpkxRtRtYw0AQBEGofFZ/Pxt75n6LLL0PYPRmHnT4uGWj+//uQbdhY+3TZe88h+R5BwCzFcYoH4Tc0AaGkOI7sqwWKw5uSlAxFKlns9U4D28DOg5phE5DGsLDWywUgiAI1S4smPbyl19+UTUtNJje87rrrlNFvj7++GNUFqzwzLRoTJNGAUGY4paWEwoIWh5YMfmOO+6wz9O4cWNVjOyVV16xj+M8FEIURIRCac2aNXjjjTdEWAiCIFQyO9b+jvVvv4wsmzdsFBUUFNZstB0/AQOuvs0+HTPWZK4+hbTf49R3zzbBCB7fGjqPgvoVRVkoDm6MVy5PaQk5dkHB+IkOgxvBw0tylwj1GxZqZcer9lkQKoIyP1lZAO+nn37Cgw8+aB8XHByMt956S7kWVaawyMvLU/UJNFFBNPcrCgMKC1dWjh9//BEDBw60j1u/fr2yfDhCSwWtH4IgCEIlVsue8QCy89xh0fmoOAkfUzaaXDocI6cUThNts9iQ+vMhZP0bpO3btwECLmtWpOsTg7IPbU7AxkVxdguFp48RnYc3QoeBDZX7kyAIUB207HAVhBoRY8ECZv/3f/+HSZMmYcOGDcjPz1fFVuiK5ONTufm+hwwZgvj4eGV94HpTUlLw2GOPqd9YHM2RCRMmwNvbG9HR0apaoKPg4TIiIiIKTc/vdK3KySno4XIlavi74yBUPfv370dkZCQyMjKqZH2MvWE8T3HQFY/Z0qoKXvtNmjRRVbgFoaZhSsxB8vcHcObFjUj+br/6npaUgI9vvRQLpj2MDJMnLDodvPJz0ahTE9z+48rzRIU1z4ykL3YXiAo3KEEReEVzl6JCZXnakoC5z21UMRQUFR4+BvQe3QwTn++jituJqBAEQaihwoJVtleuXIkDBw6gT58+ymJAQfHkk0/ivvvuK9MyKQ5ojituYCwHK39//vnnyoWJooENTLo0URQ4WjEI3ZoY6E3ryuHDhwtZWMrCzJkzVVlzbWBQeF2EjWRXx58B8aVxl+M827Zts4+jEBg8eLAK/D958qR9GufhhhtuKHbZLH53zz33wM/Pr5C7xOzZs5U1zdfXVwkBFpN78803VUXuksLGOuepSbA2BzOwOUKr3UMPPYRHH3202rZLEJzJP5WJpG/24uxrm5C96SwsqXnI3pKA+Fc2YveT82HJDoHJYFDVssMa+uK2737HtY+/e95yLMz8NGsHcvenwM2oU/EUfv2iz5uO933cjkR8N/MfLJ69C8mnC4Kye17eFDc+17dAUJQwBa0gCIJQPsr1tGUGqLVr1+LUqVPYu3evquDIxo9j3EVpmDp16gUrLjdr1kz9ZSwHh7NnzypBw8bo66+/bv9dg6KDQ+vWrZWrVv/+/ZX4YYA2x3N+R/idlo2iMluxQesoTmixqKvigiLi008/LTSO2bPKyrlz51QqYoq/v/76S2XzorAgy5cvV4JRo7jMYkxpzDibd95557yEAnR3oyWNQf30Hd2+fbsSCRQLVWlNqCjYaLJYLEX+fv3116v7humdHY+fIFQ1eXFpSF91AnkHUuzjPFsHY3vczwhICkekb0vE+HdEI7/2OJ2/By3vuhJBsQ1cLiv/TBaSPtsFS1o+dL5GhN7UDu6N/utE0O6N43uSsfHnI0g4VmC5NHro0WloIzXQ/UkQhKKhl8lHH32kPt96663KNUoQykuFdOPQzYhDeWFDUAskKimaK9OcOXPg6empUtAWhdVqtbszEVpafvvtt0LTLFu2TI0vCjasy9O4rk1wPym+ioJijg+lX3/9FUuWLFHXAK1IV1xxxXnTnjhxQp0bTkPrES0KjlBkFLcuR7777jt06tSp0DXHcczstXDhQpUNTIOCgtujuawNGjRIiV9HiwQFB60bdHfi78eOHcMDDzygBq0Bo8H9ZAwO96dfv35KeFGkOsK6LhQ2vM4oft9++21lXdCuQboR0rJCV7yWLVsqoXv11Vfb0x/TosPrkgJp586dalouUzvmhOulCGeSgosuughz587Fs88+W6LjJwgVBe8NWhQyVp1A/rF/3ULdAK9OYdiVtgJ7f/4BWUZvNTo0JRztQ/siwqsVoj3aI+uTw7B2SlepYo3hBdOQ3AMpSPp6L2x5FhjCvRB6c3sYgj0LrffkvmRs+DkO8UfS1HeDu66gDsXwGHj5FtxrgiBc+P5lh5/2WRAqglprH2bDjbUz2EClGGAmqhdffNHuB8+GGa0PPXr0UNOwR5fTsBHGxiZhmlku55FHHsHkyZOVaxcbqGwoVxa8ec35BQKnquHLt6IzP7DB+/LLL6t4F1oQ2IPOhjmtQ47xEDz2dEtieuLyCjNaO7gsRygqWrVqVUhUaHCf6bZWEmjxoGhhXRT24DhCd6pXX30VX375pbK60F2LrkhctwZTFlPgUiDQGsMYJIqm559/3u5K99VXX6mMZC1atMDq1avVciioHRML0C2Q66IFjsujVWLx4sXKskMc94duiTwmglBVMKA6Z9c5ZKw6CVN8QZE56N3g0z0CJ70O4tc5TyET3oVSx3a7azza9x2B/JMZSF9xHLl7k5Gz7Rxytp+DV8cw+A+NQd7RNKQuPARYAY9mAcr9SeeQDjY+Lg1/LzyCU/sLrCJ6ow7tB0aj64jG8PYXQSEIglDd1FphsXHjRjz99NOqKB7dnD788EPlCuPoSsPedPY6s+eY7kpjx461B3kTxmVQRHAaZrNq2LChCu6uzBoWFBWz7/sT1cFtbw1UrgIlhe5GzpaFxx9/XA0a7DVngDx54YUXVO88z41jLMaNN96oBN38+fOh17teP0WiY3wMG8pdunRxOS2Fi7OwOHjwoBIW5YWCiNvI2A1nCwrNxhQEmqsfCy7OmDGj0DS0TNB6xtgfuibxd4oqWhM4P48RxYFmFaNwYCYzXr+OwoLzOVrfeB5YL8aVVYdFIHlMBKEqMKfkIvGz3TD/m3HJzV0Pn96RyGtmxvwX7kB2tjssTB0LwMechTbjr8XAa/5L/e3e0E+5NjEWQwmMPUlKXOTsOAf822nq3TUcQWNbwM1Q8ExIOp2JDT8dQdz2RPVdp3dDu/7R6DayMXwC64cFWRAEoTZQa4XFF198UezvdCdZt27dBZdD15etW7dW4JbVHXgMP/jgg0LjHC0RhLVDNBjrwviUhISEQtPQFYkuSrQGXHPNNS7XNW/ePFVHRKO4uBVm7GIvviNVYcalWHCMH6ILlPO+0trB6TQoICh+6TrFv7R6OLvrMbuTs4hyFk7FQRFdmuB0QSgrpoRsJH6ysyD2wdugUr/qOwdg3tM3ISs+D/kGT0APlekpvHc7XP3oW0Uuyz3aF6E3tkX+6UxkrDiOnN1Jarz/sBj4DY1Rlsb0xBxs/CUO+zfGK9FBg2ur3pHocVlT+F+gMJ4gCIJQ9dRaYVFboTsSLQfVte7SQKHgqiaII87BXmwMaLEsGk888YQSIIw3oAC49tprz1sOhcSF1qURGhqqUgw7wlgFZgy7ELSKOIsQWhJKgqt9LY2gobAgtJI5xyQ5u4eVJmUzq8mXNjZJEEoLXZgS5+yCNdsMQ5gXQqd0wA9v34fE9w8j1+gBGAxwN5ng19gf182cD3fPkjX83Rv4ImRiWyVarLlmeMT4IystD5t/O4rda07Daim4x5p3CUPPK5ohOKpy05kLgiAIVSQsGK/A1K0lhY0uNr6Ky2pT3+DxKI07Ul2BAcps1DMGg9fFuHHjyrws9u7v2bOn0DiKlvHjx6vAcOc4C66PwduMS2AD3LHWCa/NXbt2KeuMoztTWa9ZZqGiRUXLavX3338rNyYKJ1p7KCCY1crR7akkFLdN3P6i3MYEoSLIPZyKpM/3wJZvgbGhL/b4rMb+SY8UBGYbPaC3WODjZ8WYmR8hNDKmTOtgAHdulgnrFxzGjpUnYDYVdFA0ahuM3lc2Q3hjfzmZgiAIdUlYMDC1pL27Qu2HsSnMXOQI/fxpMSgLtFwwfoHiglYNLTajtDAGZsqUKaqhrcVs0AqyYMECtUxmUxoxYoQSEcyqxFomrHnB7E8srsh0wbQa0K2JKYpTU1MLLZ/B/QyqplChECjN/tKt6ZZbblHbwOBtxgExFoOiinEbDPZmTA/3n1mlmKKZKZvpQnbTTTcVuVxuU1xcnKoJwlggLkuzcjAeRTJCCZUFXZSSvt0LmG0wh9mweM1jSLcaYHMIzB40bTpadRtQ5nWY8i1KTGxdehx52WY1LqKpP/qMbo7oVkEVuDeCILhKbFLRiV2E+kuphAUbP0L9gVmInFOpMkC6JC5HRcHgeTayGWhPSwKDtksLa2FQ4DAIWgu050Pxm2++UalZGTzNLEychpmXGDyuTcfsX7QqcBx/ZyPf0VqhBU7/73//U8KD4qo07k5Dhw5V6xwwYICal0KHxe00KAAoeJgd6siRIyqLWdeuXQsFxLviqquuUjEq3FYKIS3d7Pr165U40dLVCkJFkrX5LFJ+OKCyNJ3N24+1fy+ESWdUKWV9zNloM/6aQoHZpcVqsWLf3/EqjiIrtSANeHADH2WhaNIxVBo7glCJ0L2X6dMFoSJxs0ny4nKhudiwcecsvHJzc1UvM7NPOQcbC+Xjvffew88//6zqStRn6FLGgPGihIlcg0JZyVhzCmmLjqjPx9K2YUPyUthgg6cpD2HdW7msll1S+No5uiMR6xceQcqZgnS1fiGe6HVFM7ToEQGdTnpPBUEQakNb1xkJ3hZqJbQosOc+IyNDuQXVR+h21aFDB3shP0GoCFRM0rJjyFh5Qn3fn/YPtiWvhNFshm+4Oya8ugBePoXTUJcGFrVb9+MhnDlUUNzOw8eA7pc0QYeBDVVdCkEQBKH2IsJCqJXQjYkxG/UZBnQzlkMQKgqb1Ya4j/6Ae1zBq2FnymrsTV4HX2MuLpnxGmJadS7zslPis/D3T0dwZGtBpV+KiE5DG6HriBh4OBTBEwShamDM7GeffaY+07XWOfOhIFSLsGCdgzvuKLuPrSAIglD9pCacxf4XfkC0ZwdltdiStAxnUtag5w0T0W/M5DIvl6lj/1kUhz1rzyjhwhjR1n2j0POyZvANkuJ2glBd8D4/ffq0/bMg1AhhwerWIiwEQRBqLwv+7wE0y+iBaK8OsNos2Br/E0xtLLj9sVVlXqYpz4KtS49h67LjMOcXpI5lQHbv0c0Q0qDsrlSCIAhCzUVcoQRBEOopq+fPRt7iU+gacgX0XgaYLHnYmfcrRrz/bJnjKKxWG/atP4MNPx9Bdlq+PXVs37GxaNAisIL3QBAEQahTwmL37t3o06cP2rVrp4b27durvw0aNKiYLRQEQRAqlFNH9uDP6c+hffAYBIe1UeMScg4g9MYuuKLbG2Ve7ok9yVj7wyEknSqoMu8f6ok+Y2LRvGuYpI4VBEGoB5RbWLCuAWsHUGCwAvD777+v/jJjT+vWrVXxLkEQBKH6MZtM+Oqh69EwuzP6Rd4GnZse+ZZsJDVJQPc7J5e58Z90OhPrfjiM47uT1HcPbwO6XyqZngRBEOob5RYWrHzMlJccnPPn7927t7yLFwRBECqAX2dNR/rqw+gROR4BIWFq3DnEoe20q9As2KdMy8xOz8fGX45gz5rTYOwn60+0HxSNHpc2haevZJgRBEGob5RbWNx+++0ux7MgXJcuXcq7eEEQBKEc7N30B9a88gKaBo3ARY1ugc5NhzxLJnwvb4LO/fuVyUphzrdg24oT2LL4mArSJs26hKHP6OYIjPCW8yUItQRvb7lfhRomLFioTBAqgyeffBJnz55VrnZC4cJ4LVu2xPfff4/u3bvLoRFckpOVia8fGAev7Ibo1/Au+BmDC8ZH5qDZrcOg9ym9RYEpKQ9uOov1Px5GZkqeGhfe2A8XXd1CArMFoRbWQnr44YerezOEOoaUORVcMmjQINx///3njWcxncDA/zK7PPPMM+jcuXDRLMbVcBrOz4YIp2GvqPOwfPnyIo9+fHw83nrrrUJF8FjAR5uXhXwiIiIwfPhwzJkzB1ZrQTrL+vIyeOihh/Doo49W96YINZRf3noCy+58Ct29b8DAhjcoUZHnlo2Qm9qixf0jyiQqzh5Nx4+vbMGyT/YoUeEb7IHhk9vi6ke7i6gQBEEQKj7dLBuDkZGRRX4X6j6//vorrrnmGjz22GN46qmn7OOZKcxZSAQHF/SguuLjjz9G37590bhx40LjR44ciU8//RQWi0VZMxYvXoz77rtP9d7//PPPqiJ3feD666/H1KlTVdIEHltBIAdWr8bpb9agjd8AeEYVuDiwLoVXt3A0uKIPdJ6lvz+yUvOwfuFh7P87Xn03uOvQbWRjdB4WA4O7Xg68IAiCUDkWi1tuuaXY70Ld5ptvvsHYsWPx8ssvFxIVhA1+ikzHgT3vRTF37lxcfvnl54338PBQ80ZHR6Nr1654/PHH8dNPP+H3339X1hQyefJkXHbZZYXmM5lMCA8PxyeffGK3yNx777145JFHlMDhMmlZceT1119XSQl8fHzQqFEj3HnnncjMLEij6Wi9WbRokcqORl/Vq6++GtnZ2fj888/RpEkTBAUFqfVQCGl8+eWXyoXJz89Prfe6665DQkKC/fcZM2aodM1JSQUZdsioUaMwePBgu2WGy73ooovUcRLqN6xmnbHjDLZM/QxevwKxgf3hqfdGjjkN2a3yEP1EX4Rd27bUooJxFJt+O4qvnv7bLipa9YrE9dP7oPulTUVUCEIth+9Fvsc48LMgVASGiu6tLu67UOCjbM4r8E2uagweHpWWS/69997Dgw8+qNyS2JteHpKTk7Fnz54Sxw8MGTIEnTp1wo8//ogpU6aoYcCAAThz5gyioqLUNGz8s8E/btw4+3xs/HObN2zYgPXr1ytXKzbW6V5FdDod3n77bTRt2hRHjhxRwoJChCmVNbhMTsMGfkZGhhJWY8aMUYLjt99+U/NdddVVarnauvkAf/bZZ5UYoaDgNnDdnJ7Q/YuWGO7HggUL1LFdt24dtm/frrZJo2fPnpLOuR5jyTIhe1M8EpbugYfFG+HG5mp8QtZhpEYmYPBjD8NN71amZ9ThLeew7odDyEjOtRe463dtC0Q2Dajw/RAEoXrgvX7s2DH7Z0GoCOqH30gNgqLi7ZuurpZ13/v59zB6elb4cplW+O6771bWgKJExc6dO+Hr+18l37Zt22Ljxo0upz1+/Lh6yJWmyCJrpuzYsUN9pgsVG+20DFAIELpP0UXLcRs6duyIp59+Wn1u0aIF3n33XaxYscIuLBxjTGh9eO6551QWNEdhQZHwwQcfoHnzgkYdLRZcL920uC7uJy0Nq1atsgsLWlQ0mjVrpoRJjx49lDWE8zCF81dffaViV+hSxt/pGhYTE1Non3l8tJeCUH/gvZG+7BjS/zgBNyvgAW/kW3JxNGMH4q2bceWr78HHP6hMyz53PANr5h/E6YOp6rtPoAf6jGmOlj0jpMCdIAiCcEFEWAjlpmHDhqqH/pVXXsEll1xitxI4woY+YyAcXZqKIicnx56yuDSNLUdrDHv7mU2KwoKNfLpKrVy5stA8FBaOcLsdXZIYEzJz5kzs27cP6enpMJvNqj4LrRRaij7+1UQFYUA5RYijgOE4x+Vu3rxZuV3RApGSkmJ3b6KgohDRBMerr76qsq5RkNBdyhkvLy+1LUL9InVxHLL+PAVe7cl5Z3AofSsS0zej5913Y8jAqWVaZk5GPv7+6Qj2rD0N2AC9UYcuI2LQdURjGD0kjkIQBEGoAcKCLia9evWqzFXUOuiORMtBda27pPj7+yMtLe288ayoHhBQ2B2CsQJshLOnX+uddxYXjKeIjY0t0bpDQ0PVXza6w8IKCnmVxGpClyWNG2+8UfX208WJbkT8rX///oXmYWYpRyhMtEb+0aNHVZzGHXfcgeeff17FYaxZs0bFDTHdqyYsXC2juOVmZWXh4osvVsPXX3+t9o+Cgt+5XEdWr16trBfcFooa58B0uoyV9PgIdYMtr3+D8IRG6vPmxKU4mbwB4T1a4bZpS8u0PKvFil2rT6sid3nZZjWuRY8IZaXwC65466YgCIJQt6lUYUHXEzaaBKeGZyW4I1U0tDAsXXp+Y2XLli2qhoIzDCamuBgxYoQKjKa4KI0rkyO0AFDYMM7C1bqcoSWCrlYPPPCAfVxISAhGjx6tXKAoLiZNmlSqbaBVgWLgtddes8c1fPfddygvtH4wKPvFF19UAeFk06ZN5003b948FTPyxx9/4Nprr1UxGdOnTy80za5du6QIZT0h4cRhbHluNtoHFSQl2JG0EgmWDZj4ybfwDwop0zJPH0zB6rkHkXSqICFBSENfDBjfEg1i/0snLQiCIAhVKizY6CnKNYU9qkLthD31jDlgRiO6FdF1icH43377LX755ReX89AdatmyZar3neKCjeKyiAs25IcNG6YsBBQHjuTl5ak0xo7pZumuROsCrRSOcLs5ntPedNNNpdoGWlcYP/HOO++o7FRr167FrFmzUF4YJ0HrDZfLeA2KA4oGR06ePKmO/0svvYR+/fopccT9oJtZ7969C9ULcZ5XqHt88+Qt8D4eiO5RBffCwaQ/EXJVC1w6siA+qLSwBsW6Hw/h4D9n1XcPbwN6XdEM7QZEQ6ernOQOgiAIQv2g3MKCvdQMVnX0KdeEBV05hNoJffx5/pihiI18uukwQHr+/PmqlkRR0E2Klg5OM3DgQCUuygJFwa233qpS1zpmQqKQoJsV3YJoJWE2KAY3Uzg4Tke43ZyWdR5KK3C4XKabZeN+2rRpKssUBYyzeCktdF1iaj+myeV2M2UuYymuuOIK+33DDFHM+MSAeEKhRqFxww03YNu2bepeoxWGrmoMFhfqJltXLcSGd99CoH9ndI0suD5OYicGzX6iTIHUFpMV21eewD+/HYU5zwIGabTr1wC9rmwGL9+iUz8LglB3cXbdFYTy4mYrZ44xptdk9hw2vJyhzz17sOsyDOplY5qNPLrvOMJA37i4OOXfX5pAZKGggc34HLo3TZgwoUyHhFmWWO+CPf68TusSDOim+KFAKQ65BmsfOVmZ+Pr+a5CZAoT6NEf/yKugdzMALT0QfXMPuJXBqnBsdxL+mncAaQkFiREim/ljwPhWCIvxq4Q9EARBEOpLW7fCLRb0Ay+Kui4qhMqDPbLM6sTYidLC2IjExEQVH0H3LM0aUFeg9YiF+xxjSoS6waIPnsHxZWuQY/REqHc0+kWMUaLCq30Igie0KbWoSE/MwV/fHcTRHYnqu5e/O/qObY5WPSPLJFAEQRAEoVKExYkTJ+zBp4JQGbCOA4fSwoQBtBIxDS7djpyzKdV2GKPxf//3f9W9GUIFcuboPvz82B3ItHoDRk+E6EMxIPJqGHTu8GgZhODxrUtV7I5uT1uXHcem34+qz4yd6DikIXqMagp3r7p1PwiCIAg1hzK/YehvP3XqVJXSU0u9KQg1AdaRkCqiQm0Kzk7afRL5Rh8V9xCh88KA2EnQ5evg3sQfITe0gZuhcPxQcZzYm4zVcw8g9WxBjZPoVoHK7Sk4yqcS90IQhNoG05hr2Q6ZiKeudcIJ1UPJ31Yu3JyWLFmiKhazV1gQBEEoOXv+XoZZVw/GmQNnkW80wt1kQtMWzTG0w0NKVBijfRF6czvo3EtWoC4rNQ9LPt6Fn9/apkSFt787hk9uiyvv7yKiQhAEl27DBw8eVINWa0kQykuZ5Wnfvn1VAbwvvvhCZQ5i+sw333zzvCJkgiAIwn+YTSZ8+eC1SD+TC7Peh5kK4KvLxqhn3oZ+cS4sybkwhHsjdHJ76DwNJSpyt2PVSWxcFAdTrgVMGNVhUEP0vKIZPMTtSRAEQagNFgsNpt/cv38/Ro0apfLsM/0lMyEJgiAIhVmzYA4+GjcSyQkmmPV6eJry0KJ/Z9z8wkLofspSokIf7ImwKe2h97lwGsgzh1Lx3QubsPb7Q0pURDT1xzXTeqD/uJYiKgRBEIQqp8Ic6lhxmemoaLlgIbV77rkHTz311Hn1LQRBEOobWekp+Oa+8cjM1MNq9ILOaoWPjxnjPvgW+v35OPfxTsBqgzHKByE3tYXe36PY5eVk5GPdgsPYt+6M+u7hY0DfMbFo0zdKsj0JgiAItU9YsArxP//8o4a9e/eq4mTt27dX1YSZX3/u3Llo27atSkfbvXv3it1qQRCEWpVCdi1yjB7KRuxtykG768ah/+gpSFl4CBmbCipge3UKQ9BVLYqNqWBSgr3rzqjK2XlZZjWu7UVR6D2muRS5EwRBEGqvsHj++edVATO6QvXu3RvdunWDl5eX/ffbbrsNL7zwgqoivGvXroraXkEQhFpBYvxx/PDgzcgye8Fm9IDBYoFfhDtufHMx3HJsODd7B/KPZ6hMUAEjm8J3QHSxFbVT4rPwx9f7cfpgqvoeEu2LQde3QmSzgCrcK0EQBEGohBgL1rH4/vvv8dBDD6Ffv36FRIXGlClTlDWjMtiyZYuq7M0CaCEhIUrIsNKyK5KSklRNA760U1MLXsoaf/zxB7p27QoPDw/ExsZKhqtKZsWKFWjTpg0sFguqgmeeeeaCtTAGDRqkqsdXFSzeFx4ejpMnT1bZOoWqZcHrD2PenVOQafGGzc0NPuYsDH7gXkx+7ydYzuTg7Dtblahw8zQgdFJ7+A0seD65wmyyYMMvRzD32Y1KVBjcdeg7NhbXPN5dRIUgCIJQe4UFC4+Vhry8PKxcuRIVzenTpzFs2DAlBJiZavHixdi9e7eyjrjilltuQceOHc8bzyBzBp0PHjwY27ZtU41LiiGm0a3v8FiyocOBBdl4rGfMmKHyXmuCTPvdcbhQ4bZHHnlETaPX6wtVkn755ZeVCx1rooSGhuKiiy7Cp59+CpPJVOJt5voXLlyImnYcR48eXWgc94+WvqeffrratkuovEJ3H44fjCN/70Gu0R1GsxnhTQJw+w+r0PGiS5C1KR7nPtwBa0a+yvwUcXdneLYMKnJ5J/enYN5z/2DTr0dhtdjQuH0IJjzVC11GxECvL3fuDUEQ6jF8t/M9xIGfBaHKXaF69OihGklsfPOzK9LS0lTBlbfeektZEe69915UNIsWLYLRaMR7772nYju0mA+Kh0OHDqlGsMYHH3ygrBQMJP/9998LLYfzsELza6+9pr6zJ33NmjV44403cPHFF6O+M3LkSNW4p0D87bffcNddd6njPm3aNPs0zAjm7+9v/15csD6P7eHDh3HVVVcVEhU81tu3b8ezzz6rBAWX9/fff+PVV19Fly5dylR9u7qhRaY4t5ZJkyYp98FXXnkFwcHBVbptQuXw3fN3ImHzYeT9W+jO15qFoU/PRGzH3rBZrEj7NQ6Z606raT3bhiB4XEvoPFw/gnMy87Hu+0PY93e8+s6aFMz01LxrWLHXlSAIgiBUJ6Xq8tqzZw98fHyUC1JkZKTq7b/11ltVBqgbbrhBuRTRxWPOnDmqB7oyRAVhQ5fqWhMVRHPFYuPVcXvZy85aG47Taqxfv15ZPhxhI5fjBSj3MJ7nxo0b44477lDH6ueffy50aHi+OY02FCcsGNDPa8fT09M+jrVPVq9erVykKFwoIpo1a4brrrtOWaNYgFGrps1pHeG0dHXSfidjxoxRDS/tu8aXX36pxgUEBGD8+PHIyMgo9DstMXfffbf6nRaFJ598slD1bl5zdPuLjo5W9wDji2i10WCRSLrl8fgwaQGP3eTJk/H555/jp59+slt0tHnatWuHBg0aYMGCBXKp1XKO79+GD8cNxokdx5HHQndmE6JaRuB/81cpUWHJMiHxk112UeE/LEZV03YlKrTg7G+e3lAgKtyA9gOjcd303ojtFi6iQhAEQag7FgvGMrz++usqcJspZdmIP3bsGHJyclRj7Prrr1cNc2aHqkyGDBmCBx98UPX23nfffcjKysJjjz2mfjtz5oy9IThhwgQ1TUxMDI4cOXLecuLj4xEREVFoHL8zbS73yVXcCJfLQYPTlgY2HGym6qlw6WbUlathwuPBeJWy8tdffynB4MjXX3+tBAstE87QOsKhJDA7GUUOLSy0tDi6WtFKQhcpWrpSUlJw7bXX4sUXX1TXsQYFAF3mNm7ciE2bNilrG68bCmdC0UGhSnGkCQKuZ+fOnXbxk52djZdeegkff/yxuleioqLUdcRrhNtFHK0TPXv2VMeE6xVqJ988eQuS9pxEvsFHffdFFi554Q3EtOoMa54FWRvOIGP1SVgzTXBz1yN4XCt4tQtxuSxWy/7j6304dUALzvbBoOtbSxyFIAiVAjvUtM4tdsoZDBVWgUCoxxjK2sBkITwOFQnFARtmxcFgcPb2siFIcUG3HDYiaR2hKNAsExxP1yZaUiqSmTNnYvr06WWen6Li9FPrUB00mNFXNW5KC8UQLQqMPaF1yhEGxTtCoclGtSv4Gxvljhw8eFAFT5eXsLAw9ZdWA1pOHLFarcqi4Ofnp75PnDhR7Y+jsGjUqJFygaPwatWqlRIM/E5hwdgiCgP+1baf1gvG9nA8s58RxoO8//77KlbE8V6hEHXeJsJlbd26tdz7LlQ9h3b8jRXPTkOmzgcwGOFhMiG8W3Nc+8T7sOaakb7qODL/OgVrdkFMkiHMS1kpjBEFAsQRi8WKbcuO459FR2ExW2Ew6tDjsqboNKyRxFEIglBp8N3IDjNy5ZVXypEWKoRyy9NTp06pv3QRKS9Tp04tMgBbg24yhD3fHM6ePatcU9ggpDVF+51B42wcMnMV0dxaaFl54oknlDhgY4/zO8Lv9PF3Za3QBAsFjQZ7o9korYuwh5+uTWww8wHE4625Hmmwx11rsJOgoKIDUdl77+gGRRzdjSoLukA5biMtCQkJCYWmYcpkR2tOnz59VOwNYyV4HfFvy5YtC81DweAoouie5ypJQFHwGqOVQ6hdfPPkZCTtOV1gpbDZ4KvLxhWvf4CI8OZIX34MGWtOw5ZbICj0IZ7wH9wI3l3C4eYi2DrhWDpWfbUPiScKMto1ahOkrBT+oa6fP4IgCIJQJ4XF2rVrlTVAyxTFBjtFARvtjsG8pe111nqeS4rmysS4DjZa6cNPfvjhB9WQdXSVoc87G8LNmze3Nx4ZlOzIsmXL1PiioO88h/K4I9FyUB1w3aWB2bIY/M4GM3vXXZlJGfxOK0FJ4DVCVyRH2Fjft2/fBeelJcpZhJQ0Y5SzOxUFBIVSSWEaY1rFNm/eXMjFijjGlFAolMbVLDk5udTXu1C9sRS//98D/1opDPAw5SP6oo644vaZyFxzCmc+/Qe2PIvdQuE3JAbeHcPgpj//mjDlW/DPL3HYtuIEbFYbPLwN6HdNC7TqHSlxFIIgCEL9Exb/+9//lKsRG/AUEuvWrVOZoPj9zz//rBALRnG8++676Nu3r2rYUQw8/PDDym9ea+Rq4sGxdgDhNmvTsEo4l8MUqBQdtHIwoxXjRyoLFcRbBnek6oCWIMcMW+WFcRSa2VWDVpDHH39cuQQ5x1lQODBrFLeDDXAtfkazFDFdsLOAKGt9DAaKO8KsVIydoJDgdnG5tHL079+/VMulKCtqm1g4siLcwITKZ96ztyNhW1whK8XomR/C47AR8S9thC2/QKgaI72VoPBqHwo3nVuRKWRppUg/V9DxEds9HP2vbakyPwmCIAhCbabMidAZEMssPcwExcYn8/Iz6JWNsKooNsYgW1onOnTogNmzZ+PDDz8sdRYq9rZTRFCY0C+eri8MvJVUs5UDj6tj1i7Ca4UpZocOHarSBzPtLAPtKfDonsQYDC1gn5mdaHGia9JNN910nvWALk+MnWBQvrNl5ELQ8kYXN6bP/fbbb/HOO++oxACaVYWJCXiN//jjj0rQ8PpjvM2FRCi3aceOHWq5FLealYUuULSAjBgxolTbKVRDXYpxg3FyFwO0C2IpGndpjuvu+grmr84hc/VJJSqMDXxUDEX4vV0LrBQuREVetgmrvtyLn97YqkSFT6AHLr2jAy6e0l5EhSAIglC/LRbs+WcPrqPfOXvjmd6V2W4qG6aQLQ3sGXblz8/xEkBbNbBxTusQG9kMkCZ0K6OwY6A0xSGDolkkj9cXhaKWYYyxLWzQX3bZZSolLGteOFssKAwpDj766CNlMTt69GiJt42iga5zvHYpWCgqmBlKg0Hazz33nIoDYlwR3boofLg9xcHgb6aY7d69u3KpWrVqlbrmmIKWWadKawERqo4fXnkAZ9bvKahL8W/Gp1EvvguvLXqk/VqQZc69kR/8hsbAs1VQsS5MR7aew59z9yM7LV99bzcgGn3GNIeHl2RhEQRBEOoObrYyRs++/fbbqrHFvP2Owct0IWEBNC2ou65Dlxw2dFkY0Dm2JDc3VzV+aRlxDlqur9BljceMIqI+Q1FC4eScfreikWuw9Jw7FYcfp96CTKsXe0tUXYqwTk1w9T1vIemLPTCdyVK23sBRzeDTt0GxgiI7PR+r5+7H4S3n1PfACG8MvqEVGrQoOsmBIAhCVUBXY1retc47qb4tlKWt60yZu8s0dyf6oY8dO1YVK6Mv+VdffaWK4wmCKxjcz5SsDJ52VbSwPkCXKN4zrLMi1Cx+fnsaTvyxGblGb1WczseahYufexVRxuZIeGerSh+r8zEg+Lo28GxedNIC9tcc2pSA1XMPIDfLpFyjuoyIQY9RTWAw1o4YK0EQ6jaMS6Sg0D4LQrVaLJiWddu2bconnn850B+evXd0Y2HsA1NvcmAhsbqKWCyEmoxYLEpGekoS5t5zLTLzPWFzc4PRbEZI6whc99xnyFx7Gmm/HQGsgDHaFyET28AQWLQFMistD6u/PYAj2wqsFCENfTH0xjYIi/kv5bEgCIIg1BaqxGLBNK8MxnUMdGYjhoG1muCgmxSLh6WmFlSSFQRBqGmsnPsu9n23EDlGuj4BPpYsDPu/Z9G8fR+kfHcA2VsLap6wFkXQ2Fi4FWFxYB/NwX/OYvW8A8jLMkOnc0O3S5ug28jG0Bvqp3VOEARBqF9UaOQg4wh69OihBkEQhJqM2WTCZ3ddiYwUwGr0hMFiQWCMH256fRHMqblImLUDplOZKp4i4NJm8L2o6HgKWin+/GY/4rYXpLUObeSLoTe1QWhDsVIIglAzMZvNqhAuYSISV7WqBKG0yFUkCEK9Y+uqhfj7nbeRzVgKHeBtykHPO+9Ct2FjkXckFUlf74M1ywSdtwHB1xcdT0ErxYGNZ/EXrRSMv9C7qTiKLhc3ht5FpW1BEISaAmMd6V1CLr300ureHKGOIMJCEIR6xRcPjUfK0TSYjd7QWa3wDbDhpncXQW/WI+PPE0hbcrQgniLKByET28IQ7DqeIis1D398sx9HdxRYKRhDQStFSPR/1dgFQRAEoT5RZmHBegGuoKsAXaJYNO/KK69EcHBwebZPEAShQojbuxlLnnoIWTofQK+Ht8mEThdfg3aNhyJl9l6YTmfZp/XqHIagsS2gc9e7tlJsiMdf3x38z0pxWVOV9UmsFIIgCEJ9pszCgkXltmzZolLMasXODhw4oIqLtW7dWqUUZTExVlpu27ZtRW6zIAhCqZg/8x7E/3MA7l4NEevVBA19GiPCvzUQZ0NG3En7dLRS+PSKhE+vKJfxFDkZ+fjj6/32jE/hjf0w5EaxUgiCIAhCuYSFZo1gkTwt9RTTUE2ZMgX9+vVTFYdZ/OuBBx7AkiVL5GgLglDlpJ6Nx6rHn0GYR1t0aDoCvkaHWAmTDTpfIzxbBMGjZRA8YwOh93MvclkUE398vQ85GaZ/YymaouvFMdBJLIUgCIIgKMocXfjKK6/g2WefLZTPljlun3nmGVUgz9vbG0899RQ2b95c1lUI1Qh7a4sbeJ7rC++99x6aNGmiXPx69eqFjRs3lnjeF198UR0vraCkUPlYc83I3paAfS/+gtTXdqNH2PWI9e+iRIVNB3g0D4D/yCYIv7cLoh7vheBxreDTJbxIUZGXY8aKz/bg91k7lagIbuCDqx/rju6XNhFRIQiCIAgVYbGgdSIhIeE8N6dz586pQhokMDBQlYwXah9nzpyxf543b54Sifv377eP8/U9P0CV59rdvege39oI953xRLNmzVKi4s0331S1W3gswsPDi533n3/+wYcffqiKRAqViyU9Hzl7kpCzOxF5R9IAiw2+CFRdJ9mmNCQZjqPzpPHwaBbgMm6iKE7sS8bKz/ciMyUP9IxiHEXPy5pBb5SMT4IgCILgjK48rlCTJ0/GggULcPLkSTXw8y233ILRo0eradiz27Jly7KuQqhGIiMj7QMtUex1dxxHYTFo0CDcfffdqjc+NDTUXiyRvftsgDvSuXNnu5WDKe5mzpyJpk2bwsvLC506dcL3339/wW3i9cR1ch7G8WzatAmzZ8/GFVdcUUlHAXj99deVW9+kSZOUiKbAoDVuzpw5xc6XmZmJ66+/Hh999BGCgoIqbfvqM+akHGT8eRIJ72/DmZkbkLrwEPIOpipRkZafiN2p67D62GxkjMxHn5fuhFfr4BKLClO+BavnHsDPb25TosI/zAtjpnZFnzGxIioEQagTGI1GPPTQQ2rgZ0GoVosFe2IZPzF+/HhVZIWZUnhh3nTTTXjjjTfUNGz8ffzxxxWyoULN5PPPP8cdd9yBtWvXlngeioqvvvpKNdJbtGiB1atX44YbbkBYWBgGDhzocp6///4bgwcPxowZM1Rj/ZFHHlGfd+/eXawoYeV3DsWxZ88exMTEuLTA0JVv2rRp9nE6nQ7Dhg3D+vXri13mXXfdhVGjRqlpn3vuuWKnFUpP5rrTSP3lMGD7b1yGPgnHzm7DsZzDyDIlw9c7HxM/XgAvn9Klf40/kobln+1BWkKO+t5+QDT6jG0Od0/Jzi0IQt2BHYY+Pj7VvRlCHaPMb0r2WLOBRxFx5MgRNa5Zs2aFXGTYSy24pjgXMTZeHStgFjctHwyOPQ1FTVtZLkoUBoypKSl5eXmqob98+XL06dPHft0wexjFalHCgu5I11xzDR5++GH1fcKECWqg5axLly5Fru/222/HtddeW+w2NWjQwOX4xMRElfUsIiKi0Hh+37dvX5HLmzt3rsqYRlcooeJJX3UC6aw1ARS4NrXwxqJPH0Jirok3BDxMJsQO6oor7p1ZquVazFb8sygOW5Ycg80G+AR6YMiNrRHTNkROoyAIgiCUgHJ1waWmpuKTTz7B3r171fd27dop9yi6zggX7rUvrrHOjFoar776Kkwmk8tpGzdujJtvvtn+/a233kJ2dvZ50z399NOVckq6detWqukPHTqktm/48OGFxlMQFSUQ6GZHCwGPgwaFF61k06dPL3Z9zFxWlbVUTpw4gfvuuw/Lli1Twd5CxcHznb70GDJWnVDf/YbGYHfWKmx75yvkGD2VqPCxZmHUi2+jUcsOpVp2SnwWls3Zg3PHM9T3lr0i0P/alvD0EfcAQRDqJvQ20bJ20pXZsUNTEMpKma8i+rfzQqS/e8+ePe3+6M8//zyWLl2Krl27lnmjhNqDKzMqLS5sBDqiCSPGHpBff/0V0dHRhabx8PBwuQ5NuDpeUwye5nXXoUPxDcjyuEIxboR1Wc6ePVtoPL8zzsQVdJ1iUgPHbaXVg+5e7777rrLYcJlC6bBZbUhbdES5QJGAS5rihwX3I/VkFixGT+gtFgQ29MHNby4q3XJtNuxefQprvz8Es8kKDx8DBl3XGrHdig/MFwRBqO0w3pFtOeLc2ScIVS4sGF/BoFm6Q2kql+qXdSwYzMuGlFA0jn77rhrmjjCwqiici3ixt7y6YayEY1YpZgmLi4tTnxkATQFx/PjxIt2eXGUgY2Nc29fk5GRlvWDQ94UojysU3cdokVmxYoU9IQEfxPzOoHVXDB06FDt37iw0joHfjDd69NFHRVSUUVSk/HgQ2ZsKBJ5bP198/e5EewVtL1MuOk+6EX0vv7FUy81Oz8eqL/fi6M4k9b1h6yAMvaktfINcC1xBEARBECrRYuEoKtTCDAYVVNu9e/eyLrbeUJqYh8qatrIYMmQIPvvsM1x++eUq5TBT1Wq99H5+fkooUZiykc5iihQODP5mTRQG/zvDWB32+jOWg3EWFE/MPEVLw7Fjx5Q7WGW5QjG2g9vEa5oWEma7ysrKUmJBg5YIZkSj4OD+tW/f/jyrTkhIyHnjhQtjs1iRPG8/cnYkAm7AqcjD2DTnG+QafWhugJ97Lq7/6Hv4+Jcu89bRHYlY+eVeVZdCb9Chz5jm6Di4Idx051fbFgRBEAShkoUFG4HsdWZPrLOPORtXQv22xtBCcdlll6l4GxZS1CwWhN9p1WCcCQP/KT7oOvT444+7XF5sbKzKAMX4Ebo1MRPZN998gxEjRmDkyJF2V6nKYNy4cao2C8VRfHy8EjmLFy8uFNDNIO/Dhw9X2jbUV2wmK5K+2YvcvcmA3g3b0xfiwJo9sBrdYTSbEdktFtc+/m6plsk0suu+P4Rdq0+p7yx2N+KWdgiJLl3mKEEQBEEQzsfN5uwMX0Luvfde1UtLl5S+ffuqcex1Ztaeq6666rw6BnUVuvmw8cxed8cq5CQ3N1c1qFmvQQJ5heqgtl6D1nwLkr7Yg7xDqbDpgX9OfIm4vIL4Cm9TNvo/9Bja9x1RqmUyMHvZnN1IiS9IbtBpaCP0Ht0MBqPEvAiCUP9g0hQtkQw7BGuCx4NQ+9q6FWaxoKCgz/uNN96oYisI056ypsGLL75Y1sUKglDPseaakfjpbuQfS4fFzYz1R7/BKfMZ9oLA19uEmz//Fe6eXiVfntWGrUuPYePPceqzT4C7iqVo1LbqsoUJgiAIQn2gzMKCypauKVS7mhtI8+bNVVViQRCEsmDJMiFxzi6YTmXCZM3Bn6fnI8l8Bu5mExr27YAxU/9LOVwSMlNysfzTPTh1ILXgGdUlDIOubw1PX0kjKwiCIAgVTbmTFlNIXCjlpyAIwoWw5lmQ+PFOmM5kIc+ciT/OzkeqKQE+pmwMe3IGYrv0K9VBjNt+Diu+2Iu8LDMMHnoMGNcCrftEnZdJTRAEoT5CLxMtk6RjoV1BqDJhwQw5JYU1LQRBEEoCQ71S5u9XoiLXnImV8XORmZeIAH8Lbp61BIZSvPTMJgvW/XAYO/84qb6HxfipAO3ACLGmCoIgaLCThclTBKHahMXWrVtLNJ30CAqCUBrSlh9Fzq4kWGwWrElYiPzseLQY2hOX3/VsqZaTfCYLSz/ejaRTBYUYOw1rhD6jm6uUsoIgCIIg1CBhsWrVqsrbkjpMGRNvCUK9uPZOrtwGrMhQn7ckLkVu1gGMev5VNG3TrVT7uXfdGfw17wDM+VZ4+RlVgHbj9iGVuOWCIAi1F9aHYv0lrbirVm9KEKo1xkIoGs1nMTs7G15eJc9iIwgVmU6Q1NQXxuqP3kOjA61g1HvgYNpmJGELpsxdWirXp7xsE/74ej8ObU6wV9AeNqktfAKkgrYgCEJxwmL9+vXq86BBg2rse0KoXYiwqER4k9J/MSEhwR7oLm5iQlXByuYs7sfrzmCoebf6l/ddh+76cTC6e+Bc9jGkNT+FKQ8tLtUy4o+kYeknu5GRlAudzg29rmyGLsNjpIK2IAiCIFQDNa+1UceIjIxUfzVxIQhViU6nQ0xMTI0StCkJp/DdPRPRNeIm+LkHI9uUCvcx4Rg9+JUSL8NmtWHzkmPY+Euc+uwf6onht7RDZNOASt12QRAEQRCKRoRFJcMGXVRUFMLDw2EymSp7dYJwXr0ZiouawrpfvsC2T79Ai/BLEOXdDBarCVF39IJfbHiJl5Gdno/ln+7Gib0p6nuL7uEYeH1reHjJ40wQBEEQqhN5E1ehW5T4Lwr1ma//bxIS98WjQWAXtAnsrcaFXd8B3rFhJV7G6YMpKutTVlo+DEYd+o9viTZ9pTaFIAiCINQEpI6FIAiVSk5WJr687UpkmL0Q5NUAPUMvUeP9BjaEd6eSiQq6O21ZegwbfjoCJroKivTGxbe1R0gDXzl7giAIglAX6lhs2bIFZrMZrVq1Ut8PHDigeuW7dSt5mkhBEOou+zevxsrnn0G20RseOm/0jxgNvc4Iz1ZB8L+4SYmWkZNJ16c9OL47WX1v1SsSAya0hLunGFwFQRAEoU7UsWBlbT8/P3z++ecICgpS41JSUjBp0iT079+/4rdUEIRaxW8fPosjS9Ygz+gNvRUYGDMGXvpAGEK9EDy+dYkyN505lIoldH1KzYPeqMMAcX0SBEGosJT4d9xxh/2zIFQEZY7qfO211zBz5ky7qCD8/Nxzz6nfKhtaS4YPH67SuYaEhOC2225DZmZBtV3HwGnnYe7cuYWm+eOPP9C1a1d4eHggNjYWn332WaVvuyDUdT679yrsX74eeUYjPEwmDO5yI4L0DeHmoUfIjW2hu0CgtXJ9WnIMC17fqkRFYIQ3rnmsO9pe1KBGZbgSBEGorfBZysQyHOS5KlS7sEhPT1c58p3huIyMgiq6lcXp06cxbNgwJQQ2bNiAxYsXY/fu3bj55pvPm/bTTz/FmTNn7MPo0aPtv8XFxWHUqFEYPHgwtm3bhvvvvx9TpkzBkiVLKnX7BaGukp6ShA/HD0bS2TxYdTp4m7Ixatz/ISQ9CnADgse3gjHcu9hl5Gaa8OsHO7B+wWElMFr0iMA107ojJFriKQRBEAShJlNmJ+UxY8YotydaJ3r27KnGsZH/8MMPY+zYsahMFi1apMx27733nj2V5qxZs9CxY0ccOnRICQ4NWjS0WhLOcJ6mTZvaLSxt2rTBmjVr8MYbb+Diiy+u1H0QhLrGrnVL8derLyHb6KO++xqyce29nyLrl+Pqu/+IxvBqE3LBgndLPtqFzJQ86A069B/XAm37iZVCEAShMipv//XXX+ozXdglc6VQrRYLNsovueQSXHfddWjcuLEa+HnkyJF4//33UZnk5eWdl5/fy8tL/aUwcOSuu+5CaGioEj9z5syBjSll/oWl7Gn5cISCQitxLwhCyfj5nSew6rXXkW30gt5iRVhjP0x8ZB6yFhWICt+LGsBvUKMi5+d9uWPVCSx4dYsSFQHhXrj6sW5o1z9aTPSCIAiVJCz+/PNPNfCzIFSrxcLb21sJiFdeeQWHDx9W45o3bw4fn4LeyspkyJAhePDBB9W677vvPmRlZeGxxx5Tv9HdSWPGjBlqWm7r0qVLceedd6o4jHvvvVf9Hh8fj4iIiELL5ne6eeXk5NjFirOo4aDBaQWhPjPnriuRlmCC1WCAhykfrcdcjIu63Yykr/YCNsCnZyQCLmtWpEAw5Vmw6qt9OPjPWfU9tls4Bk9sLVmfBEEQBKGWUe58jRQSdEGqCCgOXnrppWKn2bt3L9q1a6eyUVFcTJs2TZnvKBYoChytGE8++aT9c5cuXZQAoRjRhEVZYMD69OnTyzy/INQV0pISMPfOcciED6DTwcecjYuffwVRhmZI/Hw3YLXBu3MYAkfHFikqUs9m4/cPdyL5dBZ0Ojf0vSoWHYc0FCuFIAiCINQ3YZGamopPPvlENfZJ27ZtccsttyAgIKBMy5s6darLAGxHmjVrpv7S7YrD2bNnlbhhw4UpcLXfXdGrVy88++yzyuLALFCMveD8jvC7v7+/S2sFoZChoHG0WDRqVLSLhyDURXas/R1rX3vNHk/h556Lmz//Fbb4fCR+sgsw2+DZLgRB17QqMq3ska3nsPzzPTDlWuDt764K3jWIDaziPREEQRAEodqFxaZNm1Q8AhvgWvA2g55feOEF5XbEFK6lJSwsTA2lQXNlYvyEp6enSkFbFMz8xJS4FBWkT58++O233wpNs2zZMjW+KDivNr8g1Ed+nTUdR5atQ77RE3qLBaEtw3HDC58j/2QGEj/dDZvJCo+WQQiZ0Bpu+vNFhdVixYafj2DLkoL4i6jYAFx8a3v4BMh9JQiCIAj1Ulg88MADuOKKK/DRRx/BYChYDKtwM10r07auXr0alcm7776Lvn37wtfXV4kBZqN68cUXVRYo8ssvvyjrQ+/evZXg4DQUPQ899JB9GbfffrtaziOPPILJkydj5cqV+O677/Drr79W6rYLQm3li6njkXQ8HVYD61Pko+21V2DI+Lthis9C4pxdsOVZ4N7UHyE3tIGb4fzcENnp+Vj6yS6c2p+qvnca1gh9xjSHXl/mPBKCIAiCINQFi4WjqFALMxhUI7179+6obDZu3Iinn35aBWO3bt0aH374ISZOnGj/XUtHSwHEjDNMQUtXqVtvvdU+DVPNUkRwmrfeegsNGzbExx9/LKlmBcGJ/NwcfDblcmSYPFU8BetTDH/mBcR27A1TYg7OfbwT1mwzjI38EHpzO+jc9S5TyS6evUsVvDN46DFkYmu06F44eYIgCIIgCLUXN5tj/tVSuiB9+eWXGDFiRKHxLC534403nhe7UFdhjAVjStLS0lRshiDUNc4c3Yefpt6BLMO/9SncsnDD7B/h4x8Ec0ouzs3aAUtaHoyRPgi7rQN03sZC8/MRs/OPU1j7/UFYLTYERXpj5P86IDiq8jPICYIgCK6xWq32TJpRUVGFkt8IQlnbumW2WIwbN04Far/66qvKJYmsXbtWuSRNmDBBzogg1AHW//oVtnzyOXKNPnCz2hAYZsDk91ep3yzp+cpSQVFhCPNC6JT254kKc74Ff3yzH/v/jlffm3cNx5AbJZWsIAhCdUMhER0dXd2bIdQxyiwsKCiYiYnWCcZWsFeSRevuuOMOFesgCELt5sdXp+Lk+j0wGT1gNJvRsE87jH3oNdisNuTuSULa73GwJOVCH+yJsCkdoPd1LzR/RnIuFn+4EwnHMlRmKMZSdB7WSFLJCoIgCEIdpcyuUBrZ2dmFCuSxGF19QlyhhLrInDuvROo5M2w6N3ia8tD9tknoOWKCEhTpy4+rYG2i93dH2O2dYAj2LDT/6UOpSlTkZJjg6WPEiFvboVHr4GraG0EQBMEZVtv++++/1WcmumFNMEGoNlcoV3UsWLiO2ZXKWsdCEITqJSs9BV/dNhaZNha9c4OPOQtXvPY+AjPDkPD2Fpjis9V0bh56+F7UAH79ogu5P7GfYvdfp/HX3AOwWm0IaeiLS2/vAP9Q13VhBEEQhOoTFsuXL1efe/ToIcJCqHl1LJh16fnnny9zHQtBEKqP4/u34ddpDxYqejf+ka+QvTAeyWf3FisoiMVkxep5B7BnzWn1PbZ7OIZMbAOjh/SCCYIgCEJ9oNbWsRAEoeLYsPgbbJr9KXKN3tBZrWjduAu6RI9F+vwj/wmKftHwu6jBeYKCZKXlYfGHu1RKWbgBfUY3R5cRMRJPIQiCIAj1iFpbx0IQhIrh53eewNE/Nqsg7QbGaHRvMhpeZl+Yz2ZfUFCQ+Lg0LJ61E1lp+XD3MmDElHZo3C5ETo8gCIIg1DPKLCwYvHH8+HFVnM6REydOwM/PryK2TRCESuaLRyYgKS4NVoMBHfx6oE3oYLiZ3UokKMjedadVOlmruaA+xaV3dERgRP1K4CAIgiAIQgFSx0IQ6iFmkwmf3jYK6VlGGA1e6B8yEpG+BZ0E3t0iEDiqabGCwmqxYs33h7Bz1Un1vWmnUAyb1BbunuXKByEIgiAIQi2mwupYEKPRKHUsBKGGk56ShG//dzUy3XwQ6BGGfuFXwscYAhjcEHRFLHx6RhY7f26WCUs+2oWT+1LU9x6XNUWPS5uoWhWCIAiCINRfpI5FOZE6FkJt4vCuf7D0qceQbfRCE5926B52MfRuRuiDPBByfRu4NyzejTH1bDZ+fX+H+mtw12H4pHZo1iWsyrZfEARBqBisVqtyaScxMTGqErcgVGsdC8KCeB06dCjvYgRBqGTW/fQZtn7+DfKNvugePATNAwpSQnu2CkLwuFbFuj6Rk/uSsXj2LuRlm+Eb5IFL7+yIsEYSTyUIglAboZBo0qRJdW+GUMcol7BYsWKFGhISEpTydWTOnDnl3TZBECqIhW88guNrdsLoGYKh4aMR7NlApYX1HxoDvyExF3Rj2rX6FFbPPQCb1YaIpv645PYO8AnwkPMjCIIgCEL5hcX06dMxY8YMlVo2KipK8tULQg3P/BTm1xx9wi6Hh94bOm+DslJ4tgoudl7nIO0WPSIw5MbWMBil6J0gCEJtr7y9efNm9blbt25SeVuoXmExa9YsfPbZZ5g4cWLFbIkgCBXOx7degrQ0HdoEXYQOQf3g5qaDMdpXxVMYgj2LnTcv24QlH+/GiT3J6nuvK5qh2yWNpRNBEAShjgiL33//XX3u3LmzCAuheoVFfn4++vbtWzFbIQhChZKfm4NPJ49CpsUbXUKGomVAQdFKZnwKvLw53IzFB+mlJmTjt/d3ICW+IEh72M1t0bxruJwlQRAEQRCKpMwpAKZMmYJvvvmmrLMLglBJpCUlYM4NlypR0Tl4iF1UBI6ORdDYFhcUFSf3p+D7lzYpUeET6IGxD3UTUSEIgiAIQsVaLB588EH7ZwZrz549G8uXL0fHjh1VDQtHXn/99dIsWhCECiBu72YsfuIRZBt90CloEFoF9FDjKSguVJ+C7Fl7Gn9+vR9Wqw3hjf1U5icJ0hYEQRAEocKFxdatWwt9p08e2bVrV6HxLJwnCELVsnXVQqx75z3kGr3QKXAAWgf2UuMDx1y46B2zPf398xFsWXxMfW/RPRxDbmwDg7sEaQuCIAiCUAnCYtWqVaWZXBCEKmL5l69h78KlyDd6oFNAP7QO6qPGB17ZHL69ooqd12yyYOXne3FwU4L63v3SJuh5eVPpIBAEQRAEoVSUu0CeIAjVy/cv3YdT/xyA2WBEJ/8+aB18kRofeHkz+PZpUOy8uZkm/DZrB84cSoNO54ZBN7RGm77FCxFBEARBEIQKibF49tln4ePjUyjewhUSYyEIlc8XD41H0rF0WPV6dPLridYhA9T4gFHN4HtR9AUzPy16dzvSEnLg7qnHyNs7oFHr4utaCIIgCHUDg8GACRMm2D8LQrXEWJhMJvvnopAYC0GouhoV0OnQya8bWocOVuMDLm0Kv/7Fi4r4I2n49f0dymLhG+yBy+7uhJAGvnLaBEEQ6gk6nQ4tW7as7s0Q6hhuNpvNVt0bUZtJT09HQEAA0tLS4O/vX92bI9QDzCYT5kwaiQyTl/reMaAL2gSPUJ/9RzaB/6BGxc5/aHMCln+6BxazFWExfhh1l2R+EgRBEASh/G1dsX0JQm0rfHfTpciEj/reMbgL2gT8KypGNC5WVLAP4f/buw/wKIv1beD3lvQOCQmEktASEhI6CAhIb4IeCwoKUgQLKIqo8NmwwhEROUeQIggoShMbTUBBeu8hCR1CGiEJ6cnW75rJPzkJKUA2ye5m7991vTK7+77LuEPgfXbmeebk9hjs33BRPg4I90a/caGwc2DlJyIiW9x5+8yZM7IdFhbGnbfJvPtY3A1zLIgqf+O7n154ClkqF9grHNDKtxMaO+VXf3Lv0xDuvRqWea1Bb8CeNRdwdnesfBzWsz4efLKZTNgmIiLbDCx+++032Q4JCWFgQebfx6IszLEgqlzxV6Pw65SXoXTyR1v39mjsHg6VIn9TSrdeDeDep1GZ12pyddj2bQSunU0GFMCDTzRDq97lL5ciIiIiul/cx4LIwl08dQBH/70YbfxHwN+5eWHgblfXBW4P1YdTuE+Z1+ZkaGTlp5vXMqC2U6Lv2FA0blP2+UREREQVZVKOxZ49e7Bo0SJcvnwZ69atg7+/P77//nsEBgbiwQcfNOWtiWyeUW/EubVboD+Yhgcbji38PByDvODarT4cmniUOzuYlpSDP/5zUv7q6GKHwZPC4RfoYfOfKxEREVUNZUUv/Pnnn9G/f384OTnh+PHjyMvLk8+LjPHPPvusMvtIZFMMeTpk7I3F1Q//hscpN9Ryqg+9QYdUj0T4vt4W3mNawrGpZ7lBRdL1DPw8+5gMKtxqO+Lxt9oxqCAiIiLLDCw++eQTLFy4EEuWLIGdXf5ab6Fr164y0CCi+5d78TbiZx5B2sbLsNPYI1efjajk3YjvcANh05+AnW9+NajyxESl4JcvjyMnXYPa9V1lUOHp68zhICIiIstcChUdHY3u3fN3+S1K1Lm9ffu2qf0isjnapGwk/xAJY64OGZpkRKcfQXzqcXR5bRJadRt8T+9x4Ugidiw/B4PeCP8gTwx8MRwOTqwqTURERFWvwnccfn5+uHjxIgICAoo9v3fvXjRu3Lgy+kZkMww5OiSvPCeDilu5N7AzfjUctZno99EsNGnZ4Z7e49RfMdi77oJsN21XB31Gh0BlV+FJSSIiqsHUajWeeOKJwjZRZajwn6Tx48dj8uTJWLZsmVzrHRcXhwMHDmDq1Kl47733KqVzRLbAaDAi+aco6JJykK1Lx76bv8JRm4HH5y1BnQZN7un6A79ewolt1wv3qOj2ZDMouEcFERGVQalUIjQ0lJ8PWUZgMW3aNBgMBvTu3RvZ2dlyWZSDg4MMLF555ZXK7SVRDZa2+QryzqdCZ9BiT+LPUOXdxIil6+HuVfuu1+r1BuxcGYXoQwny8QOPNkbb/o24lwwRERFVuwqvkxCzFO+88w5SUlJw9uxZHDx4EElJSfj4449RHUSCeN++feHp6YnatWtjwoQJyMzMLHHe8uXLER4eDkdHR9SpUwcTJ04s9vrp06fRrVs3+XqDBg3w+eefV0v/iYSso4nI3Ju/G/ahpE3Q5l7ByO823FNQITa+2zz/tAwqxOxEr1Et0G5AAIMKIiK6K/HlcEREhDxEm6gymLwA297eXm4F37FjR7i6uqI6iGVXffr0QdOmTXHo0CFs3bpV/mCMHj262HlffvmlDH7E7Ip4fceOHbJEboH09HT069cPjRo1wrFjxzB79mzMmDEDixcvrpb/D7JtedfSkbzunGyfTd2H21nHMfb7zXBx97rrtbmZWvw29wSun0uB2l6JQS+FoUWXutXQayIiqgl0Oh3Wr18vD9EmMutSqJ9++gnDhw8v9bU333xT3qRXlY0bN8oSt/Pnz5drBAVR+lbMTIiEchFwpKam4t1338Uff/whl2sVEOcUWLVqFTQajcwTEQGSWGt48uRJGZCIGRCiqqK7nYsb8/fBQemKmKxoXEvbhnGrtkFdpHRzWbJu5+G3eSeRGp/Fje+IiIjI+mcsXnrpJWzZsqXE86+//jp++OEHVCWxGZ8IBAqCCkFs1FdQlUrYvn27nNqLjY1FixYtUL9+fQwbNgwxMTGF14hkc5EbIt6rgJjREKV0RWBCVBUMGj3Of/yHDCpS8xIRkbIO477/856CCrHh3YYvjsmgwsXDHv96oy03viMiIiLrDizEt/1ixqLgRl4QSdtr167Fzp07UZV69eqFhIQEOSsiZhxEECCWOwnx8fHy18uXL8vAQuwC/tVXX8mpPpEPIvIyxDWCeA9fX99i713wWLxWVlAjllAVPYjulTZPg2NTl8Ddzg+5+iycuP0jnl91b0FFcmymDCrSb+XC3ccJj73ZDrXq3X3DPCIiIiKLDiwGDx6MBQsWYOjQoTI/4eWXX8aGDRtkUBEcHFyh9xTBgUgKL++IioqSS5ZWrFiBOXPmwNnZWe6pERgYKIOCglkMEVRotVr85z//kbMQDzzwgFy+deHCBZMCn5kzZ8pNAAsOkfBNdC90Wi32TJ6Fus6h0Bv1OJG2DqO+23BP1yZeSZe7aWenaWQw8djUtnD3zp+lIyIiIrIEJu2IMmLECLnLdteuXeHj44N//vlH5jdU1BtvvFEiAftOBZvvid9bHImJiXBxcZFBh8iNKHi9bt38RFaRWF5A9NHb2xvXr+fX+xcBibi+qILH4rXSTJ8+HVOmTCl8LGYsGFzQ3Whyc7DlpTfRzvcp+fhc1hY8uXjpPX1wN6JTsXnBaWjz9PANdMfDk1rJ3AoiIiIiqw0sit5QFyVu2Nu2bStnMAqIm/z7Jd5HHPejYOmSSMAWJWPFUidBBDuCyJcQ+RWCWAp169YtWQVK6Ny5s6waJWY2RDJ4QW5GUFAQvLxKr8wj9uoQB9G9ysnKxC8TxqOL/3j5+KrmKAZ+/e97uvbKqST8uSQCep0B9YO9MPDFMNg7codUIiIisjwKo9FovNeTe/bsec9vXNV5Fl9//TW6dOkiS9yKYEBUopo1axZeffXVwnMeffRRWSVKlI91d3eXsw0i90JUfhKBRFpamgwiRMnZt99+W+7HMXbsWMydO/eeq0KJGQuxJEq8l/g9iO4MKn4cMwwPNnwZLmoP3DJcQ6uZz0ChUtz1gxL7U/y1IlLurB3Yyhv9ng+F2k7FD5iIiEym1+tx5swZ2Q4LC4NKxX9fyPR73fsKLCzJqFGjsGnTJrkpnsjpEDt+jxw5ssQHIapUidwPkXvRo0cPzJs3r9jSJbFBntg078iRI3KZlEhAF0HGvWJgQeUFFStGD0Erv2fQwCUI2cbbaDpjIJROd59xOLPrBnavPi/bQQ/4odfIYChVJm87Q0RERGQZgYVYCiV21hY5DWUti5JvqlDIxGpbwMCCygsqfN07o4PPQBiMevi90g729d3K/cDEj+OxLddw6PfL8nFYz/ro9mQzubM2ERERkSXf697XYu0TJ07IfISCdnmBBZGtBxVKhwZoUzt/c0avQU3vKag4+OtlHP/zmnzcfnAAOj4cyJ8nIiKqdKJ6plguLojCO0X3BiOqqPsKLIrmTVR1DgWRtQYVK0cPQY7SHX18hkCttIdDU0+4dvO/a1Cxb/1FnPorfwPHrk80Res+Daup10REZGt0Op0swy+IHNSimwUTVRTLyxBVYklZMVORpXRBa69u8HLwg9JZjVrDmpe7lEkkZ+9Zcx5n/omVj3sMb46WPfIrmRERERHZVLnZ0lSk3CyRNQcVy0YOkkGFn2MjBHl2ks97PdEcKneHcoOKXauicG5fPKAAej4bjJCu9aqx50RERESV475zLO4FcyzIVoMKB4UTHvAbIp93eaAunEJql3mdwWDEzpWRiDqYAJGW1Pu5Fgh6IH9jRyIiIiKbybEgovyg4rv/CypgNKJrwyFwULhAXccZnoMDy/yIDHoDdiyPxIUjiXKZVN8xIWjWIX+zRyIiIiJrVCklAPbt24e8vLzKeCsiqwsqMv8vqGjVoBN8VIGAWoFaw4OhKGMzO73egG1LI2RQoVQq0P/5UAYVREREZPUqJbAYOHAgYmPzE0+JbIFOq8V3z/0vqGhYzx/BTr3ka54DA2Ff16XU6/RaA7YuOotLx5OgVCsw4MUwNGlbp5p7T0RERGShVaGsdPNuogoHFUtH9kcm8oMKHz8XdGv0AnSJ2XAM8oJLl9KTr3UaPbYsOovrEclQ2Skx8MUwNAotOweDiIioqqhUKvnFcEGbqDKw3CzRfVo2ZgAyjc6y7eWtxpDenyHrQDyUrnbwerJ5qcULtBo9Ni84jRtRqVDbKTFoYjgaBNfiZ09ERGYhgomOHTvy0yfLWwq1aNEi+Poy8ZRqviWj+yFD6yTb7m56DJ+0TAYVQq1hQVC5ltxgSJunx6avT+UHFQ4qPPxKKwYVREREVONUyozFiBEjKuNtiCzat+MHIT0nP3Bwc8rD2C83IvGr4/Kx64P+cGzuVepMxaYFpxB7/jbsHFUY8kpr1G3iUe19JyIiKspgMOD69euy3bBhQyiVlfJdM9m4CgcWZW2WJ5aBODo6omnTpnjkkUdQqxaXe5D1WzbxEaSl5/+l66bOwXPvb8DN+adgyNbBrq4LPAYElJpTIZY/xUbnBxVDX20Nv8YMKoiIyPx0Oh1WrFgh29OnT4e9fckZd6JqCyzEZnnHjx+HXq9HUFCQfO78+fNyzV5wcDAWLFiAN954A3v37kVISEhFfxsis1sx5UmkJulE1Aw3RQ6efHYJkpacBoyAurYjaj3TAgp18W96dFqRqH2mcPnTkEmtGFQQERFRjVbheS8xG9GnTx/ExcXh2LFj8rhx4wb69u2L4cOHy/Kz3bt3x+uvv165PSaqRj9MH4XkG9kyqKijtMcjD36BrH9iZVDh3M4XdV5tCzvv/JyLYiVlF4vqTylQ2ysxZFI46jb15LgRERFRjVbhGYvZs2dj+/btcHd3L3zOw8MDM2bMQL9+/TB58mS8//77sk1kjVZ/9AKSLt6CUalEkHMztG7wBLQ3MqFwVMHrsWZwDvcpcY1eZ8DWJWdx7UyyrP708MRWqNesZO4FERERUU1T4cAiLS0NN2/eLLHMKSkpCenp6bLt6ekJjUZjei+JqtnPs19H/JkYqNRO6FKrN/zdwgCNAfYB7qj1dBDUno6l76j9bQSunr4l96kQJWX9gxhUEBERkW1Qm7IUauzYsZgzZw46dOggnzty5AimTp2KRx99VD4+fPgwmjdvXnm9JaoGf8x/DzGHolDLuRE6+wyGs52nXDTo3qcR3B5qAIWy5D4VBr0B25dG4PLJJKjUSgx6KYwlZYmIiMimqE3Zu0LkTzz99NOysoB8M7Uazz33HObOnSsfiyTub7/9tvJ6S1TFdnw/B1d2HkOL2j0Q4tkZCoUSqlqOcpbCoeH/lv3dGVTs+O4cLh1PglKtwIAXWqJhCHfUJiIiItuiMBqNRlPeIDMzE5cvX5btxo0bw9XVFbZELPsSuSViaVjRfBOyPoe2/ohj3/6ELvVHwNuxvnzOuW0deD7SBEqH0mNwg8GIv5afw/nDiVCqRFARhsBw72ruORER0f0RVT0PHjwo2w888ICs6klk6r2uyRvkiUAiPDzc1LchMqvIo7twdPFyhPgNkkGF0c6I2k8Ew7lVnTKvEUHF3ysj84MKpQL9x7dkUEFERFZBBBJdu3Y1dzeohjEpsLh9+zaWLl2KyMhI+Vgkco8bN05GNUTWIvbyOez87BO4uTVDM/d28jmf58Lg2LTsxGsx0ffPqihEH0yQORf9ng9F49Ylq0QRERER2YoK72Nx9OhRNGnSROZTpKSkyEO0xXNi4zwia5CWfBO/T50Ejb0bOnoPlM+5dPK7a1Cxb/1FnNsXL7a3QN+xIWjStuyZDSIiIktjMBjknmPiEG0is85YiMTtoUOHYsmSJTJpWxBJ3M8//zxee+017N69u1I6SFRVNLk5+OmFp5Bt54K2nt3gaucFlYcDPAYGlnvdkU1XceqvGNnuOTIYzdr7cpCIiMiqiHu2ggI706dPh729vbm7RLYcWIgZi6JBhXwztRpvvfUW2rdvX1n9I6oSOq0W3z03CFkqF/g41Eczz/ySyV6PN4PSsewfi5M7ruPIxiuy/eCwZmjRpR5HiIiIiMiUpVAiK/z69eslno+JiYGbmxs/XLJoy8YNRCZcoFKo8YDfYPmcSwc/ODYvewnUuX1xcgmU0GloIFr1alBt/SUiIiKqsYHFU089JRO116xZI4MJcaxevVouhRo+fHjl9pKoEi198WFk5OXvnN3e90E4Kz2h8rCHx+Cyl0BdOJqInT9EyXabvg3RbmAAx4SIiIioMpZCffHFF1AoFBg1apRcpycSWsX6vJdeegmzZs2q6NsSVamVbw3H7RQjRNZ1Q2dvBDh3AoyA52NlL4G6euYWdiw7J88L7VYPnR9rIv/sExEREVElBBYiiJg3bx5mzpyJS5cuyedERShnZ+eKviVRlVr76cu4dSUdUCrgjjw8GDIR+uRcOLfzhVNQrVKviY1OxdbFZ+WeFc06+KL78CAGFURERESmBhZTpky553O//PLL+3lroiq1aeGHiDtxFUaVEi66LPxr1ELkHrgJpbs9PMtYApV4JR2bFpyGXmtAQLg3eo9uITfCIyIiIiITA4sTJ07c03lcJkKWZP8fK3Fp+wHo1Wo4a3Pw2HuLkbs2Qb7mJZZAOduVuCY5NhN//PcktHl6+Ad5of/4UKhUFU5JIiIisridt3v06FHYJqr2wGLnzp2V8psSVZeLpw/ixHc/QGtnD0dtHgZ89G8Y/kyT+RLOberAKbjkEqjbidn4bd5J5GXr4BvojkEvhUFtx790iYio5hDBxEMPPWTublANw69gqcZKvRmL7TPeQa6dPex1OnR6aQJqxdSGLikHSjc7eA5pXOKazNRc/D7vJHLSNajt74qHJ7WCfTn7WhARERFRPgYWVGM3wFvz8rPItnOCSm9As4E9EB48ABm7b8jXvR4tuQQqN0uLP/57ChkpufCo44Shk1vD0aXkMikiIiJrJ6p53rx5Ux6iTVQZGFhQjbRs9AC5q7bCaIRvS3/0H/02Utadl0ugnFr7wCm0drHzdRo9Nn9zGilxWXD2sJdBhbO7vdn6T0REVJW0Wi2++eYbeYg2kU0HFsePH0ffvn3h6emJ2rVrY8KECcjMzCx8ffny5TKJvLRDROcFdu3ahbZt28LBwQFNmzaV15F1W/bSEGTonGTb00eNp97+BqkbLkB3MxtKV7EEqkmx8w16A7YtjUD8xTTYO6kx5JXWcK+dfz0RERER1eDAIi4uDn369JGBwKFDh7B161ZERERg9OjRxXYGj4+PL3b0799fVkCoU6eOPOfKlSsYPHgwevbsiZMnT+K1116TO4f/+eefZvy/I1P89MHzuJ1skG13dR6GjVyIhC+OIPt4fjDp9WhTqIosbxLTv/+sPo8rp25BpVbKRG3v+q4cBCIiIqL7ZFJW6p49e7Bo0SK5Qd769evh7++P77//HoGBgXjwwQerbDA2btwIOzs7zJ8/H0plfmy0cOFChIeH4+LFizLgcHJykkeBpKQk/P3331i6dGnhc+Ia0dc5c+bIxy1atMDevXsxd+5cGYSQddm86GMkRsTJvSoa2vuiW5uXcPvXi/I1dR0neD7cBI7NvYpdc2TjFZzbEyc24kbfcSHwv+N1IiIiIqriGYuff/5Z3nyLm3exv0VeXp58Pi0tDZ999hmqkvi9xM7fBUGFUBBEiMCgNCtXrpS7gj/xxBOFzx04cEDOfBQl/p/E8+X93unp6cUOMr/Df67BpW374ORQGz18HkFn/9HQ3cyB0lkNz6FN4Du5bYmg4uzuWBzZdFW2xY7aTdrkz2QRERERUTUGFp988on8xn/JkiVy9qBA165dZf5DVerVqxcSEhIwe/ZsaDQapKamYtq0afI1seSpNGKmYsSIEcVmMcR7+Pr6FjtPPBbBQk5OTqnvM3PmTHh4eBQeDRo0qNT/N7p/VyKP4cSS7xFSpy8G1h8HP9dgQKmAa9d68JvaHq5d6kFxx+Z2l07cxD8/Rct2+8EBaNndnx89ERERkTkCi+joaHTv3r3E8+Jm+/bt2xV6TxEclJVwXXBERUUhNDQUK1askEuYxCyEn5+fXNIkgoKisxgFxAxEZGQkxo0bB1NNnz5dzsoUHDExMSa/J1VcWvItRH6+Fr0DX0GQR0coFSo4BnnB97W2Mkm7tF21Y8+nYvvSc7JCVEi3euj4cCCHgIiIiMhcORbiZl7kMwQEBBR7XixFaty45MZj9+KNN94oloBdmoL3FrMP4khMTISLi4sMOr788stSf+9vv/0WrVu3Rrt27Ur8P4jrixKP3d3di81sFCWqR4mDzC83Ph2xn+9EuO/D+Y/tslD/2Y5wDCq5m3aBWzcysfmbM9DrDAhs5Y0eTzeXf3aIiIhsbeftzp07F7aJzBpYjB8/HpMnT8ayZcvkjZmo1CRmBqZOnYr33nuvQu/p4+Mjj/tRsJRJ9MPR0VGWoC1KlKBdu3atXMJ0J/EDtXnz5mLPbd++vfAHjSyXNjELsV/ug7uDH/L0OYixO4UeM94oseSpqPRbOfjjvyehydGhblMP9BsXCmU55xMREdVUIpjo16+fubtBNUyFAwuxbMlgMKB3797Izs6Wy6LEN/kisHjllVdQ1b7++mt06dIFrq6uMhh48803MWvWLLmvRVFr1qyBTqfDs88+W+I9XnzxRfk+b731FsaOHSurRokgZNOmTVXef6o4TVwmbszbDweVK1LzEnEyZw1GLlxX7jU5mRq5q3Z2mga16rlg0EvhUNvzGxoiIiKiyqIwmriPu0ieFkuixMxASEiIvNGvDqNGjZIBgPh9g4ODZUAzcuTIEueJ4EPkX6xatarU9xEb5L3++us4d+4c6tevL2db7rYcqyiR6C3ySkS+hVhCRVVLcyMDcQuOQG2wQ0pePI4mLMfYn4rPOt1J7Kr969wTSLySDlcvBzz+Vju4ejlyqIiIyGaJ2z9x7yKI+xguC6bKuNc1ObCwdQwsqk/e9XQkLj4JpU6BW7mxOHJjJUYsWwUX97L3njAajPhzyVlcOpEEB2c1HnuzHWrVdanGXhMREVke8cVwwTJxUZhGlPEnMvVe976WQk2ZMuWezxWJ1ESVJe9qGpKWnpFBRVJuDA7F/Ii+H35YblAhHPjlkgwqlCoFBr4YxqCCiIiIqIrcV2AhNsK7F5xOo8qUe/E2kldEAFojEnOu4kDcOoQ9PRhNWnYo9zqxAd6J7ddlu9eoFtxVm4iIiMhSAoudO3dWXU+ISpF7PhW3Vp4DdAbEZ1/G/sQN8Gvljx5PvlTu53UtIhm7V5+X7Y5DAhHUyY+fLxEREZElVoUiqmo5kclI/iES0BsRm30R+2/+CndPHZ5695tyrxN7VYi8CpFfEfyAH9oPKr7XChERERFVvvvOsfj444/lhnR3y7dgjgWZIufsLST/FCWDihuZUTiQ9AeckYFxi8qfNcu6nYdN809Bm6uHf5AnHno2mEvziIiIiCwxx0Kr1cr28ePHy7xhY44FmSL7VBJS1kQBBiAmIwIHbm2GkzYLzy7fUO51mlwdNs4/hczUPHj5OWPAhDCo1NwAj4iIiMjiAot58+YVlpkS+z8QVUVJ2ZS10flBRdpJHEjZBnttHvrO+LTcClAGvQHblkbgVkwmnNzs8PCkVnB0seMAERERlUKpVKJ9+/aFbaLKcF9/ktq0aYNbt27JduPGjZGcnFwpnSAS9JkapKzKz6mIz4zA/pQ/odLrEDb8X2ga/kCZH5LYimXv2gu4diYZKjslBr0cDndvJ36oREREZVCr1Rg8eLA8RJuo2gMLT09PXLlyRbavXr0Kg8FQKZ0gEonWKaujoU/TIFN3C/tvbRO7N8KvdcBdK0Cd+isGZ/6JBRRA37Eh8Av04AdKREREVM3uK0R9/PHH0aNHD9StW1fmUYgpNJVKVeq5ly9frqw+kg1I334NeRdvQw8t9sT/Cp1RAw9P3LUC1OUTSdj380XZ7vJYUzRpU6eaekxERGS9xGx/dna2bDs7OzM/lqo/sFi8eDEee+wxXLx4Ea+++irGjx8PNze3yukJ2XRZ2YydMbJ9LGET0nXJcDVm4fnF5VeAunktHduXRQBGoGV3f7Tu06CaekxERGTdRDGeL774QranT58Oe3t7c3eJaoD7XlQ3YMAA+euxY8cwefJkBhZkEl1yDlLW5G9kdzn1MK7kRMNJm4thC38o97qstDxs/uYMdFoDGobWRrenmvHbFiIiIiIzqnC2znfffYfbt29jzpw5iIyMlM+FhoZi7Nix8PDgGne6O6NWj+RVkTDm6pCaG4Njt/+BnU6Hzq9Mglcd/zKv02n0MqgQe1Z41XVB/+dDoVSxogURERGROVX4buzo0aNo0qQJ5s6di5SUFHmITfHEc2KPC6K7Sf3tErRxWdAYsrDn5h8wGvXwf6AF2vR8tNw1oTt/iMLNq+lwcFFj8MthsHdiNQsiIiIic6vwHdnrr7+OoUOHYsmSJYVlynQ6HZ5//nm89tpr2L17d2X2k2qYrCMJyD6aCCMM2JfwO3L0GfDwMOLxN+eWe92Jbddx/nAiFEoFBoxvCQ8f52rrMxERERFVQWAhZiyKBhXyzdRqvPXWW4UbrhCVRhObidTf8is5RSTvxs2863AxZGH0gm3lfmBXTt/CgV8vyXa3Yc1QP7gWP2AiIiIia18KJXbgvn79eonnY2JimNBNZTJka5H8wzlAZ0RCZjQi0g/BSZuHx79aBrVd2TtlJ8dlYvvS/ApQod39EfZQfX7KRERERDVhxuKpp57CuHHjZKmyLl26yOf27duHN998E8OHD6/MPlJN2gRvTTT0qXnI0qZgf/IWqHV6tJswBj7+gWVel5upxeYFp6HN08O/uaesAEVEREQVp1Qq0apVq8I2kVkDCxFQiE3yRo0aJXMrBDs7O7z00kuYNWtWpXSOahaxV0VudCr0Ri323vwNWn0u6rcJQKcBI8q8Rq83YOviM0i/lQt3b0cMmBAGFStAERERmUQsX3/00bKLpRBVhMIoyuyYQOzaeOlS/rp3URFK7N5oS9LT02V53bS0NLk8jEqXeyEVt5adlUuZDt/chCtZZ+HurMX47/4s9yP758donN0dCztHFR5/qx1q13PlR0xERERkgfe6yoru1ti7d29cuHBBBhJhYWHysLWggu6NPlODlLXRMqi4knZCBhUu+iyMWbyx3OvO/nNDBhVQAP3GhjKoICIiqiTie2WNRiMPE79jJjItsBBLnk6fPl2RS8nGiL+sUtdfgCFDi/S8mziW+jcctXl45Itvyk3WvhGVgt1rLsh250ebICDcuxp7TUREVLOJL4lnzpwpD9EmqgwVztZ59tlnsXTp0krpBNVcWQfjkRuVAr1Rh/1JG6HQ5SF85DDUDQgu85q0pGxsXXJWJns37+SLNv0aVmufiYiIiKgak7dFwvayZcuwY8cOtGvXDi4uLsVeF7twk23TJmbh9qYrsn065R+kaZNQt0VddPvX+DKv0eTqsGnBGeRl6VAnwB09nw2WRQKIiIiIqIYGFmfPnkXbtm1l+/z588Ve440gGbUGpPwUDegMSMi6hPPpR+GqzsaIj5eVu2zq75VRSI3PgouHPQa9GAa1nYofJhEREVFNDix27txZuT2hGiVt6xVoE7KQp8vCwVub4azNwbNLfi73mpPbY3Dp+E0oVQoMeCEMLp4O1dZfIiIiIqrmHAuDwYB///vf6Nq1Kzp06IBp06YhJyfHxG5QTZIbnYLMfXGyfejWZhg06eg8eTJc3L3KvOZGdCoO/HJRth98shn8GntUW3+JiIiIyAyBxaeffor/9//+H1xdXeHv74958+Zh4sSJldAVqjGlZdflL427kHYM8dmX4Ns6AK17DC3zmoyUXGz79ixEtbvgB/zQsod/NfaYiIiIiMyyFGrlypVYsGABXnjhBflYJG8PHjwY3377LbeEt3GytOy68zBkapGWdxOnUnfBzSEPT723sMxr9Fqxs/ZZ5GRo4d3AFT1GBDFHh4iIqIoplUqEhIQUtonMElhcv34dgwYNKnzcp08feSMYFxeH+vXrV0qnyDplHYhHbnQq9AYdDiT9AQdNBkYu/a3ca3avPY+bV9Ph4KzGwBfCoLZnsjYREVFVU6vVePLJJ/lBU6VSVqTMrKOjY4kN87i5im0Tidq3N1+W7VOpO5Gdk4BuU9+Gk4trmdec2xeHc3vi8nfWHhcKd2+nauwxEREREZl1xkIsdxk9ejQcHP5XsSc3Nxcvvvhisb0sNmzYUHm9JIsvLZv8UxSgMyIu+5LMrWjUoTladulX5jU3r6Vj90/5uRidhjRGw9Da1dhjIiIiIjJ7YPHcc8+Vugs32a60LVegS8xGri4Lh5M2w81Zgyem/afM83MyNdiy6Az0OgMCwr3RbkCjau0vERGRrdNoNJg5c6ZsT58+Hfb29ubuEtliYPHdd99VTU/IKuVEpSBzf0Fp2U1Q593C6O82lXm+QW/Atm8jkJmSB486TugzJgQKJXfWJiIiIrJ2LANAFWbI1soqUML5tKNIyTiPnu/MgL1j2bkSh36/jBtRqVA7qGSytoNThfdoJCIiIiILYrWBxfHjx9G3b194enqidu3amDBhAjIzM4udc+TIEfTu3Vue4+Xlhf79++PUqVPFzjl9+jS6desmE9IbNGiAzz//vJr/T6xX+o7rMGRpkaa5hdMpO1G/aziC2nUv83yxq/bxP6/Ldq+RwajtX3ZiNxERERFZF6sMLERpW1HmtmnTpjh06BC2bt2KiIgImVReQAQZAwYMQMOGDeU5e/fuhZubmwwuCipYpaeno1+/fmjUqBGOHTuG2bNnY8aMGVi8eLEZ/++sgzYxCxn7Y2X7ePIOuLjk4V9TZpd5fmpCFv5aESnbrfo0QLP2vtXWVyIiIiKqela5DmXjxo2yxO38+fMLN3VZuHAhwsPDcfHiRRlwREVFISUlBR999JGciRA++OADec61a9fkOatWrZLJS8uWLZNJS6GhoTh58iS+/PJLOQNCZVcGu7k+AgoocCPrPLIyI/H86j/L/Li0Gr3cBE+bp4d/c090+VcTfrRERERENYxVzljk5eXJQKDoTpFOTvnr+sXMhBAUFCSXSC1dulQGDzk5ObLdokULBAQEyHMOHDiA7t27F6uEIGY0oqOjkZqaWubvLWY6ih62JjcqBcaYPOiNOkQkbUeP6e9DbWdX5vl7Vp9HSlwWnN3t0XdcKJQqq/xjR0RERETlsMo7vF69eiEhIUEuXRJBgwgCpk2bJl+Lj4+Xv4plT7t27cIPP/wggw5XV1e5ZGrLli1yt0lBvIevb/ElOQWPxWulEaXZPDw8Co+C2RBbYdQZELN8f2HCtmuoJ1q0f6jM86MOxiNyfzwUCsigwsXjf/ufEBERkXmIL2ebNWsmj6Jf1BKZwqL+JIngQKFQlHuIJU5iydKKFSswZ84cODs7w8/PD4GBgTIoKPjhEDMU48aNQ9euXXHw4EHs27cPLVu2xODBg+VrFSVqPaelpRUeMTExsCWRy7fAWeGJHF0mrmf8jafeW1jmuSnxWfjnx2jZ7vBwIOoHeVVjT4mIiKgs4kvWESNGyKPgC1ciU1nUn6Q33nijWAJ2aRo3bix/LfhhSExMlDt+i6BD5EYUvP7jjz/i6tWrcrlTQbAhnhPVoX777Tc8/fTTMiAR1xdV8Fi8Vhqx43jRXcdtSdbNVDhF2wMqIOrWX3jiPyvKzav4c8lZ6DQG1A/2QruB+cvPiIiIiKhmsqjAwsfHRx73o2DpkkjAFiVjRQlaITs7WwYUIuAoUPDYYDDIx507d8Y777wjq0SJZHBh+/btMj9DBCBU3KmPV6ChSzuk5sXDfVATeNSuc295FWNDoeQmeEREREQ1mkUthbofX3/9tdzL4vz587I61KRJk2T+g9izQhABhsi9mDhxIiIjI2U52jFjxsjpvp49e8pzxIyHSNwWS6bE62vWrMG8efMwZcoUM//fWZ4/Z3+KBs5tZfuCfhd6DHupzHOj78irEMEFERERWQ6Ro/rZZ5/JQ7SJatyMxf04fPiwLB8r9qsIDg7GokWLMHLkyMLXxXN//PEHPvzwQzkzIWYr2rRpIxO469atK88Rydfbtm2TwUe7du3g7e2N999/n6Vm73A96gR8YxpB4axAbMYZDP1qXrl5FbuYV0FERGTxCvb1IoKtBxYrV6686zli1qJgaVRZxL4We/bsqcSe1TzHZ36L9nWfht6ghfeo9mWWli2aV+EfxLwKIiIiIltitUuhqHqsfHUYgr0HyHaCyyU06dSxzHP3rMnPq3CSeRUhzKsgIiIisiEMLKhMe39Zhjo5YXC180SOPh0dpo8tP69iXzygAPqNDeF+FUREREQ2hoEFlSo9NRnRq35HkFcX+bj2v0KgtFfdPa9icCDqB9fip0pERERkYxhYUKlWTxqG5r79YKe0R55bLrw6B9xTXkX7QdyvgoiIiMgWWW3yNlWddTNfgZ0yEIFuYfJxg1EPFNsPpCjmVRAREVkf8e96o0aNCttElYGBBRVzJfIYEo6eR7cGY+Rj57Z1YN/ArdRP6cLRROZVEBERWSGxMfDo0aPN3Q2qYbgUior58/2pqOvRCt6O/jCqAY8BpS9tSk/Owa5V+XkV7QcGMK+CiIiIyMYxsKBCP0wfBaVDQ7T1zt/7w6N3I6jcHUp8QgaDETu+OwdNjg6+ge5oP5h5FURERES2jkuhSIo8uguZFzPQo+EY2CsdYR/oDrdu9Uv9dI5vvYb4i2mwc1DJ/SpUKsanRERE1kSj0WDevHmyPXnyZNjb25u7S1QDMLAg6LRa7J75MTrUHw03Oy8o3FWo/UwLKNQlA4aEK2k4vPGKbHcf3hwePs78BImIiKxQdna2ubtANQy/aiZ8P/VpNPcdAl+nAOgVOtQZ2woq15LfXGhyddi+NAJGgxFN29dBUCc/fnpEREREJDGwsHHHdmyAd0Ywmrm3g9FoRJ2R4bDzcyn13D2rzyP9Vi5cazngoRFBLE9HRERERIUYWNj4EqhL321CK59+8rHjQ75wCqldZmnZqIMJEKWu+44JhYOzXTX3loiIiIgsGQMLG7b21TFoV+8pKBVKpHsmw3tA81LPy0jJLSwt225gAOo186zmnhIRERGRpWNgYaP2rluK1k6Pw0HlhNvaWLR4Y0ipS5tEadntyyJYWpaIiIiIysWqUDYoOyMDTv8A7q7eyNGmo8m0gVDYqUo9l6VliYiIah7xZWK9evUK20SVgYGFDTr6/75BY9fO0Bu0SGubjmY+HqWeV6y07NMsLUtERFRT2NnZYfz48ebuBtUwXAplYw5+9R0au3SW7bO5W9D+meGlnidLyy4797/Ssg+wtCwRERERlY0zFjYk5fQ11I0LkOHk+eR/0H/BZ2Weu2fNeaQn5bC0LBERERHdEwYWNkKfoUHq9+fgoHJFbGY0aj0TDrWdXdmlZQ8UlJYNYWlZIiKiGkar1WL+/PmyPXHiRLk0ishUXAplI05/tV4GFbc1N3HB/m+0fmhoqedlpeXhnx/zS8u2HdAI9Zp5VXNPiYiIqKqJTXHT0tLkIdpElYEzFjYg9eIN1M70BxTAucQ/8OyqH0s9T/zFsvOHKORl6+DT0A0dHg6s9r4SERERkXXijIUNuPDfP+QmePFZF9DylWfKPC9yfzyunUmGUq1A79EtoFLxjwcRERER3RveOdZwexZ/g3pOLeVsxFXVIYQ80LfU89KTc7B33QXZ7jSkMWrXc63mnhIRERGRNWNgUYPlZGXC5YyTbMdmnMa/Zn9d6nmipOzfK6OgzdXDr7E7WvdtWM09JSIiIiJrx8CiBtvyxlTUcWkCg1EPhwH1yqwCdXZ3LGKjU6G2U6L3cyFQKrkDJxERERHdHyZv11B7NyxDU8eesh2TdwJdh75e6nm3E7Oxf8NF2e78WFN4+jpXaz+JiIio+ikUCvj4+BS2iSoDA4saSKfV4tavJxBQbxh0Bg3avTuu1PMMBiP+WhEJncYA/yAvhPXwr/a+EhERUfUT+1a8/PLL/OipUnEpVA30/atPoLl3b9nO8E+Do7d7qeed3HEdCZfTYOeoQq9RwVBwCRQRERERVRADixrmxM5fUSsvBO72taExZCH0hdI3wkuOy8Sh3y/L9oNPNoN77fwkbyIiIiKiimBgUcOWQB2ZvwAtanWTjz36NYHSseRqN73egL+WR8KgM6JRWG206FLXDL0lIiIic9FqtViwYIE8RJuoMjDHogZZ9fYz8K/1IJzVbtCoc+H/UJNSzzu+9RqSrmfAwVmNns8GM2mLiIjIxoj9rZKSkgrbRJWBMxY1ROTRXciO0aKFR2f52Pdf4VCoSw6vCCiObroq2z2GB8HFw6Ha+0pERERENQ8Dixrin1kfokntzrBXOULpbQ/nNnVKnKPXGrBj+TlZDapJ2zpo2r7kOUREREREFcHAogZY9e4Y6O190dy9vXzsNahpqRWeDm+8jJS4LDi52aHHiOZcAkVERERElcZqA4vjx4+jb9++8PT0RO3atTFhwgRkZmYWO+evv/5Cly5d4ObmBj8/P7z99tvQ6XTFzjl9+jS6desGR0dHNGjQAJ9//jmsyfXok0iJjEeoZxeolXawb+QOxxa1SpyXcCUNJ7Zdl22RV+Hkam+G3hIRERFRTWWVgUVcXBz69OmDpk2b4tChQ9i6dSsiIiIwevTownNOnTqFQYMGYcCAAThx4gTWrFmD33//HdOmTSs8Jz09Hf369UOjRo1w7NgxzJ49GzNmzMDixYthLba8/zrsHX3Q2K2VfOwxMKDETIRYAvX3yiiI3KygTn4IbJW/0yYRERERkU1Xhdq4caPcMXL+/PlQKvNjo4ULFyI8PBwXL16UAYcIJMTj999/X74unhOzEcOGDcMHH3wgZzFWrVoFjUaDZcuWwd7eHqGhoTh58iS+/PJLOQNi6U7v24IcgxM6enWHUqGEY3AtOAR4lDjv2NarSI3PXwL14LBmZukrERERWQ7xJaSHR/49w51fSBLZVGCRl5cnA4GCoEJwcsrf4G3v3r0yiBDniOVNRYlzcnNz5ezEQw89hAMHDqB79+7yvQr0798f//73v5GamgovL69Sf29xFJ31MJfwrgORezkF9SIaAgrAvX9AqRvhHdt6Tba7PdUcji52ZugpERERWRLxBe1rr71m7m5QDWOVS6F69eqFhIQEuXRJzDiIIKBgiVN8fHxhgLB//3789NNP0Ov1iI2NxUcffVTsHPEevr6+xd674LF4rTQzZ86UEX7BIfIyzCkwL1z+6ty6DuzruhR7TVR/2vl9FAx6IwJbeaNpO1aBIiIiIiIbCCxEcCCm48o7oqKi5JKlFStWYM6cOXB2dpaJ2YGBgTIoKJjFELkTIvB48cUX4eDggObNm8ucC6HoTMf9mj59OtLS0gqPmJgYmIs+UwNdSi6gUsC9b6MSr5/ZeQOJV9Jh76hC96eDONVJRERERFVGYbSg7RbFDpDJycnlntO4ceNiS5cSExPh4uIib5rd3d2xevVqPPnkk4Wvi/89MUMhljVdvXoVISEhOHz4MDp06IBRo0bJpUy//vpr4fk7d+6UMyIpKSmlLoW6k7hezFyIIEP8/tXNqDNAI3bRblw8tyL9Vg5++ugQdBoDHnomCKHd/Ku9b0RERGSZtFotli9fLtui+I1YGkVk6r2uReVY+Pj4yON+FCxdEgnYIqdClKAtSgQc9erVk22xLEosXWrbtq183LlzZ7zzzjvyh6vgB2r79u0ICgq6p6DCEojdte8MKkQwtWtVlAwq6jXzREjX/P9/IiIiooJ7BVFls6BNVOOWQt2Pr7/+Wu5lcf78eVkdatKkSTL/QexrUUAshTpz5owsRfvxxx9j1qxZ+M9//gOVSiVfHzFihJz9GDdunDxHVJKaN28epkyZAmsWdSABMZGpUNkp5Z4VpW2WR0RERERUmSxqxuJ+iOVMomys2BQvODgYixYtwsiRI4uds2XLFnz66aeyilOrVq3w22+/YeDAgYWvi2mdbdu2YeLEiWjXrh28vb1leVprKDVblux0DfatvyDbHR8OhKevs7m7REREREQ2wGoDi5UrV971nL///vuu54i9Lvbs2YOaYvfq88jL1sGnoRta9zFvxSoiIiIish1WuxSKSrp8MgmXjt+US596jgyGUsXhJSIiIqLqwTvPGiIvW4t/foqW7Tb9GsKngZu5u0RERERENsRql0JRcfs3XEJ2mkbmVHQYXHIHbiIiIqKixF5gRJWJgUUNcCM6Fef25peM6/lsENR2+VWviIiIiEojqmK++eab/HCoUnEplJXTavTY+UOUbLfs7o96zaxj/w0iIiIiqlkYWFi5I39cQXpSDly9HND5X03M3R0iIiIislFcCmXFxJ4Vp3fdkO0ew4Ng78ThJCIiorvTarVYtWqVbD/zzDOws7Pjx0Ym452oFXN2t8cTb7fHlVNJCAj3Nnd3iIiIyEoYjUZcu3atsE1UGRhYWDnv+q7yICIiIiIyJ+ZYEBERERGRyRhYEBERERGRyRhYEBERERGRyRhYEBERERGRyZi8TURERGSDWGKWKpvCyBpjJklPT4eHhwfS0tLg7u5eWeNCRERERGRV97pcCkVERERERCZjYEFERERERCZjjgURERGRjdHpdFi7dq1sDxs2DGo1bwnJdPxTRERERGRjDAYDLly4UNgmqgxcCkVERERERCZjYEFERERERCZjYEFERERERCZjYEFERERERCZjYEFERERERCZjVSgTFWxcLnYlJCIiIrIGGo0Gubm5hfcw9vb25u4SWaiCe9yCe97yKIz3chaV6caNG2jQoAE/ISIiIiKqsWJiYlC/fv1yz2FgYSJR+zkuLg5ubm5QKBQwRxQpAhsx2O7u7tX++9O94ThZB46TdeA4WQeOk+XjGFmHdDPf64k5iIyMDNSrVw9KZflZFFwKZSLxAd8teqsO4g8aAwvLx3GyDhwn68Bxsg4cJ8vHMbIO7ma81/Pw8Lin85i8TUREREREJmNgQUREREREJmNgYeUcHBzwwQcfyF/JcnGcrAPHyTpwnKwDx8nycYysg4MV3esxeZuIiIiIiEzGGQsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsrlJKSgmeeeUbWMvb09MS4ceOQmZl51+sOHDiAXr16wcXFRV7bvXt35OTkVEufbVFFx6lgM5qBAwfKTRd//fXXKu+rLbvfcRLnv/LKKwgKCoKTkxMaNmyIV199FWlpadXa75pu/vz5CAgIgKOjIzp16oTDhw+Xe/66desQHBwszw8LC8PmzZurra+26n7GaMmSJejWrRu8vLzk0adPn7uOKVX/OBW1evVq+W/Qo48+yqGwwHG6ffs2Jk6ciLp168qk7ubNm1vE33sMLKyQuAmKiIjA9u3bsXHjRuzevRsTJky4a1AxYMAA9OvXT/5hPXLkCCZNmnTXHRSpesepwFdffWWWndxt0f2OU1xcnDy++OILnD17FsuXL8fWrVtlQEKVY82aNZgyZYqsgnL8+HG0atUK/fv3x82bN0s9f//+/Rg+fLgcgxMnTsgbIXGI8SHLGKNdu3bJMdq5c6f890jsIiz+PYqNjeUQWdA4Fbh69SqmTp0qg0GyvHHSaDTo27evHKf169cjOjpaBu/+/v7mHy4jWZVz584ZxbAdOXKk8LktW7YYFQqFMTY2tszrOnXqZHz33XerqZdU0XESTpw4YfT39zfGx8fL9/jll1/4gVrgOBW1du1ao729vVGr1VZRT21Lx44djRMnTix8rNfrjfXq1TPOnDmz1POHDRtmHDx4cIm/81544YUq76utut8xupNOpzO6ubkZV6xYUYW9pIqMkxibLl26GL/99lvjc889Z3zkkUf4QVrYOH3zzTfGxo0bGzUajcWNDb+utjLimx6xXKN9+/aFz4kpZTHzcOjQoVKvERGveK1OnTro0qULfH190aNHD+zdu7cae25bKjJOQnZ2NkaMGCGnRP38/Kqpt7arouN0J7EMSiylUqvVVdRT2yG+iTt27JgchwJiPMRjMV6lEc8XPV8Q3/aVdT5V/xiV9nedVqtFrVq1OBwWNk4fffSRvF/gLKzljtPvv/+Ozp07y6VQ4p6uZcuW+Oyzz6DX62FuDCysTEJCgvyBL0rczIi/nMVrpbl8+bL8dcaMGRg/frxcttG2bVv07t0bFy5cqJZ+25qKjJPw+uuvy+DvkUceqYZeUkXHqahbt27h448/vudlbnT3z1P84yj+sSxKPC5rTMTz93M+Vf8Y3entt99GvXr1SgSEZN5xEl84Ll26VC6rIcsdp8uXL8slUOI6kVfx3nvvYc6cOfjkk09gbgwsLMS0adPkmvryjqioqAq9t8FgkL++8MILGDNmDNq0aYO5c+fK5NNly5ZV8v9JzVaV4yS+gfj7779lfgVZ7jgVlZ6ejsGDByMkJEQG7kR0d7NmzZKJwb/88otMVCXLkJGRgZEjR8qgwtvb29zdobvc14kvxRYvXox27drhqaeewjvvvIOFCxfC3DhvbyHeeOMNjB49utxzGjduLJfH3JnMo9PpZKWaspbOiIoBgrj5KapFixa4fv26yX23JVU5TiKouHTpklyaU9Tjjz8uE+hE8iOZf5yK/iMsCiK4ubnJGyQ7OzsOTyUQNzQqlQqJiYnFnhePyxoT8fz9nE/VP0YFRNEDEVjs2LED4eHhHAoLGifx749IBh4yZEiJLybFTK5IEG7SpAnHzMzjVHBfJ/7NEdcVvacTMxxiaZW9vT3MhYGFhfDx8ZHH3Yg1daLEmFiPJ6LUghtS8cMvypOVRpQvE1PO4i+Fos6fPy9LmpJljJP4lv35558v9pwomylml4r+RU/mHaeCmQqxhl+U+BMzTfzWtfKIfxDFWPz111+FZS7FeIjHopJdWeMoXn/ttdcKnxNVvsTzZBljJHz++ef49NNP8eeffxbLayLLGCdRrvnMmTPFnnv33Xfllyjz5s2TlbzI/OMkdO3aFT/++KM8r6C6p7inEwGHOYMKydzZ43T/BgwYYGzTpo3x0KFDxr179xqbNWtmHD58eOHrN27cMAYFBcnXC8ydO9fo7u5uXLdunfHChQuyQpSjo6Px4sWLHAILGqc7sSqU5Y1TWlqarDgUFhYmf35E9a6CQ1RTIdOtXr3a6ODgYFy+fLms3DVhwgSjp6enMSEhQb4+cuRI47Rp0wrP37dvn1GtVhu/+OILY2RkpPGDDz4w2tnZGc+cOcPhsJAxmjVrlqyctn79+mI/MxkZGRwjCxqnO7EqlGWO0/Xr12VVtUmTJhmjo6ONGzduNNapU8f4ySefGM2NgYUVSk5Oljc+rq6uMlgYM2ZMsb+cr1y5Im9Id+7cWew6Ubasfv36RmdnZ2Pnzp2Ne/bsMUPvbUdFx6koBhaWN07iV/G4tEOcS5Xjv//9r7Fhw4byZlSUYjx48GDhaz169JA3PHeW/G3evLk8PzQ01Lhp0yYOhQWNUaNGjUr9mRFBIFnOON2JgYXljtP+/fvll1wiIBGlZz/99FOL+HJLIf5j3jkTIiIiIiKydqwKRUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREREREJmNgQUREVuHw4cN46KGH4OTkhODgYBw9ehSLFy/G0KFDzd01IiICoDAajUZ+EkREZMkOHjyInj174qOPPsKjjz6Kt956C3q9HhEREVi/fj3atGlj7i4SEdk8BhZERGTxunTpgqZNm2LlypXy8dq1azF8+HA88sgj2LBhg7m7R0REXApFRESW7saNGzhw4ABefPHFwufUajXEhPuHH35o1r4REdH/MMeCiIgsWmRkpPy1bdu2hc9FR0ejY8eOCAsLM2PPiIioKAYWRERk0dLS0qBSqaBQKOTjlJQUfPHFF3B2djZ314iIqAgGFkREZNFat24tE7U///xzREVFydyKgIAAnDt3DteuXTN394iI6P8wsCAiIosmkrZFNah58+bJ6k/16tXDtm3b4O/vjwEDBpi7e0RE9H9YFYqIiIiIiEzGGQsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIjIZAwsiIiIiIoKp/j9TEBB2yyvUGgAAAABJRU5ErkJggg==", "text/plain": [ "
" ] @@ -355,8 +361,9 @@ "ax.plot(np.array(alpha_grid), np.array(profile_kf_dynamax), label=\"KF (Dynamax)\", color=\"C1\")\n", "ax.plot(np.array(alpha_grid), np.array(profile_ekf_cb), label=\"EKF (Cuthbert)\", color=\"C2\")\n", "ax.plot(np.array(alpha_grid), np.array(profile_ekf_dynamax), label=\"EKF (Dynamax)\", color=\"C3\")\n", - "ax.plot(np.array(alpha_grid), np.array(profile_ukf_dynamax), label=\"UKF (Dynamax)\", color=\"C4\")\n", - "ax.plot(np.array(alpha_grid), np.array(profile_pf_cb), label=\"PF (Cuthbert)\", color=\"C5\")\n", + "ax.plot(np.array(alpha_grid), np.array(profile_enkf_cb), label=\"EnKF (Cuthbert)\", color=\"C4\")\n", + "ax.plot(np.array(alpha_grid), np.array(profile_ukf_dynamax), label=\"UKF (Dynamax)\", color=\"C5\")\n", + "ax.plot(np.array(alpha_grid), np.array(profile_pf_cb), label=\"PF (Cuthbert)\", color=\"C6\")\n", "ax.axvline(\n", " true_alpha, color=\"gray\", linestyle=\"--\", label=f\"True $\\\\alpha$ = {true_alpha}\"\n", ")\n", @@ -364,7 +371,7 @@ "ax.set_ylabel(\"Profile log-likelihood $\\\\log p(y_{1:T} \\\\mid \\\\alpha)$\")\n", "ax.legend()\n", "ax.set_title(\n", - " \"Profile likelihood: KF vs taylor_kf vs PF (2 states, 1 control, 1 output)\"\n", + " \"Profile likelihood: KF vs EnKF vs EKF vs UKF vs PF (2 states, 1 control, 1 output)\"\n", ")\n", "plt.tight_layout()\n", "plt.show()" @@ -376,7 +383,7 @@ "source": [ "## Profile Score\n", "\n", - "For each filter type (KF, EKF, PF), we can also compute $\\nabla_\\alpha \\log p(y_{1:T} \\mid \\alpha)$ with direct auto-differentiation." + "For each filter type (KF, EnKF, EKF, UKF, PF), we can also compute $\\nabla_\\alpha \\log p(y_{1:T} \\mid \\alpha)$ with direct auto-differentiation." ] }, { @@ -411,12 +418,13 @@ "text": [ "Filter | max score | time (s)\n", "------------|-----------------|--------\n", - "KF (Cuthbert) | 182.1674 | 2.160\n", - "KF (Dynamax) | 182.1674 | 1.832\n", - "EKF (Cuthbert) | 182.1674 | 2.974\n", - "EKF (Dynamax) | 182.1674 | 1.940\n", - "UKF (Dynamax) | 182.1674 | 1.840\n", - "PF (Cuthbert) | 201.7599 | 12.774\n" + "KF (Cuthbert) | 182.1674 | 3.272\n", + "KF (Dynamax) | 182.1674 | 2.088\n", + "EKF (Cuthbert) | 182.1674 | 2.855\n", + "EKF (Dynamax) | 182.1674 | 2.117\n", + "EnKF (Cuthbert) | 183.2659 | 2.371\n", + "UKF (Dynamax) | 182.1674 | 2.245\n", + "PF (Cuthbert) | 201.7599 | 7.428\n" ] } ], @@ -427,16 +435,19 @@ "profile_ekf_cb, t_ekf_cb = profile_score(\"ekf\", \"cuthbert\", alpha_grid)\n", "profile_ekf_dynamax, t_ekf_dynamax = profile_score(\"ekf\", \"cd_dynamax\", alpha_grid)\n", "\n", + "profile_enkf_cb, t_enkf_cb = profile_score(\"enkf\", \"cuthbert\", alpha_grid)\n", + "\n", "profile_ukf_dynamax, t_ukf_dynamax = profile_score(\"ukf\", \"cd_dynamax\", alpha_grid)\n", "\n", "profile_pf_cb, t_pf_cb = profile_score(\"pf\", \"cuthbert\", alpha_grid)\n", "\n", - "timings = {\"KF (Cuthbert)\": t_kf_cb, \"KF (Dynamax)\": t_kf_dynamax, \"EKF (Cuthbert)\": t_ekf_cb, \"EKF (Dynamax)\": t_ekf_dynamax, \"UKF (Dynamax)\": t_ukf_dynamax, \"PF (Cuthbert)\": t_pf_cb}\n", + "timings = {\"KF (Cuthbert)\": t_kf_cb, \"KF (Dynamax)\": t_kf_dynamax, \"EKF (Cuthbert)\": t_ekf_cb, \"EKF (Dynamax)\": t_ekf_dynamax, \"EnKF (Cuthbert)\": t_enkf_cb, \"UKF (Dynamax)\": t_ukf_dynamax, \"PF (Cuthbert)\": t_pf_cb}\n", "profiles = {\n", " \"KF (Cuthbert)\": profile_kf_cb,\n", " \"KF (Dynamax)\": profile_kf_dynamax,\n", " \"EKF (Cuthbert)\": profile_ekf_cb,\n", " \"EKF (Dynamax)\": profile_ekf_dynamax,\n", + " \"EnKF (Cuthbert)\": profile_enkf_cb,\n", " \"UKF (Dynamax)\": profile_ukf_dynamax,\n", " \"PF (Cuthbert)\": profile_pf_cb,\n", "}\n", @@ -455,7 +466,7 @@ "outputs": [ { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -484,7 +495,7 @@ "outputs": [ { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxYAAAHqCAYAAACZcdjsAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjgsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvwVt1zgAAAAlwSFlzAAAPYQAAD2EBqD+naQAA19tJREFUeJzs3QV4U2cXB/B/PRVKkRanaHEt7u7usOEyZcKADTZkG7KNjY1tuMMGDNfhLsXdSvEixaFU0jRpvue8XfqlpYVKSu3/e54Lt9Gb2yS9577nnNfKaDQaQURERERElATWSbkzERERERERAwsiIiIiIrIIjlgQEREREVGSMbAgIiIiIqIkY2BBRERERERJxsCCiIiIiIiSjIEFERERERElGQMLIiIiIiJKMgYWRERERESUZAwsiBJJr9dj+PDhyJcvH6ytrdGuXTt1uZWVFcaOHRt1uwULFqjLbt68yX1NZCEffvghGjduzP1JFhfzO5ySbsuWLXBxccGjR4+4O9M5BhaUrpgO4k2LRqOBl5cXPv74Yzx48MCizzVv3jxMmjQJnTp1wsKFC/H555/jbZEgpW/fvihcuLB6jTlz5kSdOnUwZswYpBf16tVD6dKlX7l8586dcHJyQsWKFfH06VN1WYECBaL93s0XrVaLlLJnz544t0uWZcuWRd1WXkOrVq1eeYzFixfDxsYGzZo1i3otcT2evA/SwufvdfulW7dub3yeGzduYM6cORg5cmTUZf7+/vj2229RpUoVZMmSBdmzZ1fvoR07dljs9f37779JPuCcMGEC1q5di9Ti6NGjKkjz9vaGnZ2d+h2kdvfu3VO/h9OnTyO1SW378+LFi2pfva0TW9OmTVPfAzHJ91eRIkUwceLEt7IdlHJsU/C5iZLNd999h4IFC6oDsQMHDmD69OnqoOD8+fPqoNQSdu3ahTx58uDXX3+NdnloaChsbZPvo3X16lVUrlwZjo6O6NevnzogvX//Pk6ePIkff/xRHVylV7LPW7dujWLFiqkDxqxZs0ZdV758eXzxxRev3Mfe3h4p7ZNPPlG/s5iqV6/+2vv9/fff6NOnDxo1aqQORuVA3UTO1vfq1Sva7eU9kZY+f7HtF3k/v8mUKVPU49evXz/qsnXr1qn3v4wc9u7dW40oLlq0SO0nOQkggXhSyWuYOnVqkoILCSzkZIRphDOlyWuSIK1s2bIoVKgQrly5grQQWMj3nLxX5HOfmqS2/SmBhewrCbLj89myRGAhQb18b8X03nvvYejQoWp7MmXKlOzbQimDgQWlS82bN0elSpXU+oABA5AtWzZMnjxZHXx079491vsEBwfD2dk53s/x8OFDuLm5vXK5+cFfcpBAJigoSJ2t8/T0fGWb3qaE7rOk2Lt3rwoq5Ax4zKBCSJD37rvvIjWqXbu2OphMCBnNkAPkBg0aqPdtzPeV7IfU+nrj+/lLzH4JDw9XAdf7778f7XIJMm7fvq0OakzkNnLgOXr0aIsEFunRBx98gC+//FIFpTKylNIHwskhJCTEYieU3iQj7M/E6tixIwYPHowVK1aok2KUPjEVijIEOTgzpVAIOZsi+Z7Xrl1DixYt1NmTd955J+pgWc58S+2Eg4ODOjv+888/w2g0qutlSFmGt3fv3o0LFy5EpXBIekdC8nM3b96sDqzkwFyev2XLlurx3kS2OW/evK8EFcLDwyPW56lbt656DldXV3WGeMmSJdFuI1/0MnQvfwzlwEwOWO/evRvtNq/bZxEREfjtt99QqlQpdQCcI0cOdXbq2bNn0R7jxYsXuHz5svo/Ifbv36/2jwylS1AhB6pJJak5MrIU2wiPr6+v+j3++eefUQezcruiRYuq1yfPX6tWLWzfvh3JYfny5ep3IGcZ169fb5Fg9fjx4+o1SdpeTFu3blXXbdy4Uf388uVLfPbZZ+oMp3wG5H0lZ/5lVMwSn7+kkBGQx48fq1Ecc/LeMw8qhGy7vFfv3LmjXtPrvOl3LO9/Ga0Q5qlbJvIdUaNGDXU/+RzJ52nlypXRnkNuL98v8jsw3d/8zK585uSASz4/su3ymmS0JaY//vhDXScHy5L2JUFczM90fMlzJXWk66+//lIpaKbtkbTMbdu2vXImW7ZZXlfu3Lnx0Ucf4fnz57GmP8pZdgkU5fHkhMFPP/0UdRv5njWNckmwaNqPpvQb02OcOHFCbYc8hillTk689O/fX71m+R2XK1cu1s9DUiR1f8pI2/fff6/SXGVfyWdQtj8sLCza7eL6OyO3N72nZJ907txZrcv+jPm3ypR+Kb8rCcBln5QsWRKrV6+O9pjyPLGldMWsH5THk79hchLI9Fzy+zCR7xEZyZETDJR+MbCgDEEOhoX5Aal8gTdt2lR92clBgZxNkeChTZs2alRAckLlLKsEFsOGDcOQIUPU/dzd3VXee/HixdUBvqzLUqJEiXhvj9xeDpTlQF3SN0aNGqX+mMqBzJtyYSWgkHxySQt6E/nil+eRWoQRI0bghx9+UH9ApJDO/DZdunRRefyS/zpw4ED1h0W2JeYf/tj2mZAgQvZRzZo1VZqK/MGXs8pyWzlgM1mzZo3aT/J/fB08eFAdHErqi9RXxDx4NJHnkQNO80XOVL7uAEACLjmIj+mff/5R+8P0R1n+sMpBp/xxlmDj66+/Rv78+eN9oC0HtTG3TRZTsGpu1apVKmCTg6INGzbEeZAiaUYxHy/mwYc5OfiU1Iy4Xq8cEMrvy3SmX9KX5PcrB4SSviDbcenSJVjq8xfXfpEg9XUOHTqkDlgqVKgQr+cOCAhQB5dvOmP9pt+xvMdNxeKmz7wsJvK+l22SNDBJd5KgVd4/mzZtirqN3F4OFuWEgun+8rimQLdatWoqcJYz3fJ4EkjLgbAE7SazZ89WKWRyACiXyzbLZ/rIkSNICfL8PXv2VPUE8trlZzkpY/79JPtWAgkJKH755Rf1vpo5cyaaNGkS7ftByMkI+e6Vg365rXzPygiAnCAR8v0hzyMGDRoUtR/l82Ly5MkTNWom+0X2kfxOJUVVDnLltvL5kvq4zJkzq4Nw2dephYzwyQib1JDJ3yH5jpLv5fjUHsUk+0TeK0KCk9j+Vvn5+aFr165qf8nzmN63iTlpIvta/ibK78z0XPI5MicBt3yGKR0zEqUj8+fPlyM1444dO4yPHj0y+vv7G5ctW2bMli2b0dHR0Xjnzh11u969e6vbffXVV9Huv3btWnX5uHHjol3eqVMno5WVlfHq1atRl9WtW9dYqlSpV7ZB7j9mzJhXtunGjRvq55cvXxrd3NyMAwcOjHa/gIAAY+bMmV+5PKbz58+r1yKPWb58eeOnn36qtjs4ODja7Z4/f27MlCmTsWrVqsbQ0NBo10VERKj/dTqd0cPDw1i6dOlot9m4caN6/NGjR0ddFtc+279/v7r877//jnb5li1bXrnctC/k/zeR/Zs1a1b1GmQ/P3z4MM7benp6qseNuZj/HmIzc+ZMdbtz585Fu7xkyZLGBg0aRP1crlw5Y8uWLY0JtXv37li3y7Tcv38/2mvInTu30dbW1livXr1Xfp/m4nq8N+3XESNGGO3s7IxPnz6NuiwsLEy9H/v16xd1mbwPP/roo2T7/L1uv5g+J3F599131ePFh5+fn1Gj0Rh79uz5xtvG53cs+ySuP5shISHRfpbPlnyuzN9HwtnZWX2WYurfv78xV65cxsePH0e7vFu3bur3YXr8tm3bxvq9Ywmve31x7V9ra2tj+/btjQaDIdbvGPnc2tvbG5s0aRLtNn/++ad6rnnz5kX7zMtlixYtivb+zJkzp7Fjx45Rlx07dizO97vpMWbMmBHt8t9++01d/tdff0X7HVWvXt3o4uJiDAwMjLo8Pt8dybE/T58+rW4/YMCAaJcPHTpUXb5r1643bqN8j5i/v1asWKFuK5+52G4r161atSrqshcvXqj3YYUKFaIuk+eJ7XXE/Nsm5L0pv4O4TJgwQd3nwYMHr9kTlJZxxILSJUmTkJEFOXMmZ3pkZEDOksuwesx82JiFd3Km2nSWx0RSo+S73HTWLCnkTJCMBEiuufmZWnneqlWrqhSr15F0AqmvkFQZGd2Qs21SCCpn4OVspvnzyBnhr7766pVUGtOwtqTHSHqAdDExv42McshZJ/OzrXHtM0mjkjN/cjbX/PXImSnZ7+avR84Oyn6MrbAvNpI2Iq9BXpukcb2O7Dt5zeZLzOLmmDp06KDO0MkZexMpMJbRIzmLZyK1NDLEL2f3EkPOQMbcNlli1onIyJKMCslZvzelU7Rt2/aVxzONOMRFXpOcITZPdZA0CHk/xny9cgZcimST8/MX2355U2crORstoytvIqNVcuZV9qOM1L1JUn/H5r8vOesu6X4yMhGfUS35TMhIldQQybr550h+p/JYpseR7ZTUrmPHjiGlSUMBGWGS36O03I7tO0ZGYHQ6nUqtM7+NjIzKZzrmd4y8V8xrh6T5gqRZXb9+Pd7bJaNCMWtq5Ltd3lvmNT4yyiLf9VKzJuk7KU22UZhGx01MTSli+z5OKhlFat++fdTP8juR781Tp06p0T5LM3125b1N6ROLtyldklxoKW6Vg0Y5KJV0pph/+OQ6OYAzd+vWLfVFG7NjhWnoWK5PKtOBiynvPKY3HUALeW0yzGwwGNRBsOTGSx6ypAZIypAc2JnST2Jr2Wpiej2yf2KSwELy2d+0z+T1yIFPbPUdSS0ol1QQ+SMnqRByQCBBjARgsZEUqZh5928i92nYsKFKD5K8ZiFBhrxOCTpMJPVCDuRlv8v+lFQNSf+QfOH4KFOmTLy2TbZF0m8kDUmCjtelaMjvIaGvV9JL5Pcqr1FSbISsy34wfz/Ke0kKxyUwkABRUtHk9yCpVJb6/CVkv8QUWwqZOflcSEAjnw05GSCf6TdJ6u9YPoPjxo1TQb95Slp82o1Kb38J7mbNmqWW132O5LMgB+tysC2fD0kn6tGjh0pDfNvkO0Z+r5KWldDvGAkY5P0U8ztV3tcx95kcjJ49ezbe2yUBbMxucPI8Uj8T831oye/2pJJtkO2T36s5CYgkoEyObZTnirm/5TMg5MSVpVtYmz67Kd2Gl5IPAwtKl+SPrqkrzevOasV2sJPcTDnkEhjE9qWdkFa1cpAtB2eySOtSySWW2obEHKzFR2z7TF6PBBXyvLGRM9dJIZMQyllqOdiVs5xz58616B8lOQCVs5tyQCg52RJkyAG+eS2H5CrLQZQUHcoZfmknKfnPM2bMUDnRliT5/XLG+/fff1cHVJaeqEtGJsaPH6/OGEoALcXhErSZv++k5kbOtssog7xeyUeXWiAZ6ZBcbEt8/hJL6jRiNgWISd4ncqAv78m4AviYkvI7luYCUpsljyE1Kbly5VJnw+fPnx+vomrTd4KcqZeALjamAEcOhKW5gLw+qZWSkQ55Thk1SA+tpuM6cfCmYDI1tl1OrKR8v0lQ/ba2JzHPZfrsxlUrR2kfAwuiGIXRcjZQ0m/MRy2kk5Hp+qSSbh9CDsYtGQCYDuRkTgvz55HUnphnwExMr0cOVGIegMll8Xm98jyyz+SMaXL9QZeDWkkTkoM9OdiWok5LkTQyKaA1pUNJe0gpdI9JRhAkAJFFUifkIFIO+i0dWEjgJvMvyCiQHCjK88ZMzUtqYCGPKwekMpoQGBgYa2GoHBxLipwscrZcikklIIlPYJGcZMRFAgbZP5KCF5M0EZADeikkjau1dFze9DuO6wBL9qWkEkp3LQm+TWQ7YortMST4lu8bOVCLz3eCdJKT36MskmYko2vyu5H3bXK3u4752ZegSEaG4ppPwvw7xnzES7ZbuoQl5jswMQfesh0y6iHba35yxJLf7Ukl2yDbJ6PA5gXWUtgvI1rm2yjfgzGba8g+NX3/x3dfybxIErSZ387UItc074UpfUmez7zFemwjKG96PvmdS1CR1BNOlHqxxoLIjKR8yB93U5tREzlzKV+YljiokpxpSXeSzjExO6KY0iLedHY0tvuZ8nNNKQeSIiEHK9LpI+bs06azfxKMSIAjZ2XN0zckfUQ6AEmtxZvI2W3ZZ6ZUInNSL2D+xy+x7WaFdJGROQ+kU5eknFiK/KGU34mMVMjcEZJCEXPyMhkxiZkHLsHa67owJYWc7ZZWpRKsSW66efehpJIDFhnhkkBKFgkgzDvqyO8y5u9H3iOSTpRcrzchZGRO3r/STjQmGVmRbmXSAefTTz9N0OPG53dsmrMl5gGdnGWX7wfzM7iSRhLbDNvyGLHdXzolSYAiJwJe950Qczvl/SqpSLJPYvteSE7yOZGDdEkji9nNy/QdI4GDbKOMwJmPOsjIo7zP4vMdE1Ncv4c3fbdLzYB5PZV8P0nrXvldS/ellCbbKMy7gAn5zhPm+0qCun379kW7naTRxRxFeNO+kjoq8y59cqJBTmxIoGgaUTedpDJ/PlPb5Pi8v83J5/ZNE4NS2sYRCyIzUjwp6UTSIk8ODCQnXdIiJD1CDvBMX7BJIUGF5NBL/racBZazxXL2Rib3kuI8OZiMGdjEPHsvX85yltKUHiGFnfLHQM64ynaankcCIjnbKn3fJQ9bzjydOXNGFbbKHwU5gJXHkzO08odVzvDK2THJ7ZezVZ9//vkbX4/cT874SwAj6UQS0Mjjylk3qYmQxzJNgiZ/wOS55ExufAu4TeQAxnSmWtrzymuVs+mWIGd+JQ1FUkokyIg58aEcuEmrSqk3kOeVonc58Je2oPEhwWDM4E7I7y+uHH5pjyrvB9m/MreBnJ2XdBtLvV5JnZGz21JrYX4GV0brJM9dfmfy/peDLhmRkmJhS44UJZa0QZZ0KNkm81E2eW9J2pzk0UvwJHMrmJPmAjJCE5f4/I7lOiEjSPI+kYBAPr9ywCcHf1KXIZ8zGeGROhMJTGLWBshjyLbL7SVYk5ooaTwgBebS6EDWJZVLtkdG6eSzLbeXdSGfLzngk+8JeT1yAkC+L2QbzEdZJdCR945pzoK4yFlnU+Aqr1mYAnc5Qy7fU3GR1yfflXJSQVLn5DtJRmzkvSKvTb4T5LtNRlJklEz2j7yHZfRCPmvyvZSYSR7le1g+o3JCRF6zHMzKfpN9GRepP5OTE/K9I9+f8v0mv19pZy0H8m+aCfpt7E/5vEkqnAQIcnAuz3f06FH1XS1BnPlM8/K9Lm2hJSCV97Z8r8uIWcwUIwkQ5H0q3/Py3Sm/H/ncmGripJ5CvgPkdybvJ5k3Rf4GmI+2yXtOar/kdjIiKI8ntzP93Yr5/pa/b/Ka5f0hz2P6nMrnQj4P0nqY0rGUbktFZEmm9nfSjvB1pB2ftH2MjbSD/fzzz1XrT2nNWbRoUeOkSZOi2icmtd2sibT/a9q0qWolKS0xCxcubOzTp4/x+PHjr932gwcPqjaG0spS7ivbmD9/fnXfa9euvXL79evXG2vUqKHafbq6uhqrVKliXLp0abTb/PPPP6q9oIODg2rx+s4770S1Bo3PPhOzZs0yent7q+eRFrFlypQxDh8+3Hjv3r1Et5uNbf8GBQUZq1WrptpcmlrZStvExLSDNZFWk6YWvubtKE2k/bDsN2nLKrcrXry4cfz48apdZVLazZq/T+J6DdKGuEiRIuo9YmoZKfdNTDtY8zahpm04cOBAtOukveewYcNU+1X5PcrvXNanTZtmsc+fab9IK8zE+OSTT9Q+MWdqiRnXElu7zYT+jvV6vXHw4MFGd3d31X7a/E/o3Llz1XeFfIbkvrIvYmvTefnyZWOdOnWi3m/mrUGlBaf8XvPly6c+19JmtWHDhuqzZd4iWe4vLXflueR7Q35f0ibU/DtMHlta1b7J696jr2sbak5axpq+P7JkyaLut3379mi3kfaysl/kdeXIkcP4wQcfGJ89exavz7zsI/l8mFu3bp1qCy3tmc2/U+J6DNP+7du3rzF79uyqBa58R8X2XRTzs/k292d4eLjx22+/NRYsWFDtK3kvSJtorVYb7XbSuvfLL79Ur8XJyUn9LZF26DHbzYrZs2cbCxUqZLSxsYn2WTB952zdutVYtmzZqPdubJ/LEydOqNblst/k783kyZNj/dsm31fymPLdEfM1T58+XW2reWtfSn+s5J+UDm6IiIjiS1qPSq2FpOxJoT29mhYpMyrLWWxJe6OkSa/7U0ZtpAOaNAJ4G2QCSRkVlJF0Sr9YY0FERGmKFAFLWkZ85qfIiCSlSlK00tNBcEri/kw66WAm6bGxNcag9IUjFkREREQZzNsesaCMgSMWRERERESUZByxICIiIiKiJOOIBRERERERJRkDCyIiIiIiSjJOkPcaMpOozEopE+e8aZp6IiIiIqL0RmamkMlTZeJL8wlVY8PA4jUkqMiXL5+lfz9ERERERGmKv78/8ubN+9rbMLB4DRmpMO1IV1dXy/52iIiIiFKATqfDL7/8ota/+OIL2Nvb8/dAcQoMDFQn2k3Hxa/DwOI1TOlPElQwsCAiIqL0ElhoNBq1Lsc3DCwoPuJTFsDibSIiIiIiSjIGFkRERERElGQMLIiIiIiIKMkYWBARERERUZKxeJuIiIgoA7GxsUHdunWj1okshYEFERERUQYiwUS9evVSejMoHWIqFBERERERJRlHLIiIiIgyEKPRiEePHql1d3f3eM1PQBQfHLEgIiIiykDCw8Mxffp0tcg6kaUwsCAiIiIioiRjYEFEREREREnGwIKIiIiIiJKMgQURERERESUZAwsiIiIiIkqfgcXEiRNRuXJlZMqUCR4eHmjXrh18fX2j3Uar1eKjjz5CtmzZ4OLigo4dO+LBgwfRbnP79m20bNkSTk5O6nGGDRsGvV6PtCIsJBgX9u7Ew5vXU3pTiIiIiIjSXmCxd+9eFTQcPnwY27dvV63QmjRpguDg4KjbfP7559iwYQNWrFihbn/v3j106NAh6nqDwaCCCp1Oh0OHDmHhwoVYsGABRo8ejbRi71/zsGXarzi7Y3NKbwoRERGlo5m3q1evrhZZJ7IUK6PMkpLKySQuMuIgAUSdOnXw4sULNaHLkiVL0KlTJ3Wby5cvo0SJEvDx8UG1atWwefNmtGrVSgUcOXLkULeZMWMGvvzyS/V49vb2b3zewMBAZM6cWT2fq6sr3rabZ09h1fhRcHTNjPdnLII1P/xERERE9BYl5Hg4VY5YxCQvRGTNmlX9f+LECTWK0ahRo6jbFC9eHPnz51eBhZD/y5QpExVUiKZNm6qdc+HChVifJywsTF1vvqSk/KXKQpPJFaGBL+B/4VyKbgsRERERUZoOLCIiIvDZZ5+hZs2aKF26tLosICBAjTi4ublFu60EEXKd6TbmQYXpetN1cdV2SERmWvLly4eUJCMUXlVqqHVfn30pui1ERESUPkiyyvPnz9WSBhJXKA1J9YGF1FqcP38ey5YtS/bnGjFihBodMS3+/v5IacVq1Fb/+x31gSENFZ4TERFR6iRZH1OmTFGLrBNliMDi448/xsaNG7F7927kzZs36vKcOXOqomyJtM1JVyi5znSbmF2iTD+bbhOTg4ODyh0zX1Ja3hKl4ZTZDdqgl/A/fyalN4eIiIiIKO0EFjIsJ0HFmjVrsGvXLhQsWDDa9d7e3rCzs8POnTujLpN2tNJeVjocCPn/3LlzePjwYdRtpMOUBAslS5ZEWiHpUEWr1lTrl332p/TmEBERERGlncBC0p/++usv1fVJ5rKQmghZQkND1fVS/9C/f38MGTJEjWZIMXffvn1VMCEdoYS0p5UAomfPnjhz5gy2bt2Kb775Rj22jEykJcWq11L/Xz0m6VAcsiQiIiKi1CdVBhbTp09XNQ716tVDrly5opZ//vkn6ja//vqraicrE+NJC1pJb1q9enXU9dKXWdKoTL2a3333XfTq1Qvfffcd0po8xUvC2S0LwoKDcevs6ZTeHCIiIiKiV9giFYpPhwKNRoOpU6eqJS6enp74999/kdZZW9vAq1otnNqyAb4++1GoYuWU3iQiIiIiotQ/YkGv8opKhzoMvU7HXUREREREqQoDizQij1cJuGTNBl1oiJqRm4iIiCgxrK2tUalSJbXIOpGl8N2URlhZW6t0KHGF3aGIiIgokWxtbdGyZUu1yDqRpTCwSIvdoY4fQbguLKU3h4iIiIgoCgOLNCRX0eLIlN0d4dpQ3Dx9IqU3h4iIiNIgaZITHByslvg0zCGKLwYWaYiVlVVUOpTvIU6WR0RERAkXHh6On3/+WS2yTmQpDCzSmOLVa6v/r508ivAwbUpvDhERERGRwsAijclRuCgye+SAPiwM108eT+nNISIiIiJSGFikxXSo/0Yt2B2KiIiIiFILBhZpULH/6iyunzoOnTY0pTeHiIiIiIiBRVrkUbAw3HLkgl4Xhusnjqb05hARERERMbBIq+lQxWpEpkP5+hxI6c0hIiIiImJgkVaZ2s7eOH0cYSEhKb05RERElEZYW1ujXLlyapF1IkvhPO5plLtnQWTJnRfP7t3B9RNHUKJ2/ZTeJCIiIkoDbG1t0a5du5TeDEqHGKam5XSo6pGjFpd9OFkeEREREaUsBhap2LNHAZjXbyyunDz62u5Qt86chDY46C1vHREREaVFRqMROp1OLbJOZCkMLFKxNUMXIdS+DvZPOYtH9/xfuT57/gLIljc/DHo9rh0/kiLbSERERGlLeHg4Jk6cqBZZJ7IUBhapmGcTD9joQ6B1LIT1X61CaHBwnEXcvkyHIiIiIqIUxMAiFWv4Th+4F78Iq4hwaJ3KYsnHk6GPcWah2H+zcN86ewraIKZDEREREVHKYGCRynUcOhyu2Q4AxghoHWpi8eCx0a7PljefSomKMBjgd+xQim0nEREREWVsDCzSgHcnfg8nm11qPQQNsWjY17EWcV/hZHlERERElEIYWKQRfadNgEYX2VY26EVdrPhhYtR1XqZ0qHOnERL4IsW2kYiIiIgyLgYWaUjPaV9CE3oKRmtbPLlaFlvnz1aXZ82dB+4FCsEYEYGrxw6n9GYSERERUQbEwCINsddo0O7HbtCE+MFg64hb+7Pi2PZ/1XVeVWuq//2Oss6CiIiI4mZtbY2SJUuqRdaJLIXvpjQmW848qDusChy09xFunwVn/n6K6xfOwqtaZGBx+9xpdociIiKiONna2qJz585qkXUiS2FgkQYVKVMBZXtkhp3uOcI0ubH7Jx9Y2dmqyfKkO9S1E5wsj4iIiIjeLgYWaVSVJq2Qv9Zj2OhDoXUsitVD/0Ih76rqOqZDEREREdHbxsAiDWvWbxCyFT4Dqwg9tI4VcXnbBXX5zTMnoQsNSenNIyIiolRIp9Ph22+/VYusE1kKA4s0rvOIkcjkulet6xyawtraHobwcFw/dTylN42IiIiIMhAGFulAz5/Hw8lqB6ysrGBlX05d5neE3aGIiIiI6O1hYJFO9J0+AY76PbCxK6p+vnLkEMLDtCm9WURERESUQTCwSEd6TR8FR/1VwNoVMEZg8ZfDU3qTiIiIiCiDYGCRjtja2aHXjOFwsMqkfg58bIcVEyak9GYRERERUQbAwCIdzs7dfOgAtW4Iv4FH18ti7W+TU3qziIiIiCidY2CRDhUq7w2nzG7SJwp6YwACznth85wZKb1ZRERElApYW1ujaNGiapF1IkvhuykdsrK2hle1mpHroSdgsHXCbZ9c2LP875TeNCIiIkphtra26NGjh1pknchSGFikU0WrRAYWsH8C+5Cb0NtlwpUtGvhsWpvSm0ZERERE6RADi3Qqb4lScMzkqmbgLt7aHvba+wi3z4LzK8NwZv+ulN48IiIiIkpnGFikU9Y2NihSpbpaD3vxAt693GEf9hg6B3ccm3sPV04eTelNJCIiohSg0+kwYcIEtcg6kaUwsEjHvKrUUP/7HfVB+XoNUbKdDex0zxGmyY19Uy7g+oWzKb2JRERElALCw8PVQmRJDCzSsXyly8LB2RkhL57j3uVLqNm2Iwo3DIRteBDCHD2x66cTuHruVEpvJhERERGlAwws0jEbWzsU9q6q1q8cPaj+b/hOHxSo9QC24S9VcLHn5zO4euZ4Cm8pEREREaV1DCzSOVPbWUmHMkZEqPWmfQeiYJ1H/wUX+bFn8nkGF0RERESUJAws0jnPMhVgp3FE0JPHCLjmF3V5k94DUKjuY7Pg4gILuomIiIgo0RhYpHO29vYoVLGyWr9yJDIdyqRxr/4oXP8ZbMMDEeaYD/umXMblE0dSaEuJiIiIKC1jYJEBeFU1dYc6BKPRGO26Ru/2QZEGz2Grk+AiLw78cQWX/qvHICIiovTHysoKnp6eapF1IkvhPO4ZQMHylWBr74AXDwLw8OZ15ChYONr1UtBtZbMYftuMCNPkwcFp12EwGFG6eq0U22YiIiJKHnZ2dujTpw93L1kcRywyADuNBgXLe6t1vyOHYr1Ng249UaxZyH/zXOTB4Zk3cd7nwFveUiIiIiJKqxhYZBBFTelQMeoszNXr8g6KtdBGTaJ3eOYtnD209y1uJRERERGlVQwsMohCFavAxtYWT+/dwZM7t+O8Xd1OPVC8VRjsdM8QpsmFI7Pu4OzB3W91W4mIiCj56HQ6TJo0SS2yTmQpDCwyCAcnJ3iWrRBrd6iY6nTojhKtDCq40ElwMScAhzevf0tbSkRERMktJCRELUSWxMAiAyla5b90qMNv7vpUu0MXlGwbAbuwJ9A55MDZFeHYtXTRW9hKIiIiIkqLGFhkIIUrV4O1jQ0e3b6JZwH3XntbbVAQChQriVyVbsMqyAdh1qHw3WaHDdP/eGvbS0RERERpB9vNZiCOLpmQr1RZ3Dp7SnWHqtK2E8LDtHh6944KNh7738IT/1vq/6CnT6LfOdwHkoV5ZY81phzai1xFveCa3QOu7u7//e8Bt5y54ZrdPaVeHhERERGlIAYWGTAdSgKL4xtW49yurXj+IACIMWmeSabs7siWJx/0Oh2e3r+LkOfPAURAr9PC/8LZWO/TYvBQlKhVL5lfBRERERGlNqk2sNi3b5/qVnDixAncv38fa9asQbt27aKul4ldFi5cGO0+TZs2xZYtW6J+fvr0KQYPHowNGzbA2toaHTt2xJQpU+Di4oKMqkjlatg1fwZCXwaqRThmckX2/AWQPZ9n5JLfE9ny5oeDk3O0+z66extrv1oCrX1uGCNewM72NApXLo/ARw9Vt6mXjx/h7M4tDCyIiIiIMqBUG1gEBwejXLly6NevHzp06BDrbZo1a4b58+dH/ezg4BDt+nfeeUcFJdu3b0d4eDj69u2LQYMGYcmSJcionN2yoMNX3+LJ3dsqeHDPXwBOmd3idV/3PPnRY+oH+OeT6QjVVEEESuHu6R3oO20CAh8/xOyP+uHOpQsqjcola7Zkfy1ERESUcFZWVsidO3fUOpGlWBmNceTBpCLypo9txOL58+dYu3ZtrPe5dOkSSpYsiWPHjqFSpUrqMhnNaNGiBe7cuRP1gXqdwMBAZM6cGS9evICrq6sFX1HaptNqsfjDH6G1r61+dtTvQa/po7D8uxG4f+Uy6vcZhIrN26T0ZhIRERFREiXkeDhNd4Xas2cPPDw8UKxYMXzwwQd48uT/Bcc+Pj5wc3OLCipEo0aNVErUkSNHYn28sLAwtfPMF3qVvUaD3jNHwjFip/o51LYeFg38EUUqV1c/+x7az91GRERElMGk2cBC0qAWLVqEnTt34scff8TevXvRvHlzGAwGdX1AQIAKOszZ2toia9as6rrYTJw4UUVkpiVfvnxv5bWkRbZ2dug3azyc7XYAxgiEamrgzMpLMryEe1cuIfDxo5TeRCIiIiJ6i9JsYNGtWze0adMGZcqUUSlSGzduVGlPMoqRWCNGjFDDPKbF39/fotucHvX5YwJcM++BVYQBOqcasEFkvcYVH45aEBERpUZSd/rbb7+pRdaJkNEDi5gKFSqE7Nmz4+rVq+rnnDlz4uHDh9Fuo9frVacouS42UvwtuWPmC71Zz5/GIWteH1gbwmClqaAuO7l1I3cdERFRKiTltaaTqGmg1JbSkHQTWEhBttRY5MqVS/1cvXp1Vdwt7WpNdu3ahYiICFStWjUFtzR96jZ6NPJXug57KwnarPDy0UOsmfJzSm8WEREREWX0wCIoKAinT59Wi7hx44Zav337trpu2LBhOHz4MG7evKnqLNq2bYsiRYqouSxEiRIlVB3GwIEDcfToURw8eBAff/yxSqGKT0coSriW730E7+6usLGOnH3b/0QYlo39lruSiIiIKANItYHF8ePHUaFCBbWIIUOGqPXRo0fDxsYGZ8+eVTUWXl5e6N+/P7y9vbF///5oc1n8/fffKF68OBo2bKjazNaqVQuzZs1KwVeV/lVp0gplW9ZU63r9NTy5XxMLBo9M6c0iIiIioow6QV69evVem/e3devWNz6GdIDKyJPhpZRqbTrh9KZ1MBoeIiLiBYLDG2Fev7F4d9pXqlUtEREREaU/qXbEgtIuJ9fM8CxTXq1bh21W/4fa18Hi96bi0T122iIiIiJKjxhYULIoVj1yVm4XDwe4uuyEVUQ4tI4VsH7EVlw6epB7nYiIKIVYWVnB3d1dLbJOZClWRvYZs8gU5hSdNigI0we9iwiDHn1+mYY9S5fhkW8JGGydYR/2CF4tDajbqQd3GxEREVE6OR7miAUlC42LCwqUiyy89/XZj47DhsOr0XPYhz2GzsEdlza7YMXECdz7REREROkEAwtK9nQo30P7VSF+g249UX1QbjiE3oDBzgUPb1bG/A9HQs9ZP4mIiIjSPAYWlGwKV6oGGzs7PL13B49v31SXla5eC+1/bAZN6DHAygYhEY2wcMAveBJwl78JIiKityA8PBzTpk1Ti6wTWQoDC0o2Dk5OKFjeOyodyiRbzjzoPWcInKx3AEYDtI5VsObLrTi5axt/G0RERMlMsggePXqkFpbakiUxsKC3mg5lYmtnh77TJsCjwDHYhAchzLEAjv0VjLW/TeZvhIiIiCgNYmBByaqQdxXY2jvg+YP7eHjj2ivXdx4xEmXahMNBexd6+8y4d7E0Fn7KmbqJiIiI0hoGFpSs7DWOKFSx8ivpUOZqtu2IFmOqQRNyGkZrWwSFNcK8PuPw8vlz/naIiIiIMkJgIQU//v7+8PX1xdOnTy23VZSuFKteS/3v63MgzlzO3AWLovfcwXCM2AkYIxCqqYHln/yD8z4H3vLWEhEREdFbCSxevnyJ6dOno27dumqSjAIFCqBEiRJq9kZPT08MHDgQx44dS9TGUPpUsEIl2DloEPjoAQKuXonzdlJ30W/WeGTNeRA2+lBonYri0OwAbJo17a1uLxERERElc2AxefJkFUjMnz8fjRo1wtq1a3H69GlcuXIFPj4+GDNmDPR6PZo0aYJmzZrBz88vEZtE6Y0EFYUrVVXrvj773nj77t+OgVfDJ7APe4Bw+6y4dbwwFnzK+S6IiIgswcrKSs2kLIusE1mKlTEBfca6d++Ob775BqVKlXrt7cLCwlTwYW9vj379+iEjTGFOr3f12GGs+3kcXLJmw6Cp82Fl/eaY9vqFs9j94x5onUqrnzWhJ9B8bDuVNkVEREREqet4OEGBRUbDwMJyZHbt6QPfgS40BF2//RF5i78+ODW55+eLzZPmQmvVWE2oZ68NgGf9EDTpPcCCW0dEREREST0eTnTxdrt27XD27NnE3p0yGKmfKFK5mlq/8oaC7AiDQXWQWvL1F1j6zRfQRtxBdk8f2OmeQafJiWsH8mLh52xJS0RERJSaJDqwaNGiBTp27IjOnTvj4sWLUZffvn37jalSlLEny7ty+AAiIgyvXB8WEozjG9dg7qcDsfG3H3H/qq+6XPsyEC/v30KVnpmgCbmICBt7BIU2wtzePyDg1qtzYxAREdHru3rOnj1bLbJOlOKBhbe3N0qWLIk1a9agbNmyqFGjhirorlatGhwdHS22gZR+eJYtD42zC4KfP8PdSxeiLn/x8AH2LJqNWR/2wd7FcxH46CEcM7miWsfu6DXpT7jlyKVuc37bv+j6+ztRLWm1jlWwccxh7Px7QYq+LiIiorREsuDv3bunFmbEkyXZJvaOPXv2VG1mly5dCltbW1y+fBm//PILChYsiG3btll0Iyl9sLG1Q5Eq1XF+93aV6mRjZ48Tm9bC78ghGI0R6jZZ8+SDd8u2KFG7PuzsHdRlHUZ+i6XfDMWD61exbfpv6D3tW6z9dTKeXCqCME0uXNkThrunvkavn8en8CskIiIiyrgSXbzt5OSEc+fOoXDhwlGXySR5PXr0QN68eTFnzhxLbmeKYPG25d08cxKrJoyGlZV1VDAhPMtWgHfLdihQtkKsHaPu+/li+XcjodeFoXT9xmjy3ic457MPx2ZcgNapuLqNRnsEbSa8C/fc+ZJhy4mIiNIHnU6HiRMnqvURI0aoLp5EKVq8XbVqVaxbty7aZVmzZsWUKVPUKAZRbPKXLgenzG4qqLCxtVVBgqQ7dfr6exQs7x1nG9pcRYuh1WfDVUAiIx4+K5egbI266DG9NxwNuyNTozRVsXbkPvw7ezp3PhEREdFblujA4scff1RzWvTt2xdHjhxR0a8UAK1cuRLOzs6W3UpKN6xtbNB26Deo32cQBk6dj6bvfwr3/AXidd/C3lXRsP8Hat1n5VKc3bkVjs7O6Df7e3gUOApb3QvoNLlw81hhzH//a4QGByfzqyEiIiKiJAcWVapUwa5du9Ss29WrV1cF2xJQjBo1Cp9++mliH5YygNxexVGxeRs4u2VJ8H3LNW6Oqu27qvUdc6bi+qljar3ziJGo+1FuaELOwmhtixA0xJIP/8Lhzestvv1ERERElEwT5N29exeXLl1SuVfly5ePVneRlrHGInWSt+zW6b/hwt6dsHVwQNfRE5GziFfURHx/Dx+DkKBaiLDRwEYfCme3Q+jJwm4iIiJFskwkdV3IyWDWWNDrcOZtC2FgkXoZ9Hqs+fFb3Dp7Co6umdHj+5/hljNX1PUH1q2A7+oX0DoWUj9rQk+g7tBGKFKmQgpuNREREVHa8laKt4lSkhR+txkyAh4FCiM08AVWTRyNkMAXUdfXatsZ3f7oDEfDLsBogNbRG7t+vY6VP0Z2wbC0Z/fvYs4nA7Br/sxkeXwiIiKi1I6BBaVZ9o5OaP/VGLi6e+B5wH2s/fE7hIdpo653ds2MfrPHIV/pC7APe4hw+yx4cL0y5vUbi2ePAiy2HcaICGyd8TtePAjAqS0bEHD1isUe+9iG1Vg5fhRCXwZa7DGJiIiIkgMDC0rTXLJkRYcR30Ljkgn3r/pi39/zX7lNm8GfoeVob2i0PoCVNULt62DVsG3YPGeGRbbh7M4tuHv5/zOJ71+6wCIzmT66dQP7/16g0r1ObdmY5McjIiIS0sVzwYIFapF1IkthYEFpXrY8+dBi8FC1fmHPTuhCQ165Te6CRdF/wdfI6rEPtuEvEabJixtHC2PegNFJGr0IfPwoKpip1LqDStG6ff6sCgaSQgKTPYvmRE0ieGb7v6ownYiIKKnkb8ytW7fUYokTYUSJCiwqV64MGxubeC/W1tbqf6LkVqBcRWTJnVelQl0+uC/O23X/biyq93eDJuQ0jNY2CLWth1XDdmL9H78l+Dnly3jn3GnQhYYil1dx1O7RG+WbtlLX7VuyQKVIJdb1k0dx+/wZ2NjZqQkFQ148xxWf/Yl+PCIiIqJUFVjs3LkTDx8+jPfy6NEj9T9RcrOyskLZhk3V+pkdm197W5mxu/+iIXDLshu24YEI0+SC//nSqvbiof+teD/n5YN7cf3kMTVK0fS9T2BtbYOq7buo2o9HN6/jciIDAYM+HHsXz1Xr3i3aqjk/xMnN63lmiYiIiNJHYCEtprJly5bghehtKFmngTrIf3jjGh5cv/rG278z8XvU/yg3HEOPR9VerBvtg1WTfnrjfaUD1a4Fs9R6tQ7dkC1vfrXumMkVldt0VOsH/1msgoSEOr31Xzy7f0+NVFRp1wVlGjaFrZ29ek33fC8l+PGIiIiI3gbWWFC64eSaGUWr1ozXqIWJV8Uq6LdwOLLl3A873TPoHDwQcK0S5vUZB38/3zjvt3vBLGhfBsI9fwFUbhsZSJjIKIPMKi5dos7u2JKg1yDdn3xWLVHrNbv2hIOTk3pdxWvVixq1ICIiIkqNGFhQulK2UTP1/+UDe2Mt4o5Lt7Fj0GR4MThK5yg5wNfUwOYJ5/DPuHGv3PbaiaMqDcrKyhpN3v8UNrZ20a6302hQvVN3tX549T8J2o5DK5YgLDgY7p4FUbp+o6jLK7aITIfyO3pIFYwTERERpTYMLChdyVuidLyKuGNToHhp9FvwNTw8D8M+7DHCHbLi8Z0amNfnB1w+cUTdJiwkBDvmTFXr3q3aIWfhorE+Vun6TZAlV25VdH1849p4Pf+TO7dV9ydRr9dAVbNhIiMj+UqVVQXhp7dtStDrIiIiisnOzk4tRJbEwIIybBF3XDqPGInWY73hGHZAZr9DqKYK9kx7gIWfjlQtYIOePoFbzlyo0blHnI8htR41u/ZS68c3rlEBxpvsWTxXBQ5FKldD/tJlX7neVMR9bseWaBMBEhERJYS9vT1GjhypFlknSjWBxfTp0y2zJUQpVMQdm5yehdFv/mjkKXkODto7MNi5IDCoGM7v3qaubzJoMOwcNK99DK9qNZGjUFGEa0NVStTr3Dh1HDdPn4C1jS3qvNsv1tsU8q6MzB45oA0OwqX9exL1uoiIiIhSbWAxe/Zsy2wJUQoWccel3aefo/vvbeFotQ3hIZFBhY19aWz/eS1uXj7/xtETmdtCbcf2zXj+IPaJ+Ax6vRqtEBWat0aWnLljvZ2kRlVo1lqts/UsERERpTZMhaJ0KbFF3LFxds2MIrWKwRjxAlawh61TXYQ61MLWn/ywaNjXr50R27NMeXiWrYAIgx6Hlv8V623O7tyCp3f9Vavaah26vnZbStVrpEZKpB7j9rkzSXpdRESUMen1eixZskQtsk6UagKLCxcuoHr16hgwYAB+/fVXbN++Hffu3bPM1hGlQBF3TJJOJXUSou3wL5HD8xTstQ+gt8+Mly8bYmH/Odi1bHGc96/do4/6/9KBPXh483q067RBQTi0/G+1XqPLu9A4u7x2W+T6UvUaqvWTm9cl6XUREVHGFBERAT8/P7XIOlGqCSyKFSuGWbNmoVGjRnjy5AmmTZuGunXrwt3dHbVr17bMVhKlQBG3KU1p64wpqqi6WI06KOxdFZ1HjkSnSfXhGLET1gYdtE7FcHlXDszrPzrWmbtzFCys7iv2L10Y7brDq5dCG/RSTbBn2t43MaVDXT91HM8CGMQTERFROgksbGxsUKZMGXTr1g3jxo3DmjVrVATs7++P33//3TJbSZRCRdzHN6zGo1s3oMnkigZ9BkVdnsU9J/rNGo/ybUKhCTkPo7UtQu3qYe2Y41g89GvotNG7NtXs+i6sbWxUgfbt82fVZU/v3cWpLRvVer1eA9T18ZE1d14ULO8NGI04tWVDol4XERERUaoLLN5///1YL9doNKhQoUJSH54oxYq4JQg4tCIyTalB74Fwyuz2ym2qt26P3nM/QBb3vbALe4Jw+ywIDGqIRe8twdopv0bdTgqyTXUf+5cugNFoxN6/5iLCYEDBCpVQoFzFBG2bqfXshT071NwaRERERGk+sHjvvfcssyVEqaiIW0Y41k+eqA78S9Suj+K16sV5W1s7O/T4/lu0H1/tv/QoLcIcC+DupXKY2/tnHN68Xt2uWoduqvA64OoV7Jo/A9dPHIWVtTXq9uyf4NflWa6iGrnQhYbiwp7tCb4/ERERkaWxKxSla4kp4pa2sKt/GKvuk790OTR9/xNVs/Em7rnzqfSomn0codEeUZPraR0r4uQaB8wbMBpPHz1Us3WL01sjZ88u36QlsuXJl+DXJdtT4b9RC0mniogwJPgxiIiIiFJtYBEQEPDan4lSexF3SOALrJ44Ws2U7e5ZEG2++Bo2tnYJes6yNeuj/4IRKFT5KjQhV2C0tkOobT38O/Eirhw8B41LpqgOT9U7dU/kKwNK1WkAB2dnPH9wX02wR0RERJRuAov+/fu/9mei1FzELSMUa3/8Ds/u34Oruwc6fDUWDk5OiX7e5gPeR++5/ZEl+17Yhz1U7WlDwprCKqy4ur5urwFq7orEstNoUKZBZNB08t/IdCsiIqI3sbe3x5gxY9Qi60SpMrDYtGnTa38mSq1F3FJLsXHKT7h/1VeNKHQY8S1csmZL8nOr+otx36LLz43gaL0DNvoQGF1qw8HtcxyZdxFb5s1K0uNXaNoKVlbWuH3+DB7fvpnk7SUiIiJKLNZYEDJ6Ebd0aNoxd5oqpra1s0e7YaMSVffwOpmzZUe/aRNQ7wN3aMIOwgoR0DqVxrUjhTC39yTsX708UY8rIytFKldT6yfZepaIiIhSEAMLQkYv4j68ehnO7dyqzvy3+HQY8hQvmWzbUdy7KvrPH4XSjR9BE3oSsLKG1tEb57Zkwby+43B6z44EP6ap9eylfbsR+jIwGbaaiIjSE71ejxUrVqhF1onSRGBx5MiR5Hx4oiQXcZ/bvQ2Hlv83V0Xf91C0cvW3slfrduqB/guHoqC3LzQhF2C0tkGoQw34/K3HvP5jcPVM/Iux85QoBfcChaAP1+Hszq3Jut1ERJT2RURE4OLFi2qRdaI0EVh07tw5OR+eKElF3NdPHcP2WX+q66q064zyTVu+9T3aYuAH6L9oMHJ5nYAm9BoibOwRalcXO/4IwLxBX+PeDb94BU2mUYtj61aq4MLIPxRERET0ltkm9QG6dOkS6+WSt/706dOkPjyRxYu4Lx/ci71/zVOF2nIALgFHrW69UnRPdxgyDPrwcKwcPwFBNwsiTJMXoWiIDeMuwN5xPtp88xGy5cwT5/2L16yLM9s2IeCaH7bP+gNnd2xBw37vI1fRYm/1dRAREVHGZWWUCCAJsmbNisWLF8PFxSXa5fKwXbt2xYMHD5BWBQYGInPmzHjx4gVcXRPfFpRSD/+L57D82xFRP3uWrYD2X45RIxmphU6rxbJR3yPscVnoHNzVZXa657DTHEfLEe/BI59nrPcz6PU4tWUDfFYuUTNyi1J1G6F2j95wdsvyVl8DERGlXjqdDhMnTlTrI0aMYMtZstjxcJKPpurVq4dMmTKhTp06r1xXtmzZpD48UbIUcT+7dwceBQqjzZARqSqoEPYaDXpNGo+Xz59j1agfoQv2Rrh9VoRHNMKaMadg7zATzb/qj5yehaPdT15HpVbtUaJWPexfsgAX9u7Ehb074Hf0EGp0fkeleqW210pERETpR6JrLPz9/dX/q1evjjWoENu3b0/0hu3btw+tW7dG7ty5VQ752rVrXxkRGT16NHLlygVHR0c0atQIfn7R89ElFeudd95R0ZWbm5uasC8oKCjR20Rpn7yXmr73Cco3bYUOI8bC3jHxE+Alt0xubujzx0R0nVQPLg47YB/2GHp7V4QYG2Hdd+cx7/2v4e/n+8r9ZHSi2Yefo/v3k5CjUBHVXnfPotlY/OUnar4LIiIiolQVWBQvXlwd2IeERJ8TwFKCg4NRrlw5TJ06Ndbrf/rpJ/z++++YMWOG6j7l7OyMpk2bQqvVRt1GgooLFy6oAGfjxo0qWBk0aFCybC+lHdJOVuoP0kp6kMyB0XvKBHT7tQlcHCXAeAS9XSZVg7HpB1/MG/QNbl4+/8r9cnuVQI/xv6DxoI/VDN9P7tzGiu+/xobJExH4+GGKvBYiIiJKvxJdY3Ho0CF8/vnnuHPnDsaPH48+ffogOc8yr1mzBu3atVM/yybLSMYXX3yBoUOHqssk7ytHjhxYsGABunXrhkuXLqFkyZI4duwYKlWqpG6zZcsWtGjRQm2z3P9NWGNBqVFw4Aus+u4nhD0uBZ0mp7pMZvS2tzqCOh+3RZFyke93c9qgIBxc/hfObPsXRmMEbO0dVHvdMg2apMArICKilCTHUeHh4Wrdzs5OHWcRWeJ4ONEjFjVq1FAjBVL8M2rUKHh7e2P//v14G27cuIGAgACV/mQiL7hq1arw8fFRP8v/kv5kCiqE3N7a2prza1Ca5uyaGb1+Ho93p7aFa+ZdcNDeh8HWCaE29bH9z0eY1+9bHN68Ptp9NC4uapSm549TVJ2JXheGbTN/V3UYRESUsUggYW9vrxYGFZSq5rHo1asXfH190bJlSzRv3hydOnVSB/7JSYIKISMU5uRn03Xyv4eHR7TrbW1tVRcr021iCgsLU1GZ+UKUWjk6O6Pnj+PQa0ZHuGXZDYfQO4iwcUCofW2cWOuIub0nYfOcGdHu4+5ZEF3GTIya92Lr9Cm4fCj6TOREREREKTpBXpMmTTBgwACVsiQpSMOHD09zhdIy+iIjH6YlX758Kb1JRPHqIvXOxO/RZ053ZM97CJoQX8DKBlpHb1w/7oW5vaZi+fjxap4MIWen6vUeiLINm6m0qM1//oKrxw5zTxMRZRB6vV41xZFF1olSPLCQomnpsiQtZeUgvGHDhioV6v3338eUKVNw/PhxFWDI/5aWM2dkXnnMOTLkZ9N18v/Dh9ELVOXDI52iTLeJSXo5S/6YaTF1viJKC2zt7ND1m2/Qf9EHKFDxMhxDTwJGA7ROJfDIvzoWDFiCRcO+Vm1sJbhoNOBDlKhdHxEGAzb+9gNunj6R0i+BiIjegoiICJw5c0Ytsk6U4oGFFGzLwbekQu3evRvPnz/HiRMnVBcn6by0a9cuFWQkR1F3wYIFVXCwc+f/88MlbUlqPqpXr65+lv9N22Qi2yQfIKnFiI2Dg4MqSjFfiNKiloM+RL+FQ1GxTTAcww7A2hCGMMd8ePmyIZZ+vgPzPxyJO9f80OyDz1C0ag01ud66n8erCQSJiIiIUmTm7deREQOZZ8JgMCT4vpJGdfXqVbVeoUIFTJ48GfXr11c1Evnz58ePP/6IH374AQsXLlSBhhSQnz17FhcvXoRGo1H3k5oPGcWQ0RXpftC3b19VzL1kyZJ4bQO7QlF6ceXkURyYsR7hhkrQ20UGzDb6UNgbTqBQi0J4dv0arp88BjuNIzp9/T1yexVP6U0mIqJkwpm3KVV0hbp9+3aCNkSKoWWUIDEkhUoCClnEkCFD1LrMnSGkhmPw4MFqdKRy5coqEJF2sqagQvz9999qvg1J05I2s7Vq1cKsWbMStT1EaZlXxSroN2scOo2vDGeZbE8bAIOtI0IdauHCztx4ct4NDs6uCNeGYvXEMXhwPTKoJyIiIkqWEQvpuiRzSUiRthzMx0aimeXLl6s6Czno/+STT5BWccSC0iudVos1P/+MoCuu0DqWBKysYTSGQx+4DIaIR7B3ckL37yYhez7PlN5UIiKyMI5YUHIdD9sm5IElzUhqKxo3bqxGBmTuCploTtafPXumrpeZritWrKhmxpZRAiJKnZ2kpNBb7F+9HFc2XkK4dSXAtQsiXq6ELuQBFg39DMUbNUKLgR+l9OYSERFReq2xCA0NxaZNm3DgwAHcunVL/Zw9e3aVqtS0aVOULl0a6QFHLCgjuXfDD9smz4cusAiCwvfBaHgMWLkgk10lOOcPRJthQ5DJzS2lN5OIiJKIIxaUXMfDyVq8ndYxsKCMSOa7WDlxPO5fvIQIYzCsrDPDPlNX2OkjYGd1EsXalEPNth1TejOJiCiR5NAvJCRErTs5OXH2bUo9gcXdu3fV/3ny5EF6w8CCMrKgp0+wcNjH0Aa9hLV1Vti5doeVlYO6ThNyCZo8AWg19FNkzpY9pTeViIiI0lpXKHMHDx5UbV6l9assUtj95ZdfqicnorTPJWs2vDvxVzi7ZUFExFMYtQvhEHxeXSeT7j1/Vh/Lhh/AvP5jVJ1Gco6gcGCViIgo9Ut0YPHee++hRIkSOHbsGHx9fTFp0iTs2LFDFW6bRjGIKG3L7JET7b8aC3tHR+i0QcjpHYoKLZ7DUb8HtroXak6MULu6OLstO+b2mo7FX36DJwGW+/xfPrgXU3p2wNkdmy32mEREGZ1er1e1srLIOpGlJDoVytHRUU0F7+XlFXWZPFSXLl3U+ooVK5DWMRWKKNKts6ex+ocxiDAY4N2yHer1GoCXz59jw8+/IvR2dmgdS6iWtcJGHwI7/Rlkq+CEVh9+DFs7u0TtRoM+HHM/HYSXjx+ptre9f57KXwcRkQWweJtSXSqUjFbIzNrmrKys8N1336mJ6ogo/fAsWx7NPvhMrZ/YtBbHN65RHaJ6jPsW/RcNhne7EDjqd8NO9xQGWydoNdVx91I5zB+4EvMGfY2j2zYm+Dkv7Nmpggrx2P8Wnt7jSCgREVFqlujAok+fPmrma39//2iXxyeaIaK0p0Tt+qjzTl+1vnfxXFw6uDfqumrN26DfnO/RZ1oreBQ4Ak3oMVgbwqDT5ECodUMcW6VRqVKLhn6NgFvX4jVacWRtZN2GrX1kwfjVYz7J9tqIiIgo6RI0QZ65zz6LPHtZtGhRdOjQAeXLl4fBYMBff/2lJscjovSnUusOqlvUyc3rsWXqr3DO7Ib8pctFm3iv81cj1Lq/ny92TVsI/VNPaB2LQutUDNqgYlgzzhcO4SvgWsKIlh9/Akdn51ee58LeXQh89FAVjldu0wl7Fs2G39FDqNK201t9vURERPQWaiwePHiA06dPqzoL+V8WPz8/lQ4laVJlypRB2bJl1dKsWTOkRayxIHqVMSICG6f8hCuHD6ii7q5jf4RHgUKv3VU+G9bg0voTMOjLQOfgHnW5ne45bHEO7t5uaD7oQ1WPYdDrMe+z9xD46IGq5Shesy5mvN9LirgwcOp8uGb///0TSxca2b/d3tHJ4vvGyjrRA8FERG8FaywoTUyQp9Vqce7cuWgBx/nz5/H8+XOkRQwsiGKn1+mwauJo3Ll4Hs5ZsqL7d5OQ2SNHvFrHbvhzCp6d0UFnW1bVY5jYhz2Cjc0FuBS2gv/pE3DK7IYBf8yBnYMGy8YMx93LF1G/zyBUbN4mSb8WnTYUC4Z8CCtrK/Se9KfFgotTWzZg98LZaD98NApWqGSRxyQiSg4MLCjVFW/HRqPRoHLlyhg4cCD+/PNPHDhwIM0GFUQUN1t7e7Qd+o3q1hT87ClWTRyD0JdvnsNGRiTafz4U/RaMRNeJlZDVYx8cQ09G1mM4uCPEpg7unL2qbmvUOeL4jshGEEWr1FT/SzpUUl3xOYCXTx6pVKsz2y3TxjYsJBgHl/+lRiyksJ2IiCgjsuiIRXrDEQui13v55DGWjBqKoCePkdurBDqNGge7/4qtE+LeDT/snL4QQXfDEaK/AFg5wiHzAFhZ2UETeh1GzWW8eH4ZVlbWeH/mIjWakVhLRw3DvSuX1LoaFflzbqK22dyRNctxYNmiyB+srDBo6nxk4ozkRJRKyaGfnH0WciZa0tiJUt2IBRFlLHLw3HHEt3BwdlYH65umTEJEhCHBj5O7YFG8M/Fb2Gd/qn52cs4MR+0NKVqA1rEQwqxawMomB4zGCCz87EscXLcqUdv75M5ttZ1SB+GSJStCXjzH+V3bkBThWq1qwSuk5kRqQS4d2JOkxyQiSk4SSLi5uamFQQWliq5QQ4YMifVyeYNKSlSRIkXQtm1bZM2aNSnbR0SpnKRDtRs2CivHj8K144exa/4sNOz3foL/WMks288D7sMxkyv6//kH7DWOOL1nB04t34GI4ALQ2xWG3vAAWp0dTm/OgkurF8DawQ+Fm5RC3U494vUc5/4LIgp7V4FnmQrYOW86jq1fjbKNmsHGNnET+Z3duUWlgbnlyAXvVu2xc+40XNq/G5XbdOQfbCIiylASHVicOnUKJ0+eVC1mixUrpi67cuUKbGxsULx4cUybNg1ffPGFqrMoWbKkJbeZiFKZvCVKo8Xgodjw6w84s22T6tyUkNawMspxePU/al0OziWoEOXrNVKLOLR+JXz+PoQI/S0YDSEIc8wPID/O7wD8Ni6Gtd0V5KtdAPW794p1tm8pHL+4b5daL9OgqWqTe3j1MlVvcXHfbpRp0CRRRezHNqxW61XadUbRKjWwZ+EsNaHfo1s33tgti4goJcix286dO9V6w4YN1bEbkSUkOhVKRiMaNWqEe/fu4cSJE2q5c+cOGjdujO7du+Pu3buoU6cOPv/8c4tsKBGlbl5Va6J+rwFqff+SBQlKB/I9uA/P7t+FxiUTKjRtGettarTphGx5JZgwonitu3Bx2AFNyEVYRRgQpsmDUJv6uHKoIOYPWo25/b7Fqkk/ITgwModYXDt+RI0suGTNhgLlKqoC9Eqt2qvrjq5bgQhDwlO4zu/ZoYrXM2VzR8k69aFxcUEh7yrquov7dyf48YiI3lZg4ePjoxZZJ0rxEYtJkyZh+/bt0Yo4pLBj7NixaNKkCT799FOMHj1arRNRxlCxRVsEPn6IE5vWYcu03+DslhX5S5d942iFz3+jFXKg/7r2r0WrVFd1EiGPn6D3lAnqsqtnjuPQwtUIf5wTYZpi/82T4Y6Aa8Diz/fCTn8JmtxBMFoHq9uXrtcI1v+dnSvbuDmOrF2hUrB8Dx9AiZp14/1aZb4NCUhE5bYdo1KpStZuAL8jh1RqV513+sDammcCiYgoY0j0iIVUhj98+PCVyx89eqSqx4UUBUmvZCLKOOq+2x9e1WohwqDHup/H4fHtm29s//rs3h1onF1Qvmmr1962SJUa6v8bp08gPEwbeVm5Sug1eQL6L/oE7UYUglvW3dCEHoNNeBAMdi7QOlbG08cV8eDqFXX7y3tP4+R/tRaScuXdoq1aP7L6H9UuNr6kjuLl40dqdvDS9RtHXV6wgjc0mVzVSMbtc2fi/XhEREQZOhWqX79+WLNmjUqBkkXW+/fvj3bt2qnbHD16FF5eXpbcXiJK5aTjUvOPhiBP8ZJqhutVP4zFy6ePY72tHMhH1Va0bAcHp9dPVic1C67uOaDXheHm6ZOxd5ea8D36L/wSPX+ti5xFjkOj2w+EHFPXW9vmhzaiJXyW22JO7wWYN2AU7t2+rbo5yUjI1RNH4vUaZZTlyNrlUaMs5u1qZeSiWPXaap3pUERElJEkOrCYOXOmKvjp1q0bPD091SLrctmMGTPUbaSIe86cOZbcXiJKKxPoDRuFrLnzqjku1kwcqyaRi+nKkYPqgF7a1VZo3vqNjyudpopWrRGvyfKcXTOj49Dh6DvnG9i73lWXOdoDDqE31LoUf4fa1sd930ow6iMbUPz7+2T4xiO48PU5oNKnZGRC0qliKlm7ftQ2ykzfREREGUGiAwsXFxfMnj0bT548UR2iZJH1WbNmwdnZWd2mfPnyaiGijMfRJRM6jPhWpQo9un0T63+ZAIM+PNpohc/KpWrdu4WMVkR+b7yJdF4S104cjfZ4cbl15hSCnj5RheEDZk3BgIX9UbunFVwcpfj7FGz0obB2qqZKzvS6UGydeg5ze8/GvEFf49/Z01U3KXOy3ZI2FbndbaM6WJnLVbQY3HLmgj4sDFeP+sTrdREREWXY4m3zAKNs2dcXZxJRxpTZIwfafzkG/4z9CrfPn8G2Gb+j2UdD1MiDnM1XoxVO8RutMMldtJgKVoKfP8Pt82dRsLx3vOauKFmnQVQb2rI166tFvHz+HNtmz8L9M1kRFv4Q+rBj0GbqKrNd4MYJYJ7PBthGXIGdRxAqdWoOW1j9f7ubxV4TIq+vRK368Fm5RKVDyXMTERGld0maefv58+f45ZdfMGDAALVMnjw5aop4IiKRo1ARtB4yQtVeyEH2wX/+ihytWLVMXV+heRtVuB1f8jhFKlePVzqUBB/X/kttKmNWYG0uk5sbOg4bjj6/T4KNrS2M+ruwx1poQs7B2hCGcHs3hGqqIDCwAXbO1WLj5GnqfhrXLAgJfhnnc5vSoaSAW0ZMiIhSCzs7O3zwwQdqkXWiFA8sjh8/jsKFC+PXX3/F06dP1SLrcplMnEdEZCKjCo0HfazWj6z5Bxt/n6S6RUlrWVNXpoQwpUNdPXZYFVLHRSbEk/kpJDUpe/4Cr31Mmd/C1N0pS77M6L/oU3SbUAHu+XzgqNsHB+0dROhvIsL4XP4sQxvaGstGnsLcXr9j/kcjsWvZ4mhpU5IKldurBIzGCNV6logotZBRVQ8PD7XIOlGKp0LJxHdt2rRRdRa2tpEPo9fr1cjFZ599hn379llsI4ko7StTv4lqzyp1FVd89qvLKjZvrSaVS6i8JUurmonQwBe4e/ki8pUs88ptjEZjVBpU6frxm0+ncpuOOLtzK26dPYWAq1eQs4gXunz9ddTjzf1sEF4EAQ42OWBnCIfezhVap9KAoTQu7QGublsH24irsM3+EqVb1FWT5t27ckmN1FRq3QFpjYwsrRw/SnX36jL2h2jdr4iIiCw6YvHll19GBRVC1ocPH66uIyKKqXqnHihVr5FalxavFVtGtqZOKElZKuxd9bXpUHcvX1Czeds5aFC8RmT71zfJ7JEzKoXp8JrIdrIm/hfO4UXAfdjY2aHvtInoO7UZCle5CicrKQK/DKuIcITbZ1VpUy+DGqqWtkf+eiDnBvHo1g0c37oBac2dS+dVbUzANT9c2seZxInSC5lte8+ePWrhzNuUKkYsZMbt27dvq5ay5vz9/ZEpUyZLbBsRpTMy5N544MfIlicf3AsUUp2jEqto1eq4sHcH/I76oH7vQa8M55tGK4rVqPPa2bxjqtKuMy7s24Vrxw+rblbu/6VQHVkTWRNSpkFTVTwumvUbBPSLvN9D/1vYvXAxgm+EI0JfAGGafAh39IS1oTAiwq/iwN/HcXrJE1jb3IRTAWvU6dldzbuRmsl+MDm+aS3KNGiialyIKG2TYGLv3sgUzRo1asDGxialN4kyemDRtWtXNRnezz//rN6U4uDBgxg2bBi6d+9uyW0konRERhsk5SipPMtUUKMRMk/Gg2t+Km3JRBschCuHD6p1ORhOCJl7Q2YOl3StI2uWo9Wnw3HX95LqQGVtI9see0qTRz5PdP3mm6ifr547haP/rEVgMCAzeBh0l6HNXBtWVvkReg9YO/Em7MP2wdr2Npw8bVCjWyfk9yqJ1EJmNjftQ2sbGzU7+vVTx1HYu0pKbxoREaW3wEICCjlD2KtXL1VbYd5l4IcffrDkNhIRxToJX8GKlVUAcOXooWiBxeWD+9Ts3Nny5leF2wlVtV1n9bi+PvtRo/M7quBclKrbAK7ZPeL1GEXKVFCLFHTPfK+nCnYc7TcAL3MiwqowdA7ZEeZYUErbVaCx4Ze7cNAehrXNbTjlt0K1bh1RoHjpFPvNXz1+BOHaUNUyuEiVGjixcY1aGFgQEZHFAwt7e3tMmTIFEydOxLVr19Rl0hHKySn+KQdEREnhVbWGCgD8jhxE7e69o9Khzu3aGpW2lJiOJx4FCqGQdxVcP3EUW6f/pgqwraysUaVt5wQ/lsyd4VW9Fs7u2ILcxT3R/KMh6vLDm9fjyo4jCH/sggjrQtA5uCPMUdKuCiA0ANj0awActMdhZXMLjnkiUKl9S3hVfHujBdJRS5So3UCN+pzavB7+F8/hwfWrqoUwERGRxSfIk0CiTJlXO7IQEb2NNrZSTP084D6e+N9SLWXlwPfhjWsq5apE7XqJfuxq7buqwEKCClG8Vl3VQjYxStZuoAILvyOH0Kj/h7DTaFCteRu1mBzb/i8ubT0I/WNnGKwk0PBAmGN+APmhfQRsnxWEfdq/YYWbsMsegqL1K6Nyk5ZRk/5ZUtCzp2rGcrXtderDNbs7ilWvjUsH9uD4xjVo+ckwiz8nERFlsMBiyJDIM23xIZPlERElJynK9ixbQQUAUsQtgcW53dvVdTKJnpNr5kQ/tqRQ5S9dTnVFgpWVSo9KrNzFSiBzjpx48SAAV48fRolarwY8lRu3UIvJyV3bcH7zXgTfDwGs8sLgVB5hGglsckEbBJzcAJxbtR62hhuwdn2K3JUKoV7Xd2Gv0SCpZN4NmX8jl1dxZMmZW13m3aq9CiwkPax2jz4q2CAiIkp0YHHqVOQZrDfhZCtE9LbIZHmRgcUhVGrdHpcP7IlKg0qqml3fxV3fiyhRq76q10gs+U6Uxzi8aqma0yK2wCImT6/iuLxrK0LC/aSpLrJ4XIUVHGF85ooIQz7oHPIj3D4LwpEFCAf8fIDr+3fATncDVpoAuBZ2Qa1uXZDTs3Ci06BK1WkQdVmOgoWRr1RZ+F84i5Ob16Nez/4JflwiIkrfEhRY7N4ddx/zhw8fqhkciYjepsKVqqoWqDJXxLH1qxAWEgxX9xzIX7pskh9bZs7+cM4Si0wMV7J2PRVYSIpR8PNnUS1rYzLow3F03UocXvUPIgx6Nd9HhN6AZ3f81fUyQlOra23Azgb7ly7Hy2vBMIblhs6hIAy2TjDYlpKQAKE3gFUTrsNBux/W1rdh76FDiSY14d2w+Wu38+HN62pfSiqZV/Xo839UatVeBRbndm5F9Y7d4cCaOqI0SeYdkwmNTetEFntvWeqBPvzwQ+TNm1e1m82TJ4+67M8//8THH39sqacgInqFzIUhZ9JvnzutDsZFmfqNLTbfgr3G0SKPkyVXHpVedd/PV6UaeccyOaDM9r115u94fPum+rlQxcpoNOAjNeJxePU/qihdZgWXpUjlaqjf412V/iVePn+Ovcv+wqNz9xARlA0GmwJqRMNUpxH6Aji8Ajjx90rYGG7COtNTZC+ZC/V69EQmN7eobZARlcjnrvLKPCNS0yLteJ/eu6O2RQINIkp7rK2to47ViCzJymg0Gi3xQPPnz8fYsWPx4MEDVKtWTbWglQlXTBOwpEWBgYHInDkzXrx4oSYEJKLU6fS2f7Fz7jS1Lt2bBk6dh0zZsiO1Ob11E3bOmw6PAoXR88cp0eaMOLRiCU5sXKtqGxwzuaJB3/fU5H7mqaUvHgbAZ+VSXNy3W91Oaj9K1KyL6p17RNVCmJOC8Ms7DiH8oT0iIvIjTJMHsIo+EZa1QQd73W1Y2dyDfU4Dnt33hS4kBG2HfqOCl5jO7tyC7bP+RKbs7hjw+xw1xwUREaVfCTketlhg4eXlhQMHDqh0qLNnz2L06NFo1apV1FBbWsTAgihtkC5GMz/oDRiN6ix/+y/HIDUKfRmIGe/1RITBgN4/T0X2fJ4qtWjbzD/w/MF9dRupv6jXe+BrC8+f3PHHoeV/4cqRyAnsZHSmdP3GqNah22uLqu/d8MOh5SsReC0IRm1OhNt5wmDnEnW9IfwmwoNWwwoOcHWoBhuXZ8hawh11unVHFvec6jZ6nQ6zPuqL0MAXqjtU8Zp1kRqEa7VY/cNYOGfJqraLtX5Er595+/Dhw2pdTgZz5m2y1PGwxVKhZEgtNDRUrZctWxZr1qxByZIl03RgQURpg0uWrJEdnM6dRrkm/++slNrISETBCpVw7fgRNcpiNBjUCIBwyZYdjQd8pAKjN8mWNx9aDxmhWuseXP4Xbpw6ruoepOi6yXufoGTt+rHeL3fBouj05Yion2XyvoPrVuDWofPQP3NCkPaeutzaoTjCnLyBCCDkArD06zOwD/sX1tZ3YJs9DFk9C+LuudM4vnHtK6MqKeXklg24c+m8Wi/bsJlFamyI0nNgsWPHDrVeuXJlBhZkMRYLLKS9bLNmzVCnTh2UK1cO9+7dg8YCbQ+JiOJDzlLLfBa5vYqn6h0mB/0SWJzZtinqsnKNm6N2j74JLoaWieo6fDUWdy9fxIFli9SB9Y7ZU5FH2tt6RI4wvI7MgVG3Uw+gE6ALDcH0QT0BHeBewIiwgJ0waj2gtysAvV2mqMn78BIwvgiRaQjx4LofZvX7HG6Fs8O7bUs103hKkIL94+tXRf18fONqBhZERGk5sKhQoQJOnDiBrVu3qlQoZ2dnrF+/3lIPT0T0WpI6lJR5K94WKYrWOLtAGxyELLlyo/GgwchXMmmTjOYpXhJdRk/A8u9GquBi64zf0fmbcQkqYL9y5BD0ujBkyZ0XPb4dHzUKIaMah/9di+sHTkP/2AEREXmgc8gLG/tSMOjOIiQ8Avo7NbB16jPs1i6FtfE2rDM9R/YSOVGrS9eoFKrkdGLTOrU/pe7j5ZPHagTnsUyYmM8z2Z+biIgSGVhIOYa/vz/y588f5yzc7du3VwsREb3K1t4e7b8ag0e3bqJk3QYWaWUrJIho8v4nWDRssKrbkBSrcmYT7iVk7grz1CYZ1ajVtrNaTJ49CsCuBSdw8zgQEX4dNqFXYXAsAp0mh4yjAHrg9jlg6ZmzKoXKyvou7LJpkb9KSdRo09Eik/iZ162c2LRWrdd9tz98ffapGc7lsqbvf2qx5yEiojdLUD9GnU6HQoUKoXnz5li1apXq/ERERAmfH0PSnywVVJhIZ6ja3Xup9b1/zceLhw/idb/Axw/hf/GcWo/P5H0yCtFx2EgU8q6ifi5cS4eGA52RLdcBOBp2QxNyAbbhQTBa26oUKq1DTbwMaogLu3Jh3sc7MLfXNMwbMBpLvhmD03si87wT6/iG1SqNy92zILyq1ohqgXtp/241XwgREaXiVKiIiAgVUPTo0QNubm7o3bu3KtCWrlBERJSyKjRrrbpFSd3Ftpm/o5OkRL2huPrS/j2qo5akZLm6x3+iUzmIl1nPL+7dhZpde6L4mKpR10kK1ZF/1+P6oVPQPbIDDLmhs8/33yR+UgdTHKGPgYPLgKOLVsNGfxtWmkdw8XRCxTYt4lWvIYGDFG2bZkmXURsJ2mS5d+USTm3ZiFrdesb79RARUQrUWMydO1fVUCxevBgLFizAL7/8gho1amDQoEHo3Lkzi7aJiFKIHFxLCpCkRN0+f0Z1iyrbqNlrU1xNaVAl6zRI0HPlLVFaFZBLd6oz2/5F9U7do6VQ1WzbUS0mMonfgZXL8ODcbRieZ4LRKi/CHHIh3N5NLSLUH5H1GmFLYR1xB9bOz5DFKztqdukM99z5oj2/zFCuDwtDziJeqnbFpFLr9lj/yyVVIF+1XWfYsZEIEdFbkaB5LMLCwuDo6IibN29Gq7OQou158+Zh2bJlakTj3XffxR9//IG0jvNYEFFaJTUGexbNgb2jo5ozwzV77CMRAdf88PfIz2FrZ4/3Z/2V4M5UMov4pt8nwSmzGwb+OU/VkCREwK1r8Fm9Bs/8ngLBWWGwzgedQyxzcRgNcNDeh7XVHdhkDkKWou647rMPBn04On79PQqU/f8IR0SEAfM/f191CZOJBmUUh4j+T47Vbt++rdbleE5m4iZ66xPkxRVYmOzcuRP9+vXD/fv3VT1GWsfAgojSKjm4XjbmS9y/chmeZSug48jvYk2J2jV/Jk5t2aAmupOWvQl+HoMBcz4ZgJePH6kOV2UbNk3ytl85eRSn/t2G4NtaGLXu0Nvmh94+esev8OAdqiuVjVU2ZNIUgl12XbTicNNs7Jk9cqDflFmwtuYM4UREqX6CvICAACxcuBDz58/H9evXVWH3tGnTkvqwRESUBHIg3eyDz7Bo+GDcOnsK53dvR5kGTaLdRs72y4hDYtKgop7HxgYVm7fB3sVz1ShJmfqNE9TmNjZeFauoxbxe48y+Hbi8ywfaACsYdK7Q6iInw7NxbgitXV5og4ALu4DL23bATucP2N6FlbWNKmD39TmAEomcIVwK26+fPI7iNeuoNsFERGThEYurV6+quSok/WnLli1q9EJGKvr27YtcuXIhveCIBRGldcc2rMa+v+bB3tHpv5So/6cZXT1+BOsmfQ9ntywYNG2BChISIywkBLM+7KO6M7X4+AsUrFBZpVQlNcCIy5Zpv+HC3h1wzZETGvscCH/iAGOEFIfnRYTN/1OxwkMPwaA9DGvr7MhkWx7Wjo/h4qlBxVbNUKRcpTc+T0jgC5UmFvjooZoZvdn7n8GzbPlkeU1Eb3vmbUljF97e3px5m1J2xELehKGhoWjXrh02b96Mhg0bJuZhiIgomXm3bAu/Iwdx388X22f/qWbqNqVEXfqvaFvSoBIbVAgJIso0bIoTG9fg3z9/UZdZWVlD4+ICTSZX9b+jSyY4qvXI/12yZkPRqjVgr3FM0HM9vXcnqti81SfDkKtIsVeKwx9KcfgLF9gZPfASNoiIeIxQh2ywti4bWRw+PRC7w/6BdcTtOIvDDXo9Nv72owoqRNCTx1g5/htVr1G7R2/YOVhuLo6UIO2FbWztUv1M9ZR8gYUcv4ny5cszsCCLSXBgUaJECdVetlevXsiWLZvltoSIiJIlJarp+59h8Vef4ObpE7iwdydK12sEbVAQrp04kqQ0qJitZ/3Pn8WzgHsI14bCaIxQk9fJEhfp6tRmyAhkyxv7pKuxObRiiXrswpWqRgsqRCY3NzQf8H60y9b/+gR+hw/CqN0MjU0FRKjicI//CsTdgQgg5DKw/NvLcNDuhJXVXdhmCYbRLghPbl2HncYRXUaNx/k9O3Bm+7+qHuXmmZNo/vGQV54/rXhw45qapd3W1g4Dp81PEzPWE1E6DCwcHBxw4cKF5NsaIiKyuGx586FG53ewf8kC7Fk4W6XzXD9xTJ2Vd89fAB4FCiX5OVyyZEXPH6dE1URog15CK4GF+v8lQoMkyHipLpdg49aZk3h61x9/jxyCxu8NjlcNxKNbN+B7aJ9al9cTHzW7vKsCi3D9S7T/qTuy5s6Lq2eO4+TGLQi6FQqj1gN6W0/o7V0R5igBTn7on5+HPmSbur+9vjT+nbAE9jkMKN20JW4c9cGz+3exdNQwVG3fBdU6dIONbZLLFeMswL9x6jhunjmFii3aqAkQk0qyn6UeRuYt0Yfr1P5k1ywispTk+TYkIqJURUYUJCVK2svumD1VjViIEhYYrYhJ5rCQQEOWuIS8eK7a1MpcG//+PklN6Fev1wB137gcWvG3+t+reu14B0PZ8uRTM4TLRH5SXN544MeqvsK8xkICoVN7tuPKbh8E3wuBNtQv8nVoqsPgWB2hMr/GM+DFUcAmLAvsrLcjPOIBDq9ahss++9Hui68TNOryJtrgIFzYswOntm7EiwcB6rK7vhfx7oRfk5SyJq6fPAr/C2ejfr6wdxcDCyKyGDYuJiLKAOSAtOkHn6mz69dPHlMzU0sdRGK7JSWVzHvR8evvUK1DV/WzTGb3z5jhUTUNMUlAdPXYYbXNNTr1SNBzVW7VQf0vM4RLQBOTBDOVG7dA26GfIULzSMYKVDBSqnFuuDjugKP2EBxCb8MqwgCDgwdsMr8DO+eWgJUGz+/dxYIvPsbUdz7G/MFfYfOcGQgOfJGoffLkrj92zJ2OWR/0UXOQSFAhnahkLpJHN6+rmcSTQkao9i6ep9ZL12+i3hMPrvvhsf+tJD0uEVGaH7EYO3Ysvv3222iXFStWDJcvX1brWq0WX3zxhZq0T7pZNW3aVLXBzZEjRwptMRFRysqezxPVO/XAgWWL1M+SEiVF1ClZ/1Gza0/k9iqhir4leFj81af/dZaK3rXp4PK/1P8latdTqV0JkadEKeQsXFQ9/ultm2JNo9LrdFj3y3gEP3uq9pPM6aEKy9/9/20e3fPHwX+W46nfExgN2RDu0Bqh+mOI0N+EVn8Tuqd5EXQkO24cPggH3W3A7j6c8lihTPNGKF29VqzbZoyIwI0zJ3Bqc2TtholsQ4XmrVGiVj1cOrAH22f9qfaBV7WayJQtOxLj7I7NKo1Liufr9eoPbVCgCtak7qbuu/0S9ZhEROkisBClSpXCjh07on62Nctz/fzzz7Fp0yasWLFCtcj6+OOP0aFDBxw8eDCFtpaIKOVVbtMRV4/5qIPsmPNapBQJInr+MAUbfp2otmv1D2PVSEb1zj1U8HHn8gVVeC5n2Kt37J7gx5cuWJVad1Bdnk5v3YTKbTvBzt4hWt3BjjlTEXD1iupa1XbYqFi7VUnXqHaffxHtsjMH8uHY8lV48cAfEfo70AUukGdEqJUzrPQuCLzujAfTN2PP1C2wNobCxlmH7F65UaNjF9y/fEEVg8sM4f9tKAp7V1XzguQrVSaqe1eZ+k1U8bhMdrh74Sy0GTIyUelVh1YuVes1urwLBydnlKzbUAUWErjU7t47yWlWRERJCiz279+PmTNn4tq1a1i5ciXy5MmDxYsXo2DBgqhVK/azM5YkgUTOnDlfuVz67M6dOxdLlixBgwaR+cMygZ90tDp8+DCqVauW7NtGRJQaycFjp2/G4eHN68hXsgxSC1d3D3T99ieVAiRpUYdX/4N7Vy6rkYOD/yxWtyldrzHcciZurqSiVWrA1T0HAh89UClR5Ro3j7ru1Ob16qy9pFm1+vRLuOV49e9KXMrVaqAWCQ6kna/UL6jpoYxBMBpkibzdf/8BL4CgYzdw89j/T3LJRH4FKnijYZ9ByOzx6nPLfCCNB3ykRnP8jhxSqWyFKlZO0Os/sma5KqbPmidf1OzohSpUUu2AZZTm1rnTKFjeO0GPSWmXHD917979lZOyREmV6HfTqlWr0LNnT7zzzjs4deqUSjcyHdRPmDAB//77L5Kbn58fcufODY1Gg+rVq2PixIlqsj6Z9CU8PByNGjWKum3x4sXVdT4+PgwsiChDk7PVqSmoMK91aNT/A+QpVgLbZv2hCrvnD/lAdZKS2pCq/9VjJDagkjk9di+YhROb1qiDazlglwPqPdIlCUDdnv0TPQGeBDydR41HhMGg6jiCnj2NXJ4+wZ0rl3Dn/HmEBWoRESEVHOGAUQsr62yw0ZSHjX0J3L9uj+XDd8MGt2CTORB5KnuhTqdusNdEzpfh7lkQ3i3b4fiG1dg5b7oa0YjvXBovHgao4Em9xnf7RY1MyDwWUmMjoyYSWDGwyDisra3h5eWV0ptB6VCiA4tx48ZhxowZaj4LqWMwqVmzprouuVWtWhULFixQdRX3799X9Ra1a9fG+fPnERAQAHt7e7i5uUW7j9RXyHVxkeDIFCCZZhokIqK3S+oKpOvT+l8mqAnxRNnGzaPNGp4Ypes3Vp2lnt2/h2snjiJ7/gIqPUrqHErVbahauiaVHLRL3Yp57Ur5Ji2i3Uan1WL/6n9w55gv9M+0iAh7Ap0mF3QaqQHMAYQBvgeAq3t2wE53A1aaB3At4oIanTvB12e/KnCXjlS1e/SJ1zbtW7JQFW7nL1P+ldoVed0SWEh6nKRLSbE4EdFbDyx8fX1Rp06dVy6Xeobnz1/tumFpzZv/fxi7bNmyKtDw9PTE8uXL4eiYsJlcTWTEI2ZBOBERvX3SvvWdib+q1CgpOK7WPvGjFSZSN1GucQscXbtCLeG6MDUaIoXdjQZ8FFXTkNxkFKJhj96AWXMrmVvj+Np/EeKvh1GXGzoHTxhsnWCwLSWH/wi9DqydeB1WIYWljBxH162Cs7s7KjZu+drnuut7CVd89qv6DRmtiPkaPQoWVoXi0hnqis8BlG3ULLleNqWymbfPnTun1suUKcOZtynlAwupbbh69SoKFCgQ7fIDBw6gUKGkT7aUUDI6IcN6sk2NGzeGTqdTAY75qMWDBw9irckwGTFiBIYMGRJtxCJfvoR1HyEiIsuQQKDJoMEW3Z0VmrbC8Q1rcP+qr/rZ2S0L2gz9Grb29khJMefWePn8Ofb98zcenruLiJfZYLApiHB7NxhdasA66BEiwq9hz4L1OPHXU9hY34Jt9lCUaFJTtc2NPhneHLUus63HNveHBBpSxL3vr3kqHYqBRcYJLNatW6fWS5YsycCCUj6wGDhwID799FPMmzdPfTHdu3dP1S8MHToUo0aNwtsWFBSkisil7sPb2xt2dnbYuXMnOnbsGDXCcvv2bVWL8bqZxWUhIqL0SVKUJNXqwt4dqm6jzRcjkSlr4tq3JqdMbm5o+d5H0SbxO/LvelzbewK6EDe8gC2M+rsIsX8GW4cawEvg6Crg1JIVsIm4DtusL5C1UDbc9/NVtRgyA3lcZH/s/3uBmttERoey5Mrzll4lEaU3iQ4svvrqK0RERKBhw4YICQlRaVFyUC6BxeDBlj3DFBt5ntatW6v0JwlqxowZoyJu6XIg6Vj9+/dXow9Zs2aFq6ur2iYJKtgRiogoY6vRpQeCXzxT7XZlDo20QArba7btqBYhRdx7/5oHg3YnnPAQVhGFEabJi3CHbAhHNhhD9Hh8SFrfAjYGD6z67hcUrl8B1Vq0e2V2c5khvUC5Crhx+gQu7tul5hYhInprgYV0XGrWrJkq3h42bJhKP5IRAxlOc3F5O4Vfd+7cUUHEkydP4O7urtrbSitZWRe//vqr6nogIxbmE+QREVHG5prdAx1HpO16ugrN26gg4NHtm8hXOTuafdgX92744cCSfxB0Q4fQUC0QEQhYOcPo0hZBWjuc2QxcXLceNoZrsHV7jkJ1y6N66/Yq0JB0KAksLuzbpSYQlI5ZycnvmI+aWVwK6lkwTpR+WBlVw+2EkwP4Q4cOoWjRokivpMZCRj+kha6MehAREaUWMs/H0tHDpJgCXcZMjGohHBL4AnM/GQhdaAiy5isC43NXGMPzIsyhAIzW0c8n2oU9hU2EH2zdA/H8wRXodWGqbW7+0uWSZZul89TOudNx+eBe9bODszMqt+6oZhmPbVJCSh5ShyoNa0z1pdJJk8gSx8OJPiXx7rvvqknoiIiI6O3L7VU8arK7HXOmwaAPV+uHVixRQYVHgcLo89Nk9JvzHfovHIROo4rBw/MwHPW7oQm9CqsIPcIdskLrWBVBQY1hRDF1/3UTJmPp6LG4efm8RbdXJg9cNGywCipkRCRzjpwICw7GgWWLMOfj/iq9KzxMa9HnJKI0UmOh1+tV4faOHTtUsbSzs3O06ydPnmyJ7SMiIqI41O7eB1ePHcbTu/6q21WRytVxdsfmqAn/zFOaPPJ5ovOIkVE/B9y6hr0LlyD4pgERhgKIcCgJg+4sdIZAPHlQDZt+vQ8H7RFY291C5mIuaDJwkCoqTygpPJfZ049vXKNGV2QywRYfD0WOwkXge3AfDq1comYul5oRuU2Vdl1Ud6qYtSBElI5ToerXrx/3g1pZYdeuXUjrmApFRESp3aX9u/Hvn7/A1s4e7gUL4f6VyyhcqSraDUtYh8ZLxw9j62+TYAgPg8a+OuAcvYuijT4Udjo/2GR+iKKNvKMKyV9H5sf494+f8ejWDfVzmYZNUa/XgGhpTzJbudSL+Kxaqib/E5myuaNah64oVa+R6t5FliXNdy5duqTWS5QooWpSiSxxPJzowCIjYGBBRESpnfwZXznua9w+fzZq9u/eP09F1tx5E/xYh1f/o0YXpF6jQPWauLT5EPRPs0Bv6wW9XaZot7UPewBrXIVT/gjU7dMDuQv+v+ZSZjM/tXUj9v09H4bwcDhmckWT9z5BkcrV4nxuSeU6v3u72oagp0/UZZIuVb1jd9USV14XEb19DCxSYEcSERGllKf37mDRsI9h0OtRvmkrNOz3fqIeJ/DxQ8z+uL9KWRrwx1xk9sihLtdptdi+cC4enLgHozY/wjSFYLT+/4G+1Gs4aK/B2ukOclbKi6CAu7h97rS6rmB5bzT94DM1GWF86HU6lc51ZO0KhLx4ri7zLFtBdfJK7m5VRJSCgYXMbC0F3KbhNGk3K/NHyJOnBwwsiIgorbi4fzdunDqOhv0+gCYJrd9XfD9SjX7U6PKOGi2Ize0rF3Fg0T/Q3rWDAV7QOUROMmjQ+SE8ZDtglCJsK7jl9kTb4V8he66Ej56Ea7U4vW2TqsHQh4Wh2Yefo1Tdhol+XfR/TIWiVBdYHD9+XM0N4ejoiCpVqqjLjh07htDQUGzbtg0VK1ZEWsfAgoiIMhqpd9g8dTLccuRCvymzVN3km4qz961aggv/7oAu7Jm6zMrGA3bOzWFtkw3WhjDYh/nCNstDlO/UCOVqN0jQ9hxdtxL7lyyAU2Y39PttJhycojeLoYRju1lKde1mP//8c7Rp0wY3b97E6tWr1XLjxg20atUKn332WWIfloiIiFJQ0So1YKdxxPMH93HX9+Ibb3//yiVc3b8nMqiwskLJeg2Rs0g+OIVfgZ3uOSJsHKB1KougsEY48Dcwp/cCzBs4CptmTlVByZt4t2yLLLnyqLQon5VLLfQqiSg5JLrVgoxYzJ49G7Zm3Rpkffjw4ahUqZKlto+IiIjeIjuNBl5Va+LC3h24uHcn8hYvFWctxIFlC3Fi0zr1s9RjNPtoSLTbS+CwbcEcBBy9iwhdIYQ5FkCYY34A+XHzFDDvvXWwNV5E5pIatPjwIzjGaF0vbGztUL/PIKyeOAantmxAmQZNkS1vvmTcA0T01gMLGQq5ffs2ihcvHu1yf39/ZMoUvXMEERERpR2l6jZQgYWvzwF1UG/noIl2/YPrV1W61JM7t//fRrZnf9g7OkW7ncxF0WLgB8DAyJ/PHtyNUyu3Q//UHTr74gi3z4pw1ELoNWDRJztgq78I54I6NH5/ALLlzBP1OFIALi10rx0/gl0LZqLT19+/MUWLiNJQYNG1a1dVqP3zzz+jRo0a6rKDBw9i2LBh6N499mIvIiIiSv3yligNV3cPNa+ETMAn7V5Nc04cXbtCzTkh61L30PT9T1GoYuV4PW7ZmvXVIh7d88eOGXMRelsDnU1J1c5Wb1cV2gBg+agzsNcthUPO56g7oCfyFS2Ger0G4uaZk6rb1NWjPihaNfLYg4jSQWAhAYWcLejVq5eahVvY2dnhgw8+wA8//GDJbSQiIqK3SNq6lqzTEIdXLcWFvTtVYPH03l1smToZ96/6qttIulTDAR/CyTVxnSDdc+dD9+/GqvXgwBf4d9o0BPrqYbAqiXD7LNA6VoT2BbBh0k04aLfALsdDeNWsi0t7dmDP4jkoUL7iKyMpRJSykjxBXkhICK5du6bWCxcuDCen6MOgaRm7QhERUUb1POA+5n46EFZW1qjRuYeaV0KvC1NdmWSejOK16iVLOpLMmbF17kw8OvkUhoji0DlEzqUhrPTB0L5cACPCULZpSzTu94HFnz8jYFcoSghOkGchDCyIiCgjWzZmOO5e/n9nqPyly6nJ7lyzu7+V55fi791LF8F/300YDCVVkGHQXUF48EYp60Ym+0rIVCgCzT5+D1ncc76VbUoPDAYDzp07p9bLlCkDG85qTikdWEycOBE5cuRAv379ol0+b948PHr0CF9++SXSOgYWRESUkUka1JZpv8LW3gF13umD8k1aptjs16YOU/d97uNl6D0YIgJgbVcE9i5tYKMPgV34ObgUNaDVJ4PhnMj0LCJKocCiQIECWLJkSVThtsmRI0fQrVs3NadFWsfAgoiIMjI5RLh24iiy5/OEW47UMyLw4MY1/PWVzJllhKOmKYyO/29xa6sLhC3OIG/t3Gja9792VET0Vo6HE128HRAQgFy5cr1yubu7O+7fv5/YhyUiIqJUQmooilSqitQmR8HCqNi8NU5uXg9NlvNwzR2MwIt6hNuUgd7eFXrUxtUjgP+eRbB2voxK7zZF2Rp1U3qzU42IiAhcvXpVrRcpUgTWKTQKRelPogOLfPnyqfayBQsWjHa5XJY7d25LbBsRERFRrKp37oFLB/fi2f27KNuwKSoN74CXz59j469TEHIjE8I0pRDmmBeIyIv9C3U4NuMPOOR6iOrd28LP5wCe3vFHk/cGI3v+AhluD0s3z6VLI2cxHzFiBOzt7VN6kyidSHSIOnDgQHz22WeYP38+bt26pRapr/j888/VdURERETJRePsgto9eqt1mVcj6NlTZHJzQ/dvx6D/oiFo8lE2ONvtgEPoDcDKBiH2WfDobiDWThyLC3t2qLa5qyaOwcunj/lLIkrpEQuZCO/Jkyf48MMPVdsyycN0dHRURdtfffWVpbaPiIiIKFal6zbC2R1bEHD1CvYvWYDmHw2Juq5IuUoo/Ls3bpw6ju1zpyHo2aOo66To22h4gqCnTzD34/fR9LPPUKJKLe5lopSexyIoKAiXLl1SQYWXl1e6Gk6Lb7GKtG0LDw9/q9tGGZd8xpgPS0QUSUYelnz9hVrv9t0k5ClWAga9Hr4++3Fs3Uo89r+lrrO2sUXxmnXx/PFThFx3hta2ILTBqwBjCKxt88HVxhNZytugzeDPYGtnl6y7NywkGDdOn0DhSlVhZ+/w1n+VnMeCUk3xto+PjxqpaNWqlfrZxcVFBRZjxoxBcHAw2rVrhz/++AMODm//g/K2SUwmRezPnz9P6U2hDESCCqltSk9BPBFRYuUqUgyl6zfG+d3bsWv+DJSq2wjHN67Gy8eRIxR2GkeUa9wcFVu0Qaas2aPud/vKRWyb7IMXz24gQu+Pl9auCPVtgvkDl8PW6QJqv9dBjXpYmjYoCMu/H4lHN6+jRpd3UL1jd4s/B1FKSXBg8d1336FevXpRgYVMsNK/f3/06dMHJUqUwKRJk1Tx9tixY5HemYIKDw8PNeN4csxAShSzk8e9e/dU57X8+fPzPUdEBKB2997wO3IID29cU4twyuyGis3boFyTFqoeI6b8XiUxYMZv8Dvqg/W/jIdBdwFWcILOuTZ0EbmwbeoT7A37CVnKWVlsFCMsJASrJo5WQYW4euwwAwvK2IHF6dOn8f3330f9vGzZMlStWhWzZ8+O6hYloxfpPbCQ9CdTUJEtW7aU3hzKQKSlswQX0tXDLpmH64mI0gIJImr36IMdc6bCLUcuVGrdHiXrNoxXmlHRKtXReODH2D77T+h1x+DgcA+24TUQ5pgPWsdKuH8FWCCjGJnOo9nQAcjpWThR2xgepsWaH79V9SAS6GhDglUQJEXnLlmyJuoxidJ8YPHs2TM147bJ3r170bx586ifK1euDH9/f6R3ppoKGakgeptMKVAS3DKwICKKJOlOhbwrw9ktC6ytbRK0W8o2aobAxw9xZM1yhATfR9uhJXDp2FE8PByIMLsKCNPkQlh4Lqz9/iLsIxahdLfKqNIkMnMjPvQ6Hdb9PB53L1+AvaMTOn0zTgVBAdf8cPP0CZXK9TbZ2NhEHbvJOlGKtZuVoMI0q7YU/5w8eRLVqlWLuv7ly5cZ6mCH6U/E9xwRUeogNRQJDSpManbtiRK168MYEYFNv09C5abN0W/BSLQc5gkn6x2wD3sEg60zQu3r4tgqDeb2mozVkydB/4bmLVJIvuG3H3Dr7CnYOWjQYcS3yFGoCApWiKzfuH7qGN42CSaqVKmiFgYWlKKBRYsWLVQ72f3796tJVeSMfe3ataOuP3v2LAoXTtwwIREREVFKnShs+v4nyF+6LMK1oSptSUYxpBaj77QJePfPVsiaYz80Ib6AlTW0TuVx/4o3Fgz4G4uGjMSzRwGvPGZEhAH//vkLrp84Cls7e7QbPkp1rRKFKlRW/0vAIcEHUYYMLKS+wtbWFnXr1lV1FbKYd6eRSfKaNGli6e2kNKJnz56YMGHCW3u+AgUK4Lfffovz+ps3b6o/FlIb9LbMmDEDrVu3fmvPR0RElmFja4c2X3yN7Pk8EfzsKVZPHKu6OAlHZ+f/Jt/7AF41bsBR6wNrgw5hjvnxMqQRln91GPMGjsLZQ3vV7WXkY9uM33HFZ79qddvmi5HIX7pc1HPJqIWja2boQkNxz/fiW28EIn8fZZF1ohQLLLJnz459+/apWgtZ2rdvH+36FStWqOJtSr2kg5e0BTa3cuVKaDQa/PLLL1G3kQPymMvVq1fjfNwzZ87g33//xSeffBLtcrlP3759kTdvXtWGWFqldu/eHcePH4/3Ni9YsABubm5IbWSfrF27Ntpl/fr1UymCMqpHRERpi4OTM9p/NRYuWbPhyZ3bWPfLuFfSnRr36o9+C75Gww+ywhE7Yad7Br2dK0Jt6uPAAi3m9v4ef48ejgt7d8LK2hqtPh0elfpkIpcXLO+t1q+fiv/fQ0uQ5h8LFy5Ui6wTpVhgYSITZcSWl5c1a1b2109j5syZg3feeQfTp0/HF19ETjIkmjVrptqami8SFMRF5i/p3LmzmtvERIIHb29vXLlyBTNnzsTFixexZs0aFC9ePNpzpTVSXxQXGcHr0aMHfv/997e6TUREZBmu2d3R4auxsHd0xJ2L51Va1Pk9O/D03l01h5WJV8Uq6DdjPLpPbojMmXfDIfQGIqxs8dKoxwO/y+o2ectUQNGqNWJ9nkIVI9OhZHZwogwdWFD68NNPP2Hw4MGqbbCMKpiT0YWcOXNGW+Iq8pIORTLqYZ4CJF++MvJRtGhRdfa+ZcuWqv6mfPnyalRr3bp16nZ79uxRZ/7NJxqU1CW5TIZp5XrZNpnx0TRyYt7OOCQkRI0SZMqUSc3tMGvWrFe27/Lly6hRo4YalSldurTqZmbu/PnzqkOGBEXSoEBSuh4/fhx1vczd8vHHH+Ozzz5To3ZNmzZVaVhCRu1km0w/C9kP69evR2hoaIJ+H0RElDq4exZE6yEjYW1jg9vnTmPr9N8w//P3MH3gO1jz03c4snaFCjrCdWHI5OaGd3/8Hn3m9ILGcRUMYSfUY9g6Ncaj23Uxt9fvWPtrZEaAOc8yFdTIhYyMvHj4IAVeJVEKt5uluMmBdGi4IUV2kaOdTYI7VH355ZeYNm0aNm7ciIYNGybp+aVoXw78K1WqFC04uHDhApYsWaJmi44pvqlNEhBIHcXo0aPh6+urLjMfFZH0Lan9GTlypApuPvjgA1UDVKxYsajbDBs2TD1GyZIlMXnyZHXgL93NZA4SCWgaNGiAAQMG4Ndff1XBgOybLl26YNeuXVGPIUPG8tgHDx6MGp2TeUzmz5+vRnfMgy7ZDzK8fOTIERWUEBFR2lOgbAV0/24SfA8fwL0rl/Hguh9CXwaqYmxZhAQeHgULI7dXCRiNEXhx/7a63MOrJHS39NAaI6B1Ko27vsDc3rPgXCgAnUaOUBPuaVxc1P2kDe2N0ydQvkmLFH7FREnDwMKCJKgoOXorUsLF75rCyT7+v87NmzerEYOdO3eqg+rYSMBhfgAvZ/SlhiY2t27dUgfWcqBt4ufnp/6XtKekkNQiSb2TwElGTWLrVPbhhx+qdQkIJDjYvXt3tMBCRhs6duyo1iXla8uWLZg7dy6GDx+OP//8ExUqVIhWdC5NCGSyR0nh8vLyUpfJyIuM8MQWIMXcLumWJtss+4WIiNKunEW81CKk1kImtbt35RLuX7mMu1cuqSJvmfROFhOZrK9K205qfd/qpbi6/ga09pWgdSwC7f0iWDBwKRw8rqDTmBGq9kIFFqeOMbCgNI+BRQZVtmxZleojKUnSx9o8gDCpX7++Ogg3cXZ2jvPx5Cy/pE6Zj5qY56Em92sxMQUfDx8+jHab6tWrR61LVzMZUbh06VJU0bkEIrHtg2vXrkUFFlIrkhCOjo4qTYuIiNIHGWXI7VVcLaa/cy8fP1IBxj3fS3hw4yqKVqmByq07RN2nTofuqNMBOLlrG07/tQ9hNlURpsmLsMC8+OujdbDNclLd7vb5syqtKj6zhROly8BC8ualIFcOviQFJU+ePFi8eLEq8K1VqxYyGklHkpGDlHruhJDflfzOJHiQNB4ZwZAaBXMSSBQpUiRejyd1B3IQLUXNpvbDpgNyqW+QEYG4mNKkzAMR08zm8RFzQkYJLhLSPi8oKEilRv3444+vXJcrV654BVaxefr0Kdzd3RN0HyIiSjvk742ru4daStSs+9rbVmzQRC1XzxzHvj/XQmesCp0mJ8JCmsPK6hb0ulDcOnsaRSpVfWvbT5RqirdXrVqlCljlrOypU6cQFhamLpc8+7c5j0Fq+4KRdKSUWBIzA7inp6cqYg4ICFDBhcyanlhSkC2k65P5ZVLTIDUQsR3om4q1TQff0nXKJOa8ExKsSIF4Yh0+fDhqXWofTpw4gRIlIicpqlixoqoFkeJrCaTMlzcFExLUxLZdEmxrtdrXBlRERJTxFClXCf1mj0OzoQXhaNgNW4MW1naRJ/E2T16CZd9998bZvJNKUpcbNWqkFs68TakisBg3bpyaCEwmyDM/Y1yzZk3Vw5/SBqkjkK5LkjokgWJgYGCiHkeCAzlAP3DgQNRlEuxIYbPUKcjs7DLHxfXr11Wh9/jx49G2bVt1OzmAl+2QTk9Sl7Fp06ao+TRM5KBfRhakJkRSuBKaYjR16lTV5lZGTz766CM1B4t0khLys4wuyNwax44dU0HB1q1bVSeqNwUzsl2yTRKcyWOaj+YVKlSIs9ATEVHsfz+Kl0a/2d+j0eAc0FgHq8vCjY/x+G5NLBywACt+mJhse06CCTlek4WBBaWKwEK689SpU+eVy6Vg1bxtKKV+MnGdBBdywJ6U4EK6Kv3999/RLpP6DZnLQoKHgQMHqlGCNm3aqBEC04zZEpguXbpUHfRLvYSkJEngGrMz1Pvvv4+uXbuqICa2IurX+eGHH9RSrlw5FfxIK1hJ3xK5c+dWnZ4kiJBZ48uUKaPaykpRdmzdrMxJALR9+3YVGJmPTsjrkddLRET0xhGMmT/CysYGxogXsAp/CK1jYTy8WRVze8+ItU0tUWplZUxkha2cjZX5AmQYTXLzpQBWLlu0aJE6gDNPiUmr5ABbAiVJ73J1dY12naS5SLtSqSeRuREosoBbOjH9888/0YqlMxoJmqTTlozUyPvH0vjeIyJKf1aOH4VbZ0+haI06eHQ6DGF2VWG0jiyF1YRcRL6GGjTpPcAizyXpyab0Y6klfNNJNMrYAl9zPBxTot9Jcjb2008/VX36JeXl3r176mz10KFDVa9/ynik3kYCS/OJ5TIi+bKW/ZAcQQUREaVPBctHzgMVFvgC/eePQqVOOjhqfQCjAVqnkvA7VADz+kzA2YO7k/xcUms4Z84ctcg6UYp3hfrqq69UxCsTq0m+u6RFSbtRCSxkJmfKmDgZHNQoHhERUUIUqlgJexbNxp1LF6ALDUGVJq3Usn/1clxZ5w+tYwWEaqrh4AItji/8Bq1GDYRHPk/uZEpVEj1iIaMUX3/9tSp6PX/+vOq68+jRIzUDMhERERHFX5ZceeCWMxciDHrcOn8m6vLaHbqg/8IvkKfYKTiE3kSEjQNCrRtg7ZjjWDT0a+i0Wu5mSjWSnFQnbUClpWhck6wRERER0ZvJLNzixsljr1zX7vMv0GtmD7i67oJd2BOE22fBy6CGWPTeMqyalLCGJkSpIhVqyJAh8b7t5MmTE7M9RERERBlSoQqVcWrzBtw4dVxNGhtzjip7jQY9fxqHJwF3sW7sNOj0NRDmmB8B1/Jjbq/fULRdDjXTN1GaCCxkIrz4SMxkbUREREQZWd4SpWHr4ICgZ0/x6NYNeBQoFOvtsuXMg34zxuO8zwEcnbkHoQ5VoXUqi/NbDPDb+C0af9UZ+b1KvvXtJ0pQYLF7d9I7ERARERHRq2zt7ZG/dDlcP3FUjVrEFViYlK5eSy27li7Cjc3PoXUqDa19bWz54TKc3Jeg27gxsDWbxJgoubFxMREREVEqSocS108dj/d9GnTvhf6LPoF7vkOwD3uIcHs3vHjRAAv7z8LelUteub3Mtl23bl21cOZtStEaC+n65Ozs/MZ6C9ZYEBERESVMwQre6v/7Vy4jNOglHF0yxfu+Xb7+JrL+YtR0aK1qQ+tUAhe2hePallFo/d0guOfOp24nwUR828P7XzyHhzeuo1yTFhz9IMvXWISHh6v1kydPxllLwRqLjKtnz54oUaIERo4cmdKbkqrIpIHSPU0+N3nz5k3pzSEiolTKNbsHsufzxGP/W7h15iSK16yboPur+ovZ43Bg3Qr4rnqgJtcLta6PtV8fQqaCvug2enS8Hue+ny8O/LMYt8+dVj9rg1+iZpd3E/WaKONIUCrUlClToqby3rNnj6q5iG3ZtWtXcm0vWUCfPn3Qrl27aJetXLkSGo0Gv/zyS9RtJECMuVy9ejXOxz1z5gz+/fdffPLJJ1GXyRkR031lAsU8efKgdevWWL16dYb6XWbPnh29evXCmDFjUnpTiIgojbSdTUg6VEy12nZG77nvwS3rbtjpnkHn4I4n92phbq/JOLptEx4+fKgW6T5l7uHN61jz03dY8s0XkUHFfyeRT23ZgLCQkCS+MkrvEhRYVKhQQZ15FYUKFcKTJ0+Sa7voLZozZw7eeecdTJ8+HV988UXU5c2aNcP9+/ejLQULFozzcf744w907tz5lflMBg4cqO577do1rFq1Sp2579atGwYNGoSMpG/fvvj777/VpJJERERvqrO4efoEIiIMid5RUrj9zoTv0XxEaTjq9gFGA7RO5XF8pVH9zZfFlIny9N4dbPztRyz+8hNVPG5lZY1S9Rqh/5TZyJo7L8KCg3F62yb+0shygYWbmxtu3Lih1m/evImIiIiE3J1SoZ9++gmDBw/GsmXL1IGvORlhyJkzZ7QlriIvg8GgRj1kNCImJycndV9JAapWrRp+/PFHzJw5E7Nnz8aOHTvUbRo0aICPP/442v1kJneZgHHnzp3q5wIFCmDChAno168fMmXKhPz582PWrFnR7vPll1/Cy8tLPacEv6NGjYr60hRjx45F+fLlMW/ePHV/CYI+/PBDtf2yL2Q7PTw8MH78+FdqhsqUKaPqi/Lly6fuExQUFHW9bFPZsmURFhamftbpdCoQl1EKk1KlSiF37txYs2ZNPH4zRESUUeXyKg4HJ2eEvgzEg2txZwrEV76ixdBv3lgUrXELmtBravZuk39nTcOW6b9hwZAP4euzX11WrHpt9Jk8Dc0++AxuOXKiSrvO6vITm9YiPIwzfZOFAouOHTuqDgJy1lpSWypVqqQO3mJbMiQZTtQFp8wSYygzPuQgXIrxN27ciPbt2yfppZ89exYvXrxQ74n46N27N7JkyRKVEjVgwAAsWbIk6sBc/PXXXyp1SoIOE0nVkueQeh85uP/ggw/g6+sbdb0EHAsWLMDFixdV6p4EL7/++mu055aRk82bN2PLli1YunQp5s6di5YtW+LOnTvYu3evCny++eYbHDlyJOo+1tbW+P3333HhwgUsXLhQpfsNHz486nq5Ljg4GF999ZX6+euvv8bz58/x559/RntumaF+//7IL24iIqLY2NjawrNsBbV+/dSrs3AnVpPeA9BzZk84a/4/fcDVA3twYc8OGI0RKORdBT1//B2tPvtSjVKYSJ2Hq3sOhAa+wLld2/lLI8sUb8vZ4Q4dOqg8e8mjlxQXOZCj/4SHABNyp8zuGHkPsHeO983lwHrdunVqNMD8wN2cBBzmaU3NmzfHihUrYr3trVu31GiGnO2PDzlQl5EFGfkS8r6SEQvZpi5duqjLJEAw1XqYtGjRQgUUpsBIggap6ylWrJi6TAICExnhGDp0qBqNMQ8CZKRNRizkvStpWfXr11fBidSHyHbJY0lwIY9btWpVdZ/PPvss2uOOGzcO77//PqZNm6Yuk/0kgZAE3vK4v/32m7q/qSbJREYs4jvRJBERZVyFKlbGlcMH1HwWiS2aDgsJxouHD/DiQQCePwxQ/794GACDUVLZTccrEbC2zQ9H68JwcneNde4MCXQqt+mInXOn4diGVSjXuBlsbDk/BiUxsDDl3YsTJ07g008/ZWCRRknajtTLSDGxnEWPWRch5IBb8i9NJA0oLqGhoSp1KiEdwaRgzHR7KRyXjlJywC+BhXRPOn/+PNavX//KdpvIfSV1SYrPTP755x81eiCjEpKqpNfrXzm4l8DAPCDOkSOHCookqDC/zPxxJWVr4sSJuHz5MgIDA9XjarVahISEqLQrUb16dRXIyCiQBD21atV65TU7Ojqq+xAREb1OgXIV1f8Prl9F8PNncHbLEudtjRER6nY3zpzAE//bKnh4/vABtC8DY7+9lTWQJTKwyOxREIbghjDYueCajx7zDn2DzpOGIpObW7T7lK7XCIdXL0PQk8e4uG83yjRowl8gJT2wMJk/f75K9ZDUlEuXLkXlkEuueebMmZEh2TlFjhyk1HMngKQYSU2EBA8SLMoIRszRJwkkihQpEu+uR3LALLUFUhfxJlLT4Ofnh8qVIwvUTOlQUv8gKUny/pKRFE9Pz2j3s4sxg6gEF6ZaHx8fH1WE/u2336Jp06bqfSijFaZOV697jNc9royqtGrVSqVdSe1F1qxZceDAAfTv31+9XlNgIbc/ePCgClLi6p4lhdvu7u5v3D9ERJSxSSCRo1BRPLjup0YtStdvHO16meNC2tHKdTfOnFRpSrFxyuyGzB45kNkjp6qXkP+dsmXH4jWRJ+56/zwJl48dxLGZZ6B1KotQNMA/n65FvnrhaNp3YLRZwSu1bIe9f83D0XUrUKpeQ1hbx153SRlXogOL48ePq4M3OQMrZ7xNBa5y4LVt2zZUrBgZaacGU6dOxaRJkxAQEIBy5cqp7kWmbbYoOfuegHSklCYH7VJTYAoupOYgsaltEhAIqW0wrb+O1Ck8e/ZM1e2YSHG01E9IXYTUW8SsT3iTQ4cOqdck9Q3mKVpJJaNzEjRIgGIa1Vi+fPkrt5P3mIxoyD6Vz4YERzEL4mUUJr6TEhERUcYmbWdNgYV0aHp064Zalza0MoGe1EWY2Ds6qrqM3F4l4JYjFzKrICIH7DWOrzyunBQD/p8RULZmfZSsUgtLRo5FyNOqCHPMH+voRdnGzXFk7Qo8D7iPKz4HEjzHBqV/iQ4sPv/8c7Rp00YdBNraRj6MpIfIWWfJR9+3bx9SA0mNkVnCZ8yYofLlJfddDvokpz6+9QDpmXQ4kjlJJLiQ/SLBRczUofiQs/ASTMqZ/JiBhYxkSFAn7w8ZjZCuSFIbISMA8rzm5P0jtRYyWpLQgvKiRYvi9u3bapRCRkI2bdpkkQ5MMmojnaUkIJWuVzIqIe8nc1I3MXr0aDUKVLNmTRVkS6qg1FyYmhnIfpAgRTpbERERvUmhipVweNVSXD95DDM/6I3gZ9HblctEegXKe6NQhUrIXaykqoWIDxlZl/Rd07qpNW2vSeNx9uDuOEcvJEip2KINDi3/WwUY0j3KyiyNmMg6KSMWkkduCiqErEuRrFyXWsgBnhSZy5ljKdSVA0JJXZFcfookbWAluJCaCwkupIYgMSQokHkaYpLgM1euXChcuLAq0pZRDQn4TIXP5rp3767eR/K/1F0khAS6EvBKYCLBjYxgSLvZpJJRLnkfSUF36dKl1WuUegsTqbV49913VaG5qd2uzNEhQZPUjUjal5DCdGlxW7t27SRvExERpX85CxVVqUz6cJ0KKmwdHFC4UlU0GvARBk6dh94/T0Xdd/shX6my8Q4qTMFEkyZN1BKzjbyMXvSe+xEyZdoJm/Cg/0YvPDH/g5HQabWo0LS1Gh15fPsmrp20XMcqSh+sjDGnXIwnKW5dvHixelOa27p1q+rd/+DBA6Q0U/67nEU2n2laWp1KfYgc6JmTVqfm7U7lAFvO6Esb1Zhn8eVgUub0kNa7CT0ATq+kgFs6KknQYDoTklBSzyAByLFjx1JVOp0lyBwe0k2tR48eSXocvveIiDIO/wtncevcGeQtWRp5S5RWIwtvi/nohdCEXETl98vg+c0bOLZuJXIW8UKPcb8kqHELpT1yPCx1q7EdD1tsxKJr166qeFUOIv39/dUiKShy1lrONqcGcgZezhZLEGROfpbUnJjkLLTsONMiQQXFn9TbLFq0KGp29oSQVCP5nUi7WDkAT29BhewTGa1JLZ8NIiJKG2Q0ola3nihQtoLFggo5pywnWGV53fll0+iFs8MOWBt00DqVhM/cZ7jjex22dvYIuHoFt8+fscg2UQavsfj5559VhCqjE5I7L6SzjuTN//DDD0iLRowYoeoxYo5YUPwltjBZ6hYkdUjmtpARpvRGumaZz6VBRESUUuRknkwiazr2eV03Rwlm+kyZgB1/LcDNHTYI0+TBswf1YGO8Az0e4sia5fAs8+amLZQxJDqwkDehvCnlLL/MGSAkhcXUejO1HMxJ7mDMtCz5WeY/iEnmYZCFUiYgSWRWHhERESWzRu/2waMG/tgwciFCNTUA57bAizkqVevelUuqGxWRdWIj3YYNG6p5CCSQkDahsqSmoMIU/Hh7e6vZpU2kbaj8nNgaACIiIqKMyD13PvRb8A2yZN8L2whr2NiXUpevGv9dSm8apeXAQlKezp49i7RAUpukK5HMmyAT+UmqVnBw8CvzCxARERHRm/UY9y28u9rACdK23wo67UvM7DkK/n6+3H0ZXKKLt6W95ty5c5HaSZG51IPIHAPSgvT06dNqroaYBd1EREREFD+VG7dAzxnvwc42s/o5xBiMLRNOYtPMqdyFGViiayykYFvmgtixY4dKN5IJzcxJ3//UQuY1kIWIiIiILMPR2RndJ47DomEfIyL8CrSONXDrRBEs/HQkek/hZLAZUaIDi/Pnz0e1BL1y5Uq069jPmIiIiCj9c89fQE3ad+34ESBoK4yZuyMorBHm9hmPDpP6I4v7q81yKP1KdGCxe/duy24JERERESU7a2trVKpUKWo9qaq266ICC73VQ7joN0Nn0xRaTXWsGroJlfp6ony9RhbYakoLEvxukq5KP/74I2rWrInKlSvjq6++UjMuE8V3NvQiRYrg0KFDb2WHyUzeMoImtTVxWbBgAdzc3PA2devWDb/88stbfU4iIiJha2uLli1bqkXWkypX0WLIX7ocIgwGFKpRCNnyHIKNPgRhjgVxZFEQ1vz6M3d8BpHgwGL8+PEYOXIkXFxckCdPHjWXxUcffZQ8W0fJok+fPupgO+bSrFmzqNsUKFAAv/32W9TPMsfE0KFD1VTue/bsibpNzMfImzfva597xowZKFiwIGrUqPHKCFiLFi2QLVs21ba4ZMmS+OKLL3D37t0Eva527dohNZF9JftFZjc1JzOMy2fpxYsXKbZtREREllK1fVf1/7ld29DsvUEo3ToUDtq70Nu74v6lclgweCT04eHc4elcggOLRYsWYdq0adi6dSvWrl2LDRs24O+//1YjGZR2SBBx//79aMvSpUtjva3BYED//v3V714CAPPZtb/77rtoj3Hq1Kk4n1OCkz///FM9lrmZM2eiUaNGatLCVatW4eLFiyoAkYPutHxWX+Z7iUvp0qXVhJJ//fXXW90mIiIi+XssrfdlsdTktPlKlUH+0mVhCA/H9ll/oGabTmgxpho0oadgtLZBcHgjLBr4C54ExP+EIWWAwOL27dvqzLKJHBDKGdl79+5ZetsoGckM43Igb75kyZLllduFhYWhc+fOqvvX/v37VQcwc5kyZYr2GO7u7nE+54kTJ9Qs7TL0anLnzh188sknapEuYxK0yEhInTp1MGfOHNUmWIwdO1a1CzYnIypyW9P1MlfJunXrokZPTCMr4vr166hfv74aDSlXrhx8fHxe2T4JlIsWLQqNRoOmTZvC398/2vXy2NKwQK4vVKgQvv32W9UdzUSec/r06WjTpo3qkjZw4ED1nEL2rVwvoyomrVu3xrJly+LcX0RERMl14kta8cvyupNgCSF/4xoN/Bi2dva4ff4Mzu/ZjtwFi6L3nE/gZLUDMEYgVFMFa77chqPbNlrkOSkdBBZyICUHVjEnzLPUGzMtk6g/JDwkRRZLnXEwFxQUpIIAGUE4ePAgihUrlqTHk8DEy8tLBSMmK1asUHUXw4cPj/U+8a19kDStLl26RBuJMU+3+vrrr9VtpNZCtqF79+7RgoKQkBCVmiSjMvJaJXVJ6iDMt71Xr1749NNP1f6QURapzZD7mJMAp3379jh37pwKPGQERvj6+qptktRBkypVquDo0aMqeCMiIkrrsuTMjRpd3lHrexfPRdCzp7C1s0Pf6RPg4XkUNuFBCHP0xKl/DFj548SU3lxKBgmu2JEDWDnrKme8TbRaLd5///1oc1msXr0aGU2oPhRVl1RNkec+0uMInOyc4n37jRs3qjoZc1I7I4vJ999/r4IAmbE8rpGIL7/8UtULmEyYMEGNPsTm1q1byJ07d7TL/Pz8VN1Grly5kBTyWhwdHdVBuoycxCRBhWmkRA74S5UqhatXr6J48eLqMgmMJU2ratXI35+MfpQoUUId+EsAIPeRRgW9e/dW18uIhewfCYjGjBkT9Tw9evSINqv7jRs31P8eHh6vBEmyLySoCggIgKenZ5JePxERUWrg3bIdfH3248H1q9g1bwbafBF5XNF55Ej4bFqLC8v9EeaYDw+uV8L8j0ai71TOd5GhRyzkwEoOkjJnzhy1yCzccpBkfhmlbpKiI2fvzRcJDs01adJE5V9KsBCXYcOGRXsMOasfF+keFnO0SwLVtzHvSdmyZaPWTUHMw4cPoy6TrhjS5cxEAg4JBCSoEmfOnFH1JBLAmBZJdZJRCBntMDG174sPCYSE+f2JiIjSMmsbGzR57xP1v9/RQ7hy5GDUddVbtkObcfXgGHocsLJBiKER5vX9DqHBwSm6zZSCIxbz58+34NOnL462jmrkIKWeOyFkdEnavr5Ow4YNMXjwYLRt21YV55un8Zhkz579jY9jfltJETInaUlSpC0H6K8btZA+2zHTvRKSfifpeiamQCYhDQckLUxGLTp06PDKdebBUswZ6F/n6dOn6v/X1aUQERGlNR4FCqFK2044vPof7Jw7HflKlYWjS2QatEc+T/Sa8zkWDx6LEDREqEMtLPlgDlp+3wo5PQun9KZTEiV9VhSKdsAq6UgpsSTXWX8ZtZDOX7Nnz44zxSm+KlSogMuXL0cLEDp16gR7e3v89NNPsd7H1KZVDr4lZcj8vjHnppDHkQ5WiSH1FsePH4/6WWoi5LklHUpI0bZcJkFUzOV1kwvJNonYtktmr5f2vBJwERERpSdVO3RD1tx5EfLiuaq3MKfqLmaMR+Ys/2vvPsCjqLowAH/phYQaOoTQe0lASuihKlVUOiJIU0CqgvTeRJqiNBVQlPZTFAQpgvSOBBKahN5DSYD0ZP/n3LjrbnrYJLubfO/zDNmdnZ2dzA3JnLn3nLsPVjGRCHOujG0TDsD3MCdftnQMLLIoyUWQC3X9JTAwMMFtpfKX5GR89913GDRokFHDr+TOv5+fn25d0aJFMX/+fNUbImVo//rrL5WLIQnU/fv3V3kMQqpFPX78WAUgUllq8eLF2LFjh8H+pUKUr6+vCgDke0ltj4b0zhw/flxVr5I8otq1a6v8CiHVqSSxW3ot5PhliJRUdNLPL0mI5E5I0CfnT45fvn/9hHAJ3IiIiDIbCR6aDxgid13ht38PbvjGL0fffeZUFKrgB9t/k7qPrQjE3jUrTXK8lDYYWGRRO3fuVEOP9Jd69eolur2Pjw+2b9+uKiHJhIivU4VKJr+Tikky74m+jz/+GLt27VKT4cnrkt/Qp08fldQtSddCeg5k/hQJKKRcrCRVa1/TkpwHqVwleQ7SwyHBSUpJGVpJRJfka5lVXnIo1q1bp3tdys9KcCDHKbkYEnRIQJRc0rVMIqlN/M6fP78uMJOCB1LeVo6ZiIgoI0lPu/wtlSWpXndjFS5bHp4tWqvHu5d9jYiw0HjbtB86HBVahcE+/BEiHfLg6j43bJjFilGWykqTHnVKM4ng4GCViC45AHKRq08uDKXij8wiHTchmRInPQrNmjVTvQ5xq1JlJTLfxebNm1Wgklr82SMiIkshwcTKER/jReBjeL3ZFo0/6JfgdpdOH8fhRb4IcyoJq5gouLj+hfe/NCzpTuZ3PRwXeywoQ0l1ptmzZ+vKsGZVMvTqq6++MvVhEBERpSt7Ryc07xvbW39m52+4d+VSgtuVq14L7We/9e9M3bZ48aoJvu87HlGcJ82iMLCgDCf5C5UrV87SZ16Gehk74SAREdHrkMEqMo+SLBkxcMWjWnVUaOAjH4xdSxclGizkKVAYPZYOhFPkX+p5qE1jrOozD0FPEs4BJfPDwIKIiIgoC5HiJjNnzlRLagqdGKPR+33gnCMnnty5heOb1ye6nb2jI3p/NxnZHPYAmmiEOb2BjcM3IsDPN0OOkzJ4HgsiIiIiotRwcs0On14DsG3BLJzYsh5latdFXnePRLf/YOEMrJsyFU9vV0eYUxnsnX0ST7rdRtFSZRD6IhihwUEICQ7SexysHjtmy4a3Bn8KB2dnNpAJMLAgIiIionQnwUSpN2rjn5PHsGvJQnSZNhfW1jaIiYnGq2fPEPz4EYIDH8V+ffwItraRiIpejYjgMIQhCge+T9nnXD5yAFWatkzvb4cSwMCCiIiIiNKdzOvUpPdHuO13Hg+uXcVPnw9DRGgIXgQGIiY6KgV7sIGVlQOccjrCrYg7nLPngFP27Ko3RB7fv3oZfn/txbXTxxlYmAgDCyIiIiLKEC6586BB997YvewrPL4RoFtvbWMDl9xuyJ43L3LkzQ9Xt3zqcXb56pYXF08fx4WN4Yh0zA+biOfI5RGMpt0/MNh3obIVVGBx6/w5RIaHwc6B0wFkNAYWRERERJRhKvs0h5OLKyIjwpE9rwQO+eCSO7caFpUY79Yd4JJrH04sv4Nwx8K4ts8W4a8Wo1X/gbpt3IoWU/uTYVS3LpxDyeq1Mug7Ii1WhaIMJaXtSpUqhSNHjvDMJzAberVq1RATE8NzQ0REmXpIVOla3qhQvzGKlKuoeiSSCiq0qtRtjEbDysMh9Aai7Fxw+6QHNs37wmC/JbxqqsfXTp9I1++BEsbAIovOIyH/+eIuLVv+l+jk4eGBBQsW6J5LneuRI0eqGRf379+v2ybuPooUKZLkZy9ZskTNVu7t7a1bp//+bNmyoXTp0uoYT58+jaxEzr9MnLdmzRpTHwoREWVi1tbWqFChglrksSUpVbUGWoyrC8fQq4i2dcLDi5Wxfvo03eslq8cGFgFnTkLDG3UZzrJ+mihNL2Lv379vsPzyyy8JbhsdHY0PP/wQq1evxr59+9CoUSPda1OmTDHYx9mzZxP9TAlOvv76a7WvuH744Qf1fj8/PyxevBgvX75ErVq11GdmJRJQLVq0yNSHQUREmZitrS3ee+89tchjS1O0dFm0mdYSjiF+iLGxR+DNN7Bm7ET1WpEKlWHn6IRXz57i4fVrpj7ULIeBRRbl4OCAAgUKGCy5cuWKt114eLj6xbNnzx4cPHgQ1atXN3jd1dXVYB958+ZN9DOlB+LatWto1apVvNdy5syp3i+9IM2bN8fGjRvRrVs3DBo0CM+ePcOrV69Ub4ms17dlyxbVy/HixQvcuHFD9Xps2rQJjRs3hrOzM6pWrYqjR4/qtn/y5Am6dOmCwoULq9dlBvC4AZUEToMHD8bQoUPVOcmfPz+WL1+ujqFXr17qe5bhXDt27IgXfElvjJOTk5pVe+HChbrXw8LCULFiRfTr10+3Ts6F7Ov77/+rn9emTRucOnVKvUZEREQJy1e0GN6d3wWOIWehsbbD88B6WD1yLGzt7OBR1TP27yyHQ2U4BhaUKOk1kCDA398fhw8fVhfLxpDApEyZMupiOiWGDRumAobdu3er4KFz586qZ0OfPH/33XcN9jl27Fg1bOvvv/9WnyeBRFRUlO4CX4Kj7du348KFC+pCv0ePHjhxwnAs5qpVq+Dm5qbWS5Dx0UcfqQBLhnCdOXNGBT/yvpCQELW95EXIMLANGzao8zVhwgSMGTMG69fHzi7q6OiohjjJfrdu3aoCke7du6NZs2bo3bu37nPd3d1VICPnioiIiBKXI48bun47AI6hxwErG7x42QQrB4/RJW0HMLDIeBpKVFBQkEZOkXyNKzQ0VOPv76++asXExGiiX70yySKfnVI9e/bU2NjYaLJly2awTJ8+XbdNsWLFNPb29po8efJoHj16lOB+tNvo72PhwoWJfu6QIUM0Pj4+8dbLOd68eXOC51hemz17tnp+/Phxddz37t1Tzx8+fKixtbXV7N+/Xz2/fv262n7FihW6ffj5+al1Fy9eTPS4WrVqpRkxYoTuecOGDTX16tXTPY+KilLfW48ePXTr7t+/r/Z79OjRRPc7cOBAzTvvvGOwbs6cORo3NzfNoEGDNAULFtQEBgbGe5+np6dm0qRJie43oZ89IiKilAoPD1d/Z2SRx5YuPDRU890HkzVf99+rlhUfjdTM7dRaM7djK01w4GNTH16mvh6Oy/IG1pkxTWgoLnsZDhXKKGXPnIZVKqavl6FC3377rcG63LlzGzyXu/IyBGrGjBmYP39+gvv59NNPVV6AltzlT0xoaKi6c59SsTFHbHK3qFmzphpOJHf9R48ejZ9++gnFihVDgwYNDN5XpUoV3eOCBQuqr48ePUK5cuVUT4F8P9KTcPfuXVWlSoZ7ybCoxPZhY2ODPHnyqGFTWtKroN2vluSGyLCmW7duqe9V9i1VnvSNGDFCDd+SXBMZSiX7jUuGUml7QoiIiChp9o6OeH/Z51g9YCpC7RohLOZN2NreRFRkCALOnEDVZm/xFGYQDoXKomRokeQJ6C9xA4smTZqoYTtSyWnIkCEJ7kcCCf19SK5EYmRbyZdIqYsXL6qvkreg1adPH6xcuVI3DEpyHrSBh5ZUVtLSvqYt4frFF1+o3IdRo0apRHQZLtWiRQsVBCS2D+1+ktrv2rVr1fArybPYtWuX2q8cW9z9SiBy5coVFaxcvXo1we/76dOnSeaqEBERkSHJrXh/yXg4RRyIXWHDPAtTYI9FGrKSpN0zp0322elBei1+++03tG3bVvUgGFOxyNPTU/WSyH7iBgMJkXK3krDdtGlT3TrJS/jss8/UcUguQ8+ePVN1DJIr0q5dO7UfbWAgF/pScs8Ysl/Jv/j444916xJKwJZ8Cun5kACkb9++6nsrX7687nXJAZH3ybkiIiKiVAYXS8didb+ZeGVfDgg7jOtnTyMyLAx2qRgxQa+PgUUaUnMxpGI4kinJ8J8HDx4YrJOScwkNZZKL323btqmKRXIhLsN4XocMv5KEcCkpW6lSJYPXnj9/ro5Hjksu9JcuXaqGDEm5Wf1eEKnS1KFDBzUES4Ke5ObNiEvmyJDKUjJBn+xr3rx5ePjwodGBhexXjvWPP/5QPSw//vgjTp48adDbIkOlpEKVr68vihYtqhLIpfLVsWPHYG9vr7aRx1Kxq06dOkYdDxERUZYNLpZ9jpV9ZiPSOgc0MUFYPeozfLiQpdwzAodCZeFZniX/QH+pV69eotv7+PioC2EZhjRw4EBd/kNqSD7B22+/neAEcDJsSI5B8iCkApOLi4uqyNS1a9d428rdfhlipF9NKaXGjRsHLy8vNfxJyspKidv27dvDWP3791cBT6dOndT8G1LWVr/34tKlSyoY+uabb1RQIeRxYGAgxo8fr9tOSt9KsBE354OIiIhS03MxEnZWsX9LXzxxxC8TJ/P0ZQAryeDOiA+yRMHBwciRIweCgoLUkBx9MmTl+vXr6o50ahKSszq5Wy8lVmW4jwQPr0N6A6QU7b1793R3+jMDCTKkpK/MY6Hf0xEXf/aIiMgYcnNu5syZ6vHnn3+eqf6W6rty4ih++3I6YOUMR9feyOtxCh3HjjP1YWWq6+G42GNBGUqqLc2ePVsFZakllZIkIJk1a5bqIchsvwhlgj/pxUgqqCAiIjKWtbW1GsIrizzOrEp61VCzcEMTghjNEzy5UQMbZ8cGVJQ+Mu9PE5ktKU+rX7o1pebMmaOGSsnwJbnDktnUqFFDDaUiIiJKT5JTKUONZZHHmZWNrR2KV4udBsAq5DhibOzx+J9q2Dx/rqkPLdNiYEEWY9KkSYiMjMTevXtfexgVERERZR0lq9dUX50LRMAx5CJibBzw0K88fv1qgakPLVNiYEFEREREmVJxzxqwsrLGs7t34DPSG44hVxBt64R750ph+7JvTH14mQ4DCyIiIqIslrw9Y8YMtcSdyDWzcXLNjkJly6nHQffv4c3JTeAYeg3Rts64c7wI/tr4s6kPMVNhYEFERESUxcjQYlmyghJescOhrp05gULFS6PZ2LpwCL2FKDsXXNluC9/D+0x9iJkGAwsiIiIiyvR5FrcvnENEWCjcy1RAzf7FYR/+GBEObjix7CZuX71s6sPMFBhYEBEREVGmlbtwUeTIXwDRUVG46XtWravi3RDl2ljBNvIFwp3csWvqHgQ9CTT1oVo8BhZERERElGlZWVmhpHY41OkTuvX1O3REIa8bsI4OR5hzeWwc/h2issjwsPTCwIJMYvz48ejXrx/PfhySROfh4aFm3yYiIqK0UeLf4VDXz56CJiZGt77NwCHIVeQkoIlGmNMbWDVgGk+5ERhYZEGNGjXC0KFD461fuXIlcubMaTBvRLVq1Qy2OXjwoNpG3q/RaNQ2cicg7rJnz55EP//BgwdYuHAhxo4dazBpnva9dnZ2yJ8/P5o1a4bvv/8eMXq/ADI7mU185MiRGDVqlKkPhYiIKNMoUr4i7J2cERL0HPf/uWLwWueJE+DiHJvAHWbXED8MynyT8GYUBhaUYtu3b0eLFi0wfPhwLFiwQAUBomLFirh//77B0qBBg0T3s2LFCnh7e6NYsWIG61u2bKnee+PGDezYsQONGzfGkCFD0Lp1a0RFRWWZlurWrRsOHToEPz8/Ux8KERFlQvL3W/4Gy6L9W57ZySzcHv/Owh1w5r/hUFo958+AU8yf6nFIZBOsGTshw48xM2BgQSny888/o0OHDpgzZw4mTDD8z2Zra4sCBQoYLHLnPTFr165FmzZt4q13cHBQ7y1cuDC8vLwwZswYbN26VQUZ0psievfurQINfVIuL1++fPjuu+90PTKffPIJPvvsM+TOnVvtU3pW9M2bNw+VK1dGtmzZULRoUXz88cd4+fJlvN6bbdu2oWzZsnB2dsa7776LkJAQrFq1Sg1XypUrl/qc6Oho3ft+/PFH1KhRA66urupzu3btikePHulenzJlCgoVKoQnT57o1rVq1UoFUdqeGdlv3bp11XkiIiJKazIyQEYKyCKPs1p1KP08C33vL54Ix7CjgJU1gh7XxaZ5X2TwEVo+BhZpSIYGRYZHm2SRz04vixcvRq9evdSwpEGDBhm1r6dPn8Lf319dfKeEj48Pqlatik2bNqnnffr0wc6dO1XPhpZc/MsFf6dOnXTr5OJfgobjx4+rYEgu6Hfv3q173draGosWLVK9ArLtn3/+qQIRfbJP2UYu8OUz9+/fj7fffhu///67WiSIWLp0KTZu3GgQ5EydOhXnzp3Dli1bVO+L/OLWkuFfEpTI96E9t0eOHFHHIMekVbNmTTXsjIiIiNJG8WrV1SzcgbduIPjxfzf9tGzt7NB18VA4hvhCY22HR/7lsXdN7I1NShnbFG5HKRAVEYNlQ/4yybnqt7Ah7Bxs0ny/Fy9eVMGE9AbIEJ2EnD9/Hi4uLrrnFSpUwIkTCd8NuHXrlgqC5K59SpUrVw6+vr7qsQyhkh4EuajXBgI//PAD3nvvPYNjqFKlCiZOnKgely5dGl9//TX27t2r8jaEfo6JXOhPmzYNAwYMwDfffGMQJHz77bcoWbKkei49FvK5Dx8+VJ8l36f0NOzbt08X1EiPilaJEiVUYPLGG2+o3hB5j42NDX766SeVuzJ69Gj1ugwNc3d3N/ie5fzcvHkzxeeIiIiIUjILd3ncveSHa6ePw7Nl/NETTtmyoc30t/HbuN0IcyqBa3tdkKvgLnj5NOfpTQH2WFCSihQpooYlffHFFwa9BPrkQv/vv//WLf/73/8S3V9oaKj66ujomOIzL4GI/hhQudsvwYSQi3wZKqV/Qa8NLPQVLFjQYEiSJJc3adJEDbuSYUs9evRQw5Okl0JLhj9pgwohCeUShOgHMLJOf7+nT59Ww7wkUJD9NmzYUBdQ6Qccc+fOxezZs9G2bVs1XCouJycng2MhIiJKywqE8nddFnmclSQ3HErkK1oM9YdUgX3YA0Ta58aZ1Q8R4Bd7g5OSxh6LNGRrb616Dkz12SmVPXt2BAUFxVv//Plz5MiRw2CdXBzLRbjc6dfenZeLdH2ST1GqVKkUfbabm5v6+uzZM+TNmzfFvSbFixfXPX///ffV3f6jR4+qYUTyWv369Q3eE3fMqAQm2hwGGZ4keRofffQRpk+frvIwJFn6ww8/VL9gJaBIbB9J7ffVq1cquV2WNWvWqO9PAgp5HvcX94EDB1TvhRyLJKZLnkrcIWMpPT9ERESplVVvXknZ2QNrfsAd//OICA1RlaISUsarJp68dw++/wtCuGNh7Jt9GPm/LoZs2Q2vk8gQeyzSkLrwdLAxyZKaqg7Sw3DmzJl462VdmTJl4q2XZGIJLiQgkcToe/fuvfY5kh4A2Y/kWaSE5D7IUKt33nlHty5Pnjxo37696rWQJGvJ/0gN6VWQYODLL79E7dq11fdszPekdenSJdXrMWvWLBXoyBAu/d4MrXXr1qmcEcnZkMBDcjLiunDhAjw9PY0+JiIiIvpP7kJFkLNAQTUL941/Z+FOTJ1W7eHu/fDfCfTKYt2QhTyVyWBgkQXJnforV66oikaSu3D58mVVJemXX37BiBEjEnyPVEiS5GcJMowJLiRBuWnTpqqHIK7w8HA1x8Xdu3dVkDNjxgy0a9dO9S5IL4U+GQ4lCc/Sm9GzZ89UHYP0rkj+xFdffYWAgACVN7FkyRIYS4Y/Se+Ndr+//vprvKDhzp076vzLMKh69eqp4Ei+z2PHjhlsJ4nbzZtzPCcREVGaz8L973CogCSGQ2m92WcAsueL/Rsd6lAPq4aMYYMkgYFFFiRj/GUojtxhl4v8WrVqYf369diwYYOaSyIxMkxq165dajiT5A5IAPA6JCiQSktxJ76TyksyzEryGOQ4ZNiVJDdLyVkZNqRPjlu2lWFGqUkEF1JlSgIpubivVKmSGrY0c+ZMGEuGLkkPipxHSeyWngvJpdDPFZEKUVLxSVtdS45fAo3u3bvryt3KEC8ZqibJ4kRERJS2SnjVUl//OXUMwYGPk92+2/TJcIqOnePiVWgjlqFNgpUmPeuUWrjg4GB1MS0XeTJ8R19YWBiuX7+uxvenJhGZYi+wJZgZNmwYunTp8lqnRC7CJfFa7vjL/BqZiVSYkuBH5vFICH/2iIjIGJL3p72h9vnnnyc591RmFBMdjR9HD1FlZ/MUcUfnKXPgmO2/wiwJiYqMxKoPFyHM2RO2kS9Q9V0r1H6zLbL69XCm6bGQu9rSnaW/yB1ifTLMR8a6y4W/TIIm8xmQ6UlbLVu27LVm05ZeDslbkCFGMjxLqipltl/2MnGfBF1ERESU9qxtbPD2qAlwyZUbT+7cwq9fzkB0VGSS75E5LlpNbQ+H0FuIsnPFhbXP8eDmNTZPZgkshEx6JiVQtcvgwYMNoisZoy7T1UuyrpRUk9mX5YKWTE/mcZASr6klyc5S4lVmApcJ++JWU7J0ctdo3LhxqtwsERFRet3gk2HEsqSm+Etmkt0tH94ePQl2jk647eeLP5YsSnay4QLFSqJKNzfYRgYj3KkIto/fioiwsAw7Zktg0VdlUgq1QIECCb4m4+bl7q9cfMrFWsWKFdUcCzK2vl+/fhl+rJR2PVUcvUdERPT6pHR63759s/wpzOdRAm2Hjcam2ZNx8eA+ZHfLi3qdDYvFxFWzeWvc85+He/6VEOZcDT8Nno7ey+NXd8yqLLrHQoY+SelRKcspPRL6Q2skAbZBgwYG4wYlUVYqIMkcCgmRqkTS06G/EBEREVHm5FGtOpr3ix3xcnzzevju2Znse9oPHY5s2farx6E2jbFmzPh0PcYXTwIRERY7wbC5s9jAQkqlSmUhqRzUv39/VbLzs88+070uZUtlyIw+7XN5LSGSyCTJKdpF8jKIiIiIKPOq1LgZ6rwbW0xmz4pvEHDmZLLv6Tl/BpwiDqrHwY+9sWOF8WXrE/Iw4B+sGTsc2xfOQUxMNMydWQUWMpty3ITsuIuUSBXDhw9X8ylUqVIFAwYMUJOdyfwB0uvwuqQygmS8a5fbt2+n4XdHREREZHoyl9OCBQvUIo8JqPNuV1Rs1BQaTQx+WzALD65dTfa0dFowFI4hFxFj44Dbh91w4Wj8ObqM8c+p41g7aRRePXuKoEcPEfbihdk3lVnlWMjkbFLnP7k5GBIi5UtlKNSNGzfUzNKSe/Hw4UODbbTPE8vLcHBwUAsRERFRZiW5inIDVfuYYhPam/UdhJdPn+Cm71lsnj0ZXafNRY58CV8zimzZc6Dhp97YN+8SIhzy4/gSfxQtUw458rgZ3T5nfv8V+39cIU9QrIon2gwbDQfnbGbfVGbVYyETjJUrVy7JJbFay5KYLbM658uXTz2vU6eOmgROPxKXmaMl6JDZo4mIiIiItGxsbdFm2OfIW6w4QoKe438zJyH0ZdK9BKUqe6JkiyjYRIUizKkU/jdimZrzwpg5NvZ+vwT7Vy9XQUWVJi3x9qiJFhFUmF1gkVKSmC3dd+fOnUNAQICqACV1/2X2Ym3Q0LVrVxWEfPjhh/Dz88O6deuwcOFCNYSKiIiIiCguB2dnvD16IlzyuOHZvTvY+sVUREVEJHmifDr3QE7304AmBqGOtfHT8AmvdWLDQ0KwZc4UnNu1XbpQ0KB7bzTtO1AFPJbCIgMLGa4kidsNGzZUZWSnT5+uAgv9OSok+XrXrl1qduzq1aurYVYTJkxgqVkiIiIiSpRrbje8M3qS6iW4e8kfO76ZD01MTJJnrPOECXC2/lM9DglvhO3LvknVGQ4OfIy1Ez/D9b9Pw9beAW2HfY432nSwuHlGLDKw8PLywrFjx/D8+XOEhobC399fJV7HzY+QxO6DBw8iLCwMd+7cwahRo0x2zFnB3r17Ub58eURHZ0zVApnwUCbaS4ok+A8dOhQZJTAwUA3Hk583IiIiskxu7h5oO2IsrG1sceXoQRzbtC7Z9/RYNBmOIb7QWNvh3tE8CPDzTXHlp5/HjUDgrRtwzpETnSbOROla3rBEFhlYkHEkQV5bZUuGi5UqVUrNYq6dB2T//v0JVuSSGaGTIuV+ZRsbGxvdOpmkcM6cOahatSqcnZ3h5uaGunXr4ocffkhVJQr5/C1btsDczmP79u0N1sn39/7772PixIkmOy4iIiIynnulKmjWd6B6fGrbZkSEhiS5va2dHRp+2hD24YGIcMiLfXP2Jptv8c/JY7rKT25Fi6Hb9HkoUKqMxTYfA4ssqmXLlrh//z6uXr2qhonJ3X+ZZFCfTCYo22gXKQecmEOHDuHatWt45513DIIKmZRQJjKU2c6PHDmCEydOYODAgao0sOS+WCLpkYlJoku0V69eKu/n6dOnGXpcREREKb1ZJwVzZLG0oTYZrWLDJshdqIgKKi7s35Ps9qUqe6JgrUewiolEmFNV/DhkYqKVn05v34KtX05HVHi4qvzUecocZM8bW4TIUjGwSEPyQxIZFmaSJbXl4mTYmJTdLVasGD766CM0bdoUv/76q8E2MqRHttEuLi4uie5Pcl6aNWsGR0dH3TpJsJfKXDJESoIJGbYk5YIlsf748eMoXbq02s7Dw0Ntq0+2lWBH+7p4++231S9A7XOtH3/8Ua2TvJrOnTvjRZw6z9ITM2jQIPW69CiMHz/e4HzJ3CcjR45E4cKFkS1bNlW6WHpttFauXImcOXOq81OhQgV17nr37o1Vq1Zh69atuh4d7Xsk76dQoULYvHlzKlqEiIgoY9jZ2eHjjz9WizymxFlZW8Prrbbq8dkdv6VokrrWAwbB2f4v9Tg0qjG2fjU/3jYH1vyA/atjy8lWaWpZlZ+SYjlp5hZAIs5FPd81yWd/smoj7PQu6lPLyckJT548ee33Sy6LBAz65K69BCyenp7xtpdfZCn9ZXby5EkV5MjwKelp0R9qJb0kMkRq27ZtePbsGTp27Kh6SCShX0sCAKkOJr0lp06dUr0n7u7u6Nu3r3pdgg7J05HgSBsQyOecP39eF/yEhIRg9uzZWLFiBfLkyYOCBQuq/J7g4GB1XCJ37ty6z6xZs6Y6J/K5REREZLkq1PfBoV9W4/nD+wg4cwqlatRK9j3d50/Gqj6LEObkiYdni+LKmRMo41VTvfboRgBO/bZJPZbKTzVax944zQzYY5HFyZ37PXv24I8//oCPj4/Ba0WKFFG9FNolqcDj5s2b6qJcnwyzkrlHjCVdtUJ6DaTnRPtcyJAk6VGoVKkS6tevjx49eqgeEn1FixbF/Pnz1Rwm3bp1w+DBg9VzcevWLRUYbNiwQb2/ZMmSqveiXr16uoBBSD7IN998A29vb7Wf7Nmzq2BM2/Mji/4cK3Iu5JwQERGRZZMbt9KrIM5sT1m+p62dHZqNfRP24Q8RaZ8bBxccR0RYmHrt6MZf1NcydepbZOWnpLDHIi1PpoOD6jkw1Wenhtzhl2BBLpjl4lx6G7RDj7Tkjrurq6vueVITC8rde/1hUBk1m6cMgdI/RulJePTokcE2tWvXNvhPK5MnfvnllypXQnol5GuZMoaJUjI8SnomtCRokCpjKSVBh/RyEBERmRv52798+XL1WHrvORwqedVatMbJ3zbhtv951eOQz6NEsu9xL1MBResfwvWjuRDmXBE/DZ2K1qM/xD8nj6p5Krzf7YLMhoFFGpKLV2OGI2Wkxo0b49tvv1UXzHJ33TaByVeKFy+ueglSQnIXZCiSPrlYv3TpUrLvlRnT4wYhKa0YFfeXobRBUonVcb18+VINrTp9+rTBECuhn1MigUJq7ihI4rZ+zwoREZG5kL+5jx8/1j2m5LnmcUOZ2vVw+cgBnPl9K1p+PCxFp61l735Y5TsGL8OaIlTTGBtnTlbry9apjzxF3DPdqedQqCxKkpSlzKzkGiQUVKSW5FFInoI+6QWRYVZnz56Nt70EDq9evVKP5QJcqk5pSd6CTGwYN4B43fkxJFFcn8yBIrkTEkjIcct+pZdDzof+IsObkiJBWWLHdOHChQRzS4iIiMgyVW/VTn29dPgvvHpueDM1Kd2+mAzH0JOIiX6M0Oexw8rrvJP5eisEAwtKE1JWVkrO6pOJ6WTOiiZNmmDx4sU4d+4cAgICsH79ejU8SXIwhOR2SGUnGXolQ5N69uwZr/dAhjxJ7sSDBw/i9YwkR/Iohg8frsrn/vLLL6rU7ZAhQ3S9KpJ3IXNPbNq0SQU0kuQ9c+ZMbN++Pcn9yjH5+vqq/crEeNpeFhkCJT0gzZs3T9VxEhERkfkqWKosCpYph+ioKJzb/XuK32drZ4dWU95FzKvY6pF21oXgnOu/gi+ZCQMLShNycS7zUshFtpYkNu/evVtNnLd06VIVTLzxxhtYtGgRPvnkE5VwLWTW9IYNG6J169Zo1aqVmnROkqj1SU6E7EsSsVPbEyBBg+SASKUmKXsrQYVUhtKSJG3ZRubzkMRs+XypRCW9OUmRcamyfY0aNVSvy+HDh9V6KUEr75VkcCIiIso8qr8VOzHu37t+R1RERMrfGBmNqJi7MmgbVi4t8MvQWciMrDQcXJcoGZIjcx8EBQWpKkD6wsLC1N1tyUOIm7ScVX366afqnEkQkZVJACWBU9zyu2mFP3tERGQMmcBWeua1N/f0qxpS0mKio7FicB+8ePIYLQYMQaXGzVJ0yjbNmoTrZ0/BwTkXrBx6AZoY5HU/jo5jx1r09XBc7LGgNDN27Fg14V5qkqczGxkS1aFDB3TpkjnHThIREWVl1pKf2bK1enz6960pSn6/f/WyCipksr1uM2bBMUyqQlnjWUAl+B7eh8yEgQWlGakgNWbMGFXlKauS6lgy9Csz1aQmIqLMRf5GyR1oWfj3KvUq+7SAnYMjAm/dwG0/32S3P7LxZ/W1Qv3GyFWwMNrP+gAOoXcQZeeKk0svICqFlTAtQda9AiQiIiLKgqTSohRYkYVzWKSeo4sLKjZqouu1SMq9Kxdx4+/TqreidofOal2eAoXh0SwGVjGRsfNbDJ+IzIKBBRERERFRKni2bKu+Bpw5iWf3JSk7YUc2xPZWVGzYBDkLFNStb9r9AzjZ/qUeh4Y1wP71azLF+WdgQURERESUCrkLFUYJrzdkhkGc2fFbgtvcveSPm75nVV5G7Q6d4r3ebd4EOIZcQoyNPf7ZHokXz59bfBswsCAiIiLKQmTepeXLl6tFOwcTpZ7XW7ET5vnt34OwVy/jvX5kwxpdb0WOfPEn3bV3dITnB6VgE/UK4U7u2DBynsU3AwMLIiIioixEKhndu3dPLZx14PW5V6oKN3cPRIaH4fyfuwxeu3PxAm5dOAdrG1vUejt+b4WWl09zuOQ9oR6H2jTAlgWWHVwwsCAiIiIiSiWpqOX1Vmyuxdmdv6k5LuLmVlRq3BQ58uVPcj/dZ06FY+hJVYL28bkiuH31v8mGLQ0DCyIiIiKi11C+biM4Zc+BF4GPcfXEUbXutv95VYY2treiY4r20+TzVrALf4IIBzfsmrbJYtuCgUUWjbCTWiZNmoSsYvHixfDw8FCzp9eqVQsnTsR2R6bErFmz1PmScn1ERESU9dja26Nqs7fU4zP/lp7V5lZU9mmO7G75UrQfj3KV4Fb5hpqRO8ypFtaMnQBLxMAiC7p//75uWbBggZqeXX/dyJEj470nIiICmc26deswfPhwTJw4EWfOnEHVqlXRokULPHr0KNn3njx5EkuXLkWVKlUy5FiJiIjIPFVr/pbqnbh35SJObN2IO/4XYGOb8t4KrQ7DP4VT1EH1+OWDGjh38E9YGgYWWVCBAgV0i3bWTf11Li4uaNSoEQYNGqTuxsts0nLBLeTuvgQj+qpVq6br5YiJicHMmTNRvHhxODk5qYv1jRs3JntM0lMgnynvKVeuHE6dOoVly5ahbdvYsYvpYd68eejbty969eqFChUqYMmSJXB2dsb333+f5PtevnyJbt26qWoauXLlSrfjIyIiIvOXLWculKvbQD0++PNK9bVyk5ZwzeOW6n29M2cwHEJvI8rOBaeW+1vcrNwMLNKB3N1PbImKikrxtnFLwCW2XXpZtWoV7O3tcfjwYXXRnRISVKxevVpt7+fnh2HDhqF79+7466/YSWAScuzYMTRs2BCtWrWCr68vypcvjylTpmD27NmYPHlyou+bMWOGCoKSWm7dupXge+W8nT59Gk2bNtWts7a2Vs+PHo0dI5mYgQMHqmPVfy8REZElkRtpslDalp4VNnZ2qNX+PbyOHHncULy5BtbREQhzroCfhlnWrNy2pj6AzEgurhNTunRpdO3aVfd87ty5idaQLlasGD744APd84ULFyIkJCTedjKUJz3Isc6ZMyfF24eHh6uL/T179qBOnTpqXYkSJXDo0CE1bEiCh4TIcKT33nsPn376qXrepUsXtbRr1w6enp6Jft6AAQPQsWPS3YyFChVKcH1gYCCio6ORP79hpQZ5funSpUT3t3btWjVsSoZCERERWSK5aaj9m0tpI3/xkihSoZIaBlWlaUu45M7z2vtq0u0D3DoyBiHRTREa3gB//rIaPl3et4imYmBBiapevXqqzs4///yjAp9mzZrF6x1ILEC4c+eO6iGQAEv3Q2lrq+pqJ9VbIXLnzq2WjHL79m0MGTIEu3fvVsneRERERFpvfjwcl48dQrV/k7mN0e3LCfix33cIcy6PgJ0xqN48UPVmmDsGFung888/T/Q1GW6jL6FEaS3JfdAnF7UZKVu2bAkef9zJdLQ9LpJ7ILZv347ChQsbbOPg4JDgZ1y8eFF99fLy0q27fPkyatasicqVKyd5fNI7IktS/P394e7uHm+95I3Y2Njg4cOHBuvlueSZJESGTklit/6xSq/HgQMH8PXXX6seG9knERERZT3Z8+bDG206pMm+ZFbu6r3L4tiPL9Ws3P8btQi9V0yBuWNgkU5djKbeNr3kzZtXVY7SCg4OxvXr19VjSYCWAELyGhIb9hRXUFCQuhjXBlFPnz5VvReS9J0cY4ZCybmUHpm9e/eiffv2usRzeS5J6wlp0qQJzp8/b7BOEr8l2XzUqFEMKoiIyCLIDcE1a2JLokoxEjs7O1MfEiWgWqOm8PtjPJ4/a4xwK2/4HvkLVbxTdn1lKgwsKFV8fHywcuVKtGnTBjlz5sSECRN0F9Surq6qB0YStuUivV69eipwkORvKWnbs2fPePuTilJy119yOSTPQnplpPKU9DTcvHlT5Zmk11Aoye2QY6pRo4bqIZFqV69evVLBgpb0RGzevFkFHPL9VapUKV6vTp48eeKtJyIiMlcy8kD+xmofk/nqNnMqvu87DoW9i6GKd1+YOwYWlOphXtJD0bp1a1WqdurUqboeCyHPpVdDEtgDAgJU8CFDh8aMGZPg/kqVKqUqQEliugxr6ty5M37++Wc0b94cLVu21A2VSg+dOnXC48ePVXD04MEDFeTs3LnTIKFbkryvXbuWbsdARERElJTey6fBUlhpGKomSob5yMWz3HWXO+76wsLC1AW1zNfARF7KSPzZIyIiY0hRFW0FS7lhaA5Drckyr4fj4jwWRERERERkNAYWRERERERkNAYWRERERERkNCZvExEREWUxLDFL6YGBBREREVEWIsnaiVVrJDIGh0IZiUW1KKPxZ46IiIjMEQMLI7sQQ0JC0rI9iFJUJlBoJyYkIiIiMgccCvWa5KJOJn979OiReu7s7AwrK6u0bBuieGRGc5nUT37ebG3535eIiFIvKioK69evV487duzIvyeUZnhlYoQCBQqor9rggigjWFtbw93dnYEsERG99k2qq1ev6h4TpRUGFkaQHoqCBQsiX758iIyMTLNGIUou6U6CCyIiIiJzwsAijYZFcbw7EREREWVlvO1JRERERERGY2BBRERERERGY2BBRERERERGY45FCiYiCw4ONv5MExEREZnJfEhhYWG6axwpCkKUGO11cEom6LXScBrfRN25cwdFixZN9iQSEREREWVmt2/fRpEiRZLchoFFEqS287179+Dq6mqyOQMkSpTgRhoze/bsJjkGShm2leVgW1kOtpXlYFtZFraX5Qg28bWg9EG8ePEChQoVSrbcPYdCJUFOXnKRWUaRHyQGFpaBbWU52FaWg21lOdhWloXtZTmym/BaMEeOHCnajsnbRERERERkNAYWRERERERkNAYWZs7BwQETJ05UX8m8sa0sB9vKcrCtLAfbyrKwvSyHgwVdCzJ5m4iIiIiIjMYeCyIiIiIiMhoDCyIiIiIiMhoDCyIiIiIiMhoDCzPz9OlTdOvWTdUpzpkzJz788EO8fPky2fcdPXoUPj4+yJYtm3pvgwYNEBoamiHHnFW9bltpJ5t588031cSLW7ZsSfdjpdS3l2w/ePBglC1bFk5OTnB3d8cnn3yCoKAgns40tnjxYnh4eMDR0RG1atXCiRMnktx+w4YNKFeunNq+cuXK+P3339kmZthWy5cvR/369ZErVy61NG3aNNm2JdO1l761a9eqv0/t27dnk5hpWz1//hwDBw5EwYIFVVJ3mTJlzOJ3IQMLMyMXPn5+fti9eze2bduGAwcOoF+/fskGFS1btkTz5s3VD+LJkycxaNCgZGdHpIxvK60FCxaYbDb3rCq17XXv3j21zJ07FxcuXMDKlSuxc+dOFZBQ2lm3bh2GDx+uKp6cOXMGVatWRYsWLfDo0aMEtz9y5Ai6dOmi2uHs2bPqwkcWaSMyr7bav3+/aqt9+/apv1Myc7D8nbp79y6bygzbS+vGjRsYOXKkCgrJPNsqIiICzZo1U221ceNGXL58WQXyhQsXNn2Tachs+Pv7a6RJTp48qVu3Y8cOjZWVlebu3buJvq9WrVqacePGZdBRkjFtJc6ePaspXLiw5v79+2ofmzdv5kk14/bSt379eo29vb0mMjIynY4066lZs6Zm4MCBuufR0dGaQoUKaWbOnJng9h07dtS0atUq3u/A/v37p/uxZnWpbau4oqKiNK6urppVq1al41GSMe0lbeTt7a1ZsWKFpmfPnpp27drxhJphW3377beaEiVKaCIiIsyufXhL24zIHR0ZolGjRg3dOuk6lp6H48ePJ/geiWbltXz58sHb2xv58+dHw4YNcejQoQw88qznddpKhISEoGvXrqrLs0CBAhl0tPS67RWXDIOSoVS2trY8qWlA7rqdPn1atYWWtIk8lzZLiKzX317Inb3EtifTtVVCv/8iIyORO3duNouZtteUKVPU9QR7Zs27rX799VfUqVNHDYWS675KlSphxowZiI6OhqkxsDAjDx48UP+h9ckFjPwSltcSEhAQoL5OmjQJffv2VUM1vLy80KRJE1y9ejVDjjsrep22EsOGDVMBYLt27TLgKMnY9tIXGBiIqVOnpni4G6XsnMofQvnDqE+eJ9Yusj4125Pp2iquUaNGoVChQvECQzKP9pIbkt99950aUkPm3VYBAQFqCJS8T/Iqxo8fjy+//BLTpk2DqTGwyACjR49W4+mTWi5duvRa+46JiVFf+/fvj169esHT0xPz589XCafff/99Gn8nmV96tpXcYfjzzz9VfgWZf3vpCw4ORqtWrVChQgUVxBNR6syaNUslBG/evFklp5J5efHiBXr06KGCCjc3N1MfDqXg2k9uli1btgzVq1dHp06dMHbsWCxZsgSmxv78DDBixAh88MEHSW5TokQJNTQmbqJOVFSUqk6T2LAZqQYg5IJHX/ny5XHr1i2jjz2rSc+2kqDi2rVrakiOvnfeeUclyUmiI5lPe+n/wZXiCK6uruqiyM7Ojs2URuQCxsbGBg8fPjRYL88TaxdZn5rtyXRtpSUFECSw2LNnD6pUqcImMcP2kr9Nkgjcpk2beDcupXdXkoNLliyZAUee9bi9xv8tufaTv0XyPv3rPunhkKFV9vb2MBUGFhkgb968akmOjJeT8mEy1k4iUO3FqPznltJjCZHSZNK1LP/p9V25ckWVMyXzaSu5u96nTx+DdVIqU3qY9H+Zk3m0l7anQsbvSyk/6XHinda0JX/8pD327t2rK2spbSLPpbJdYm0prw8dOlS3Tip9yXoyr7YSc+bMwfTp0/HHH38Y5DiRebWXlG8+f/68wbpx48apGysLFy5UFb3IPNpK1K1bFz///LPaTlsBVK77JOAwZVChmDp7nAy1bNlS4+npqTl+/Ljm0KFDmtKlS2u6dOmie/3OnTuasmXLqte15s+fr8mePbtmw4YNmqtXr6oKUY6Ojpp//vmHp9fM2iouVoUy3/YKCgpS1YYqV66s/i9JFS/tIpVTKG2sXbtW4+DgoFm5cqWq3tWvXz9Nzpw5NQ8ePFCv9+jRQzN69Gjd9ocPH9bY2tpq5s6dq7l48aJm4sSJGjs7O8358+fZJGbWVrNmzVJV1DZu3Gjw/+fFixdsKzNsr7hYFcp82+rWrVuqwtqgQYM0ly9f1mzbtk2TL18+zbRp0zSmxsDCzDx58kRd7Li4uKhgoVevXga/hK9fv64uRvft22fwPilJVqRIEY2zs7OmTp06moMHD5rg6LOW120rfQwszLe95Ks8T2iRbSntfPXVVxp3d3d1ESplF48dO6Z7rWHDhuoCJ27Z3zJlyqjtK1asqNm+fTubwwzbqlixYgn+/5FgkMyvveJiYGHebXXkyBF180sCEik9O336dLO46WUl/5i2z4SIiIiIiCwdq0IREREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEREREZHRGFgQEZFZOnHiBBo1agQnJyeUK1cOp06dwrJly9C2bVtTHxoRESXASqPRaBJ6gYiIyFSOHTuGxo0bY8qUKWjfvj0+++wzREdHw8/PDxs3boSnpycbh4jIzDCwICIis+Pt7Y1SpUph9erV6vn69evRpUsXtGvXDps2bTL14RERUQI4FIqIiMzKnTt3cPToUQwYMEC3ztbWFtLBPnnyZJMeGxERJY6BBRERmZWLFy+qr15eXrp1ly9fRs2aNVG5cmUTHhkRESWFgQUREZmVoKAg2NjYwMrKSj1/+vQp5s6dC2dnZ1MfGhERJYGBBRERmZVq1aqpRO05c+bg0qVLKrfCw8MD/v7+uHnzpqkPj4iIEsHAgoiIzIokbUs1qIULF6rqT4UKFcKuXbtQuHBhtGzZ0tSHR0REiWBVKCIiIiIiMhp7LIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiIyGgMLIiIiIiICMb6P+l63PaGNt48AAAAAElFTkSuQmCC", + "image/png": "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", "text/plain": [ "
" ] @@ -499,8 +510,9 @@ "ax.plot(np.array(alpha_grid), np.array(profile_kf_dynamax), label=\"KF (Dynamax)\", color=\"C1\")\n", "ax.plot(np.array(alpha_grid), np.array(profile_ekf_cb), label=\"EKF (Cuthbert)\", color=\"C2\")\n", "ax.plot(np.array(alpha_grid), np.array(profile_ekf_dynamax), label=\"EKF (Dynamax)\", color=\"C3\")\n", - "ax.plot(np.array(alpha_grid), np.array(profile_ukf_dynamax), label=\"UKF (Dynamax)\", color=\"C4\")\n", - "ax.plot(np.array(alpha_grid), np.array(profile_pf_cb), label=\"PF (Cuthbert)\", color=\"C5\")\n", + "ax.plot(np.array(alpha_grid), np.array(profile_enkf_cb), label=\"EnKF (Cuthbert)\", color=\"C4\")\n", + "ax.plot(np.array(alpha_grid), np.array(profile_ukf_dynamax), label=\"UKF (Dynamax)\", color=\"C5\")\n", + "ax.plot(np.array(alpha_grid), np.array(profile_pf_cb), label=\"PF (Cuthbert)\", color=\"C6\")\n", "ax.axvline(\n", " true_alpha, color=\"gray\", linestyle=\"--\", label=f\"True $\\\\alpha$ = {true_alpha}\"\n", ")\n", @@ -508,7 +520,7 @@ "ax.set_ylabel(\"Profile Score $\\\\nabla_\\\\alpha \\\\log p(y_{1:T} \\\\mid \\\\alpha)$\")\n", "ax.legend()\n", "ax.set_title(\n", - " \"Profile Score: KF vs EKF vs PF (2 states, 1 control, 1 output)\"\n", + " \"Profile Score: KF vs EnKF vs EKF vs UKF vs PF (2 states, 1 control, 1 output)\"\n", ")\n", "plt.tight_layout()\n", "plt.show()" @@ -517,7 +529,7 @@ ], "metadata": { "kernelspec": { - "display_name": ".venv (3.12.11)", + "display_name": ".venv", "language": "python", "name": "python3" }, @@ -531,7 +543,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.12.11" + "version": "3.12.10" } }, "nbformat": 4, diff --git a/docs/faq.md b/docs/faq.md index 26186bc2..27796419 100644 --- a/docs/faq.md +++ b/docs/faq.md @@ -24,13 +24,13 @@ with Filter(filter_config=HMMConfig()): return model(obs_times=obs_times, obs_values=obs_values) ``` -- **Discrete-time**: Either a **Simulator** (NUTS samples both parameters and latent states) or a **Filter** (pseudo-marginal MCMC—parameters only). Note: the usage of discrete-time filters is currently under active development (likely incorrect implementations). +- **Discrete-time**: Either a **Simulator** (NUTS samples both parameters and latent states) or a **Filter** (parameters only, with latent states marginalized by a filtering algorithm). `Filter()` defaults to the cuthbert-backed EnKF for Gaussian observation models. Use `PFConfig` when you need non-Gaussian observations or a fully particle-based approximation. For explicit representation of latent states (NUTS / SVI do all the work of parameter and latent state inference), use the simulator approach (currently working reliably), do: ```python with DiscreteTimeSimulator(): return model(obs_times=obs_times, obs_values=obs_values) ``` -For filter-based marginalization (currently not working reliably), do: +For filter-based marginalization with the default EnKF, do: ```python with Filter(): return model(obs_times=obs_times, obs_values=obs_values) @@ -97,7 +97,7 @@ Hierarchical models are supported by the [`dsx.plate`](./api_reference/public/ha ## What about neural nets? -We will put examples up soon. See [CD-Dynamax's Lorenz 63 neural drift tutorial](https://github.com/hd-UQ/cd_dynamax/blob/dev-numpyro-api/demos/numpyro/notebooks/lorenz63_nndrift_sgd_fit_to_data_tutorial_newAPI.ipynb) to convince yourself that this will work well. +See our [neural drift correction deep dive](deep_dives/neural_drift_correction.ipynb). It shows how to keep the observation model and filter workflow fixed while swapping a mechanistic ODE drift for a mechanistic-plus-neural residual drift. ## What about SINDy? diff --git a/docs/index.md b/docs/index.md index 451ba56b..1310cd5f 100644 --- a/docs/index.md +++ b/docs/index.md @@ -107,7 +107,7 @@ Other JAX-based libraries for dynamical systems: - **[dynamax](https://github.com/probml/dynamax)** — Discrete-time state space models with linear/non-linear Kalman filters and Bayesian parameter estimation - **[cd-dynamax](https://github.com/hd-UQ/cd_dynamax)** — Continuous-discrete state space models with EnKF, EKF, UKF, PF and Bayesian parameter estimation - **[PFJax](https://pfjax.readthedocs.io/en/latest/)** — Nonlinear and non-Gaussian discrete-time models with particle filters and particle MCMC -- **[Cuthbert](https://state-space-models.github.io/cuthbert/)** — Discrete-time state space models with linear/non-linear Kalman (and Particle Filters) filters, options for associative scans. +- **[Cuthbert](https://state-space-models.github.io/cuthbert/)** — Discrete-time state space models with linear/non-linear Kalman, ensemble Kalman, and particle filters, plus options for associative scans. - **[diffrax](https://docs.kidger.site/diffrax/)** - Numerical differential equation solvers. Other probabilistic programming languages with support for dynamical systems: diff --git a/dynestyx/inference/filter_configs.py b/dynestyx/inference/filter_configs.py index 6b912d0c..dff99708 100644 --- a/dynestyx/inference/filter_configs.py +++ b/dynestyx/inference/filter_configs.py @@ -84,7 +84,8 @@ class BaseFilterConfig: class EnKFConfig(BaseFilterConfig): r"""Ensemble Kalman Filter (EnKF) for discrete-time models. - A good general-purpose filter for nonlinear models. Works with any + The **default filter** for discrete-time models. A good general-purpose + filter for nonlinear models with Gaussian observations. Works with any differentiable or non-differentiable dynamics and scales well to moderate state dimensions. Cheaper per-step than the particle filter, but assumes observations are approximately Gaussian given the ensemble. @@ -108,7 +109,7 @@ class EnKFConfig(BaseFilterConfig): inflation_delta (float | None): Scale ensemble anomalies by \(\sqrt{1 + \delta}\) before the update to prevent collapse. `None` disables inflation. - filter_source (FilterSource): Backend. Defaults to `"cd_dynamax"`. + filter_source (FilterSource): Backend. Defaults to `"cuthbert"`. ??? note "Algorithm Reference" The ensemble Kalman filter comprises ensemble members $x_t^{(i)}, i = 1, \ldots, N_{\text{particles}}$. @@ -159,7 +160,7 @@ class EnKFConfig(BaseFilterConfig): ) perturb_measurements: bool | None = None inflation_delta: float | None = None - filter_source: CDDynamaxOnlyFilterSource = "cd_dynamax" + filter_source: CuthbertOnlyFilterSource = "cuthbert" @dataclasses.dataclass @@ -282,9 +283,6 @@ class EKFConfig(BaseFilterConfig): This is exact (but wasteful) for linear-Gaussian models. - This is the **default discrete-time filter** when no `filter_config` is - passed to `Filter`. - Attributes: filter_emission_order (FilterEmissionOrder): Linearisation order for the observation function. `"first"` *(default)* is the standard @@ -520,7 +518,7 @@ class ContinuousTimeEnKFConfig(EnKFConfig, ContinuousTimeConfig): [Available Online](https://epubs.siam.org/doi/abs/10.1137/21M1434477). """ - filter_source: CDDynamaxOnlyFilterSource = "cd_dynamax" + filter_source: CDDynamaxOnlyFilterSource = "cd_dynamax" # type: ignore[assignment] @dataclasses.dataclass diff --git a/dynestyx/inference/filters.py b/dynestyx/inference/filters.py index c1c4d698..56b6ffe3 100644 --- a/dynestyx/inference/filters.py +++ b/dynestyx/inference/filters.py @@ -47,6 +47,7 @@ ) from dynestyx.inference.integrations.utils import ( WeightedParticles, + covariance_from_cholesky, particles_to_delta_mixtures, ) from dynestyx.models import DynamicalModel @@ -130,7 +131,7 @@ def _cuthbert_states_to_dists( chol_cov = states.chol_cov[ (slice(None),) * len(plate_shapes) + (slice(1, None), ...) ] - cov = jnp.matmul(chol_cov, jnp.swapaxes(chol_cov, -1, -2)) + cov = covariance_from_cholesky(chol_cov) t_len = _time_len_from_array(mean, plate_shapes) return [ numpyro.distributions.MultivariateNormal( @@ -297,8 +298,7 @@ def _default_filter_config(dynamics: DynamicalModel): if dynamics.continuous_time: return ContinuousTimeEnKFConfig() - # default to particle filter in discrete time - return EKFConfig(filter_source="cuthbert") + return EnKFConfig() @dataclasses.dataclass @@ -342,7 +342,7 @@ class Filter(BaseLogFactorAdder): If `filter_config=None`, defaults are: - `ContinuousTimeEnKFConfig()` for continuous-time models, and - - `EKFConfig(filter_source="cuthbert")` for discrete-time models. + - `EnKFConfig()` for discrete-time models. Notes: - If your latent state is *discrete* (an HMM), you must use `HMMConfig`. @@ -643,8 +643,8 @@ def _filter_discrete_time( ) -> list[numpyro.distributions.Distribution]: """Discrete-time marginal likelihood via cuthbert or cd-dynamax. - Filter type inferred from config class: KFConfig, EKFConfig, UKFConfig (cd-dynamax) - or EKFConfig (cuthbert), PFConfig (cuthbert). + Filter type inferred from config class: KFConfig, EKFConfig, UKFConfig + (cd-dynamax) or KFConfig, EKFConfig, EnKFConfig, PFConfig (cuthbert). Args: name: Name of the factor. diff --git a/dynestyx/inference/integrations/cuthbert/discrete.py b/dynestyx/inference/integrations/cuthbert/discrete.py index 3da33b64..ea15e4ea 100644 --- a/dynestyx/inference/integrations/cuthbert/discrete.py +++ b/dynestyx/inference/integrations/cuthbert/discrete.py @@ -5,6 +5,7 @@ import numpyro import numpyro.distributions as dist from cuthbert import filter as cuthbert_filter +from cuthbert.enkf import ensemble_kalman_filter from cuthbert.gaussian import kalman, taylor from cuthbert.smc import particle_filter from cuthbertlib.resampling import ( @@ -17,13 +18,18 @@ from dynestyx.inference.filter_configs import ( BaseFilterConfig, EKFConfig, + EnKFConfig, KFConfig, PFConfig, _config_to_record_kwargs, ) -from dynestyx.inference.integrations.utils import particles_to_delta_mixtures +from dynestyx.inference.integrations.utils import ( + covariance_from_cholesky, + particles_to_delta_mixtures, +) from dynestyx.models import ( DynamicalModel, + GaussianObservation, LinearGaussianObservation, LinearGaussianStateEvolution, ) @@ -51,6 +57,16 @@ def _config_to_filter_kwargs(config: BaseFilterConfig) -> dict: kwargs["resampling_differential_method"] = ( config.resampling_method.differential_method ) + elif isinstance(config, EnKFConfig): + kwargs["n_particles"] = config.n_particles + kwargs["inflation"] = ( + config.inflation_delta if config.inflation_delta is not None else 0.0 + ) + kwargs["perturbed_obs"] = ( + config.perturb_measurements + if config.perturb_measurements is not None + else True + ) return kwargs @@ -106,6 +122,13 @@ def compute_cuthbert_filter( "or run inside a NumPyro seeded context (e.g., with numpyro.handlers.seed)." ) filter_obj = _cuthbert_filter_pf(dynamics, filter_kwargs) + elif isinstance(filter_config, EnKFConfig): + if key is None: + raise ValueError( + "Ensemble Kalman filter requires a PRNG key: set 'crn_seed' in the filter config, " + "or run inside a NumPyro seeded context (e.g., with numpyro.handlers.seed)." + ) + filter_obj = _cuthbert_filter_enkf(dynamics, filter_kwargs) elif isinstance(filter_config, KFConfig): filter_obj = _cuthbert_filter_kalman(dynamics, filter_kwargs) elif isinstance(filter_config, EKFConfig): @@ -113,7 +136,7 @@ def compute_cuthbert_filter( else: raise ValueError( f"Unsupported cuthbert config: {type(filter_config).__name__}. " - "Expected KFConfig, EKFConfig, PFConfig." + "Expected KFConfig, EKFConfig, EnKFConfig, PFConfig." ) states = cuthbert_filter(filter_obj, cuthbert_inputs, parallel=False, key=key) @@ -133,7 +156,7 @@ def run_discrete_filter( ctrl_values=None, **kwargs, ) -> list[dist.Distribution]: - """Run discrete-time filter via cuthbert (Kalman, Taylor KF, particle filter). + """Run discrete-time filter via cuthbert (Kalman, Taylor KF, EnKF, PF). Returns: list[dist.Distribution]: Filtered state distributions at each obs time. @@ -158,17 +181,18 @@ def run_discrete_filter( if isinstance(filter_config, PFConfig): _add_sites_pf(name, states, record_kwargs) - particles = states.particles + particles = states.particles[1:] if particles.ndim == 2: particles = particles[..., None] - return particles_to_delta_mixtures(particles, states.log_weights) + return particles_to_delta_mixtures(particles, states.log_weights[1:]) else: - _add_sites_taylor_kf(name, states, record_kwargs) - chol_t = jnp.transpose(states.chol_cov, (0, 2, 1)) - cov = jnp.matmul(states.chol_cov, chol_t) + _add_sites_gaussian_filter(name, states, record_kwargs) + mean = states.mean[1:] + chol_cov = states.chol_cov[1:] + cov = covariance_from_cholesky(chol_cov) return [ - dist.MultivariateNormal(states.mean[i], covariance_matrix=cov[i]) - for i in range(states.mean.shape[0]) + dist.MultivariateNormal(mean[i], covariance_matrix=cov[i]) + for i in range(mean.shape[0]) ] @@ -231,6 +255,111 @@ def log_potential(x_prev, x, mi: CuthbertInputs): return pf +def _cuthbert_filter_enkf(dynamics: DynamicalModel, filter_kwargs: dict | None = None): + if filter_kwargs is None: + filter_kwargs = {} + + state_dim = dynamics.state_dim + obs_dim = dynamics.observation_dim + + def init_sample(key, mi: CuthbertInputs): + return jnp.atleast_1d(jnp.asarray(dynamics.initial_condition.sample(key))) + + def get_dynamics(mi: CuthbertInputs): + def dynamics_fn(x, key): + def _noop(key): + return x + + def _evolve(key): + d = dynamics.state_evolution(x, mi.u_prev, mi.time_prev, mi.time) # type: ignore + return jnp.atleast_1d(jnp.asarray(d.sample(key))) # type: ignore + + return jax.lax.cond(mi.is_first_step, _noop, _evolve, key) + + return dynamics_fn + + def get_observations(mi: CuthbertInputs): + obs_model = dynamics.observation_model + y = jnp.atleast_1d(jnp.asarray(mi.y)) + + if isinstance(obs_model, LinearGaussianObservation): + H = jnp.asarray(obs_model.H) + chol_R = jnp.linalg.cholesky(jnp.atleast_2d(jnp.asarray(obs_model.R))) + bias = ( + jnp.zeros((obs_dim,), dtype=y.dtype) + if obs_model.bias is None + else jnp.atleast_1d(jnp.asarray(obs_model.bias)) + ) + D = None if obs_model.D is None else jnp.asarray(obs_model.D) + + def observation_fn(x): + loc = H @ x + bias + if D is not None: + loc = loc + D @ jnp.atleast_1d(jnp.asarray(mi.u)) + return jnp.atleast_1d(jnp.asarray(loc)) + + return observation_fn, chol_R, y + elif isinstance(obs_model, GaussianObservation): + chol_R = jnp.linalg.cholesky(jnp.atleast_2d(jnp.asarray(obs_model.R))) + + def observation_fn(x): + return jnp.atleast_1d(jnp.asarray(obs_model.h(x, mi.u, mi.time))) + + return observation_fn, chol_R, y + else: + probe_x = jnp.zeros((state_dim,), dtype=y.dtype) + probe_dist = obs_model(probe_x, mi.u, mi.time) + + if isinstance(probe_dist, dist.MultivariateNormal): + chol_R = jnp.asarray(probe_dist.scale_tril) + elif isinstance(probe_dist, dist.Normal): + scale = jnp.atleast_1d(jnp.asarray(probe_dist.scale)) + if scale.size == 1 and obs_dim > 1: + scale = jnp.full((obs_dim,), scale[0]) + chol_R = jnp.diag(scale) + elif isinstance(probe_dist, dist.Independent) and isinstance( + probe_dist.base_dist, dist.Normal + ): + scale = jnp.atleast_1d(jnp.asarray(probe_dist.base_dist.scale)) + if scale.size == 1 and obs_dim > 1: + scale = jnp.full((obs_dim,), scale[0]) + chol_R = jnp.diag(scale) + else: + raise TypeError( + "cuthbert EnKF requires Gaussian observation distributions. " + "Expected LinearGaussianObservation, GaussianObservation, or a " + "callable returning Normal, Independent(Normal), or " + f"MultivariateNormal; got {type(probe_dist).__name__}." + ) + + def observation_fn(x): + edist = obs_model(x, mi.u, mi.time) + if not ( + isinstance(edist, (dist.MultivariateNormal, dist.Normal)) + or ( + isinstance(edist, dist.Independent) + and isinstance(edist.base_dist, dist.Normal) + ) + ): + raise TypeError( + "cuthbert EnKF observation callable must keep returning " + "Gaussian distributions; got " + f"{type(edist).__name__}." + ) + return jnp.atleast_1d(jnp.asarray(edist.mean)) + + return observation_fn, chol_R, y + + return ensemble_kalman_filter.build_filter( + init_sample=init_sample, # type: ignore + get_dynamics=get_dynamics, # type: ignore + get_observations=get_observations, # type: ignore + n_particles=int(filter_kwargs.get("n_particles", 30)), + inflation=float(filter_kwargs.get("inflation", 0.0)), + perturbed_obs=bool(filter_kwargs.get("perturbed_obs", True)), + ) + + def _cuthbert_filter_kalman( dynamics: DynamicalModel, filter_kwargs: dict | None = None ): @@ -349,7 +478,7 @@ def dynamics_log_density(x_prev, x): ) try: x_lin = jnp.atleast_1d(jnp.asarray(dist_at_lin.mean)) # type: ignore - except Exception as exc: + except (AttributeError, NotImplementedError) as exc: raise ValueError( "dist_at_lin.mean is not available. Linearized Kalman filter requires a mean-able distribution." ) from exc @@ -439,8 +568,10 @@ def _add_sites_pf( numpyro.deterministic(f"{name}_filtered_states_cov_diag", diag_cov) -def _add_sites_taylor_kf( - name: str, states: taylor.LinearizedKalmanFilterState, record_kwargs: dict +def _add_sites_gaussian_filter( + name: str, + states: taylor.LinearizedKalmanFilterState | ensemble_kalman_filter.EnKFState, + record_kwargs: dict, ): max_elems = record_kwargs["record_max_elems"] # Strip the init entry (index 0) — cuthbert output is (T+1, ...), @@ -472,8 +603,7 @@ def _add_sites_taylor_kf( numpyro.deterministic(f"{name}_filtered_states_chol_cov", chol_cov) if add_filtered_states_cov or add_filtered_states_cov_diag: - chol_t = jnp.transpose(chol_cov, (0, 2, 1)) - filtered_cov = jnp.matmul(chol_cov, chol_t) + filtered_cov = covariance_from_cholesky(chol_cov) if add_filtered_states_cov: numpyro.deterministic(f"{name}_filtered_states_cov", filtered_cov) diff --git a/dynestyx/inference/integrations/utils.py b/dynestyx/inference/integrations/utils.py index f41cc830..38f40826 100644 --- a/dynestyx/inference/integrations/utils.py +++ b/dynestyx/inference/integrations/utils.py @@ -6,6 +6,10 @@ from numpyro.distributions import constraints +def covariance_from_cholesky(chol_cov: jax.Array) -> jax.Array: + return jnp.matmul(chol_cov, jnp.swapaxes(chol_cov, -1, -2)) + + class WeightedParticles(dist.Distribution): """A distribution over a finite set of weighted particles. diff --git a/mkdocs.yml b/mkdocs.yml index f8dd2628..b22263b5 100644 --- a/mkdocs.yml +++ b/mkdocs.yml @@ -39,6 +39,7 @@ nav: - Sparse system identification: deep_dives/fhn_sparse_id.ipynb - Discrete time LTI profile likelihood: deep_dives/discrete_time_lti_profile_likelihood.ipynb - GP prior on drift (FitzHugh-Nagumo): deep_dives/gp_drift.ipynb + - Neural drift correction for a nonlinear oscillator: deep_dives/neural_drift_correction.ipynb - API Reference: - api_reference/index.md - Public API: diff --git a/pyproject.toml b/pyproject.toml index 7dc6bbaa..1ddc4a55 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -31,8 +31,8 @@ classifiers = [ ] dependencies = [ "effectful", - "cuthbert>=0.0.9", - "cuthbertlib>=0.0.9", + "cuthbert>=0.0.10", + "cuthbertlib>=0.0.10", "cd-dynamax", "matplotlib>=3.10.7", "numpyro>=0.19.0", diff --git a/tests/fixtures.py b/tests/fixtures.py index f7263731..5c03fa1a 100644 --- a/tests/fixtures.py +++ b/tests/fixtures.py @@ -79,6 +79,7 @@ def _normalize_synthetic(synthetic: dict) -> dict: ContinuousTimeKFConfig, ContinuousTimeUKFConfig, EKFConfig, + EnKFConfig, KFConfig, PFConfig, UKFConfig, @@ -958,6 +959,11 @@ def data_conditioned_model(): "ekf": EKFConfig( record_filtered_states_mean=True, filter_source=filter_source ), + "enkf": EnKFConfig( + record_filtered_states_mean=True, + n_particles=_n_particles(64), + filter_source=filter_source, + ), "ukf": UKFConfig( record_filtered_states_mean=True, filter_source=filter_source ), @@ -1075,11 +1081,13 @@ def data_conditioned_model(): (False, "kf", "cd_dynamax"), (False, "kf", "cuthbert"), (False, "ekf", "cuthbert"), + (False, "enkf", "cuthbert"), (False, "ekf", "cd_dynamax"), (False, "ukf", "cd_dynamax"), (True, "kf", "cd_dynamax"), (True, "kf", "cuthbert"), (True, "ekf", "cuthbert"), + (True, "enkf", "cuthbert"), (True, "ekf", "cd_dynamax"), (True, "ukf", "cd_dynamax"), ], @@ -1135,6 +1143,9 @@ def data_conditioned_model(): config = { "kf": KFConfig(filter_source=filter_source), "ekf": EKFConfig(filter_source=filter_source), + "enkf": EnKFConfig( + n_particles=_n_particles(64), filter_source=filter_source + ), "ukf": UKFConfig(filter_source=filter_source), }[filter_type] with Filter(filter_config=config): diff --git a/tests/test_filters.py b/tests/test_filters.py index 2ce490d8..ae51cdf1 100644 --- a/tests/test_filters.py +++ b/tests/test_filters.py @@ -4,16 +4,28 @@ import numpyro.distributions as dist import pytest from numpyro.handlers import seed, trace +from numpyro.infer import Predictive import dynestyx as dsx -from dynestyx.inference.filter_configs import ContinuousTimeDPFConfig +from dynestyx.inference.filter_configs import ( + ContinuousTimeDPFConfig, + EnKFConfig, + KFConfig, + PFConfig, +) from dynestyx.inference.filters import Filter +from dynestyx.inference.integrations.cuthbert.discrete import ( + run_discrete_filter as run_cuthbert_discrete_filter, +) from dynestyx.models import ContinuousTimeStateEvolution, DynamicalModel +from dynestyx.simulators import DiscreteTimeSimulator from tests.fixtures import ( + _squeeze_sim_dims, data_conditioned_jumpy_controls, data_conditioned_jumpy_controls_ode, data_conditioned_jumpy_controls_sde, ) +from tests.models import discrete_time_l63_model, discrete_time_lti_simplified_model @pytest.mark.parametrize( @@ -22,6 +34,7 @@ ("kf", "cuthbert", 1e-1), ("kf", "cd_dynamax", 1e-1), ("ekf", "cuthbert", 1e-1), + ("enkf", "cuthbert", 2e-1), ("ekf", "cd_dynamax", 1e-1), ("ukf", "cd_dynamax", 1e-1), ("pf", "cuthbert", 1e-1), @@ -94,3 +107,181 @@ def model(): with seed(rng_seed=jr.PRNGKey(0)): with Filter(filter_config=ContinuousTimeDPFConfig(n_particles=32)): model() + + +def _make_discrete_lti_data(): + obs_times = jnp.arange(start=0.0, stop=6.0, step=1.0) + true_params = {"alpha": jnp.array(0.35)} + predictive = Predictive( + discrete_time_lti_simplified_model, + params=true_params, + num_samples=1, + exclude_deterministic=False, + ) + with DiscreteTimeSimulator(): + synthetic = predictive(jr.PRNGKey(0), predict_times=obs_times) + return obs_times, _squeeze_sim_dims(synthetic["f_observations"]) + + +def _make_discrete_lti_dynamics(alpha=0.35): + state_dim = 2 + return dsx.LTI_discrete( + A=jnp.array([[alpha, 0.1], [0.1, 0.8]]), + Q=0.1 * jnp.eye(state_dim), + H=jnp.array([[1.0, 0.0]]), + R=jnp.array([[0.5**2]]), + B=jnp.array([[0.1], [0.0]]), + D=jnp.array([[0.01]]), + ) + + +@pytest.mark.parametrize( + "filter_config", + [ + KFConfig(filter_source="cuthbert"), + EnKFConfig(n_particles=16, filter_source="cuthbert"), + PFConfig(n_particles=16, filter_source="cuthbert"), + ], +) +def test_cuthbert_filtered_distribution_shapes_match_observations(filter_config): + obs_times, obs_values = _make_discrete_lti_data() + dynamics = _make_discrete_lti_dynamics() + + with trace(), seed(rng_seed=jr.PRNGKey(1)): + filtered_dists = run_cuthbert_discrete_filter( + "f", + dynamics, + filter_config, + key=jr.PRNGKey(2), + obs_times=obs_times, + obs_values=obs_values, + ) + + assert len(filtered_dists) == len(obs_times) + for filtered_dist in filtered_dists: + assert filtered_dist.event_shape == (dynamics.state_dim,) + + if isinstance(filter_config, PFConfig): + assert filtered_dist.particles.shape == ( + filter_config.n_particles, + dynamics.state_dim, + ) + assert filtered_dist.log_weights.shape == (filter_config.n_particles,) + else: + assert filtered_dist.mean.shape == (dynamics.state_dim,) + assert filtered_dist.covariance_matrix.shape == ( + dynamics.state_dim, + dynamics.state_dim, + ) + + +def test_cuthbert_enkf_accepts_callable_independent_normal_observation(): + obs_times = jnp.arange(start=0.0, stop=4.0, step=1.0) + obs_values = jnp.zeros((len(obs_times), 2)) + dynamics = DynamicalModel( + initial_condition=dist.MultivariateNormal( + loc=jnp.zeros(2), covariance_matrix=jnp.eye(2) + ), + state_evolution=lambda x, u, t_now, t_next: dist.MultivariateNormal( + loc=0.9 * x, + covariance_matrix=0.1 * jnp.eye(2), + ), + observation_model=lambda x, u, t: dist.Independent( + dist.Normal(loc=x, scale=0.2), 1 + ), + ) + + with trace(), seed(rng_seed=jr.PRNGKey(1)): + filtered_dists = run_cuthbert_discrete_filter( + "f", + dynamics, + EnKFConfig(n_particles=16, filter_source="cuthbert"), + key=jr.PRNGKey(2), + obs_times=obs_times, + obs_values=obs_values, + ) + + assert len(filtered_dists) == len(obs_times) + assert all(d.event_shape == (dynamics.state_dim,) for d in filtered_dists) + + +def test_cuthbert_enkf_records_filtered_gaussian_sites(): + obs_times, obs_values = _make_discrete_lti_data() + substituted = numpyro.handlers.substitute( + discrete_time_lti_simplified_model, data={"alpha": jnp.array(0.35)} + ) + filter_config = EnKFConfig( + n_particles=16, + filter_source="cuthbert", + record_filtered_states_mean=True, + record_filtered_states_cov=True, + record_filtered_states_cov_diag=True, + record_filtered_states_chol_cov=True, + ) + + with trace() as tr, seed(rng_seed=jr.PRNGKey(1)): + with Filter(filter_config=filter_config): + substituted(obs_times=obs_times, obs_values=obs_values) + + assert "f_marginal_loglik" in tr + assert tr["f_filtered_states_mean"]["value"].shape == (len(obs_times), 2) + assert tr["f_filtered_states_cov"]["value"].shape == (len(obs_times), 2, 2) + assert tr["f_filtered_states_cov_diag"]["value"].shape == (len(obs_times), 2) + assert tr["f_filtered_states_chol_cov"]["value"].shape[:2] == ( + len(obs_times), + 2, + ) + + +def test_cuthbert_enkf_fixed_crn_seed_is_deterministic(): + obs_times, obs_values = _make_discrete_lti_data() + substituted = numpyro.handlers.substitute( + discrete_time_lti_simplified_model, data={"alpha": jnp.array(0.35)} + ) + filter_config = EnKFConfig( + n_particles=16, filter_source="cuthbert", crn_seed=jr.PRNGKey(123) + ) + + def _run(seed_key): + with trace() as tr, seed(rng_seed=seed_key): + with Filter(filter_config=filter_config): + substituted(obs_times=obs_times, obs_values=obs_values) + return tr["f_marginal_loglik"]["value"] + + assert jnp.isclose(_run(jr.PRNGKey(1)), _run(jr.PRNGKey(2))) + + +def test_cuthbert_enkf_crn_seed_none_uses_numpyro_seed(): + obs_times, obs_values = _make_discrete_lti_data() + substituted = numpyro.handlers.substitute( + discrete_time_lti_simplified_model, data={"alpha": jnp.array(0.35)} + ) + + with trace() as tr, seed(rng_seed=jr.PRNGKey(1)): + with Filter(filter_config=EnKFConfig(n_particles=16, crn_seed=None)): + substituted(obs_times=obs_times, obs_values=obs_values) + + assert jnp.isfinite(tr["f_marginal_loglik"]["value"]) + + +def test_default_discrete_filter_uses_cuthbert_enkf_on_nonlinear_model(): + obs_times = jnp.arange(start=0.0, stop=1.0, step=0.2) + true_params = {"rho": jnp.array(28.0)} + predictive = Predictive( + discrete_time_l63_model, + params=true_params, + num_samples=1, + exclude_deterministic=False, + ) + with DiscreteTimeSimulator(): + synthetic = predictive(jr.PRNGKey(0), predict_times=obs_times) + obs_values = _squeeze_sim_dims(synthetic["f_observations"]) + substituted = numpyro.handlers.substitute( + discrete_time_l63_model, data={"rho": jnp.array(28.0)} + ) + + with trace() as tr, seed(rng_seed=jr.PRNGKey(2)): + with Filter(): + substituted(obs_times=obs_times, obs_values=obs_values) + + assert jnp.isfinite(tr["f_marginal_loglik"]["value"]) From d8796f3552ab1ec3f1958d6c4430107a3d009e3d Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Wed, 22 Apr 2026 14:55:42 -0400 Subject: [PATCH 02/11] Remove extraneous link --- mkdocs.yml | 1 - 1 file changed, 1 deletion(-) diff --git a/mkdocs.yml b/mkdocs.yml index b22263b5..f8dd2628 100644 --- a/mkdocs.yml +++ b/mkdocs.yml @@ -39,7 +39,6 @@ nav: - Sparse system identification: deep_dives/fhn_sparse_id.ipynb - Discrete time LTI profile likelihood: deep_dives/discrete_time_lti_profile_likelihood.ipynb - GP prior on drift (FitzHugh-Nagumo): deep_dives/gp_drift.ipynb - - Neural drift correction for a nonlinear oscillator: deep_dives/neural_drift_correction.ipynb - API Reference: - api_reference/index.md - Public API: From ec6d5dc3fe5966b2952d5a724afde9b6922503e4 Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Wed, 22 Apr 2026 14:57:05 -0400 Subject: [PATCH 03/11] Fix mis-specified filter source --- dynestyx/inference/filter_configs.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/dynestyx/inference/filter_configs.py b/dynestyx/inference/filter_configs.py index dff99708..47f048be 100644 --- a/dynestyx/inference/filter_configs.py +++ b/dynestyx/inference/filter_configs.py @@ -373,7 +373,7 @@ class KFConfig(BaseFilterConfig): - For more details on the `cuthbert` implementation, see the [cuthbert documentation](https://state-space-models.github.io/cuthbert/cuthbert_api/gaussian/kalman/). """ - filter_source: CDDynamaxOnlyFilterSource = "cd_dynamax" + filter_source: CuthbertOrCDDynamaxFilterSource = "cd_dynamax" @dataclasses.dataclass From 285af3282329597a99a3862a5fd133da446b68b4 Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Wed, 22 Apr 2026 15:13:21 -0400 Subject: [PATCH 04/11] Update dynestyx/inference/integrations/cuthbert/discrete.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --- dynestyx/inference/integrations/cuthbert/discrete.py | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/dynestyx/inference/integrations/cuthbert/discrete.py b/dynestyx/inference/integrations/cuthbert/discrete.py index ea15e4ea..3f00c8e3 100644 --- a/dynestyx/inference/integrations/cuthbert/discrete.py +++ b/dynestyx/inference/integrations/cuthbert/discrete.py @@ -62,11 +62,8 @@ def _config_to_filter_kwargs(config: BaseFilterConfig) -> dict: kwargs["inflation"] = ( config.inflation_delta if config.inflation_delta is not None else 0.0 ) - kwargs["perturbed_obs"] = ( - config.perturb_measurements - if config.perturb_measurements is not None - else True - ) + if config.perturb_measurements is not None: + kwargs["perturbed_obs"] = config.perturb_measurements return kwargs From 6c6130ffbb54ecde9baa7200b317bc99bfc12d3a Mon Sep 17 00:00:00 2001 From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com> Date: Wed, 22 Apr 2026 19:23:01 +0000 Subject: [PATCH 05/11] Revert accidental neural nets FAQ change Agent-Logs-Url: https://github.com/BasisResearch/dynestyx/sessions/804f2085-9a89-43cc-890b-77758f2f48ab Co-authored-by: DanWaxman <11810745+DanWaxman@users.noreply.github.com> --- docs/faq.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/faq.md b/docs/faq.md index 27796419..59e90b03 100644 --- a/docs/faq.md +++ b/docs/faq.md @@ -97,7 +97,7 @@ Hierarchical models are supported by the [`dsx.plate`](./api_reference/public/ha ## What about neural nets? -See our [neural drift correction deep dive](deep_dives/neural_drift_correction.ipynb). It shows how to keep the observation model and filter workflow fixed while swapping a mechanistic ODE drift for a mechanistic-plus-neural residual drift. +We will put examples up soon. See [CD-Dynamax's Lorenz 63 neural drift tutorial](https://github.com/hd-UQ/cd_dynamax/blob/dev-numpyro-api/demos/numpyro/notebooks/lorenz63_nndrift_sgd_fit_to_data_tutorial_newAPI.ipynb) to convince yourself that this will work well. ## What about SINDy? From 27601d3e1e567adb29d1061597facacbe3b7f278 Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Wed, 22 Apr 2026 15:52:16 -0400 Subject: [PATCH 06/11] Properly handle state-dependent chol_R with error --- dynestyx/inference/filter_configs.py | 5 + .../integrations/cuthbert/discrete.py | 95 ++++++++++++++----- tests/test_filters.py | 27 ++++++ 3 files changed, 103 insertions(+), 24 deletions(-) diff --git a/dynestyx/inference/filter_configs.py b/dynestyx/inference/filter_configs.py index 47f048be..f2fefe26 100644 --- a/dynestyx/inference/filter_configs.py +++ b/dynestyx/inference/filter_configs.py @@ -90,6 +90,11 @@ class EnKFConfig(BaseFilterConfig): state dimensions. Cheaper per-step than the particle filter, but assumes observations are approximately Gaussian given the ensemble. + The observation noise covariance must be **state-independent** (it may + still depend on time or controls). Using a state-dependent scale with the + cuthbert backend raises a `ValueError`; if you need heteroscedastic noise, + use `PFConfig` instead. + The primary tuning knob is `n_particles`, with more particles providing more accurate results at the cost of higher compute. If the ensemble collapses over long trajectories, increase diff --git a/dynestyx/inference/integrations/cuthbert/discrete.py b/dynestyx/inference/integrations/cuthbert/discrete.py index 3f00c8e3..a30fe6d7 100644 --- a/dynestyx/inference/integrations/cuthbert/discrete.py +++ b/dynestyx/inference/integrations/cuthbert/discrete.py @@ -47,6 +47,57 @@ class CuthbertInputs(NamedTuple): is_first_step: jax.Array # (T+1,) bool — True only at index 1 +def _extract_gaussian_chol(d: dist.Distribution, obs_dim: int) -> jax.Array: + """Extract a Cholesky factor of the covariance from a Gaussian distribution.""" + if isinstance(d, dist.MultivariateNormal): + return jnp.asarray(d.scale_tril) + if isinstance(d, dist.Independent) and isinstance(d.base_dist, dist.Normal): + scale = jnp.atleast_1d(jnp.asarray(d.base_dist.scale)) + elif isinstance(d, dist.Normal): + scale = jnp.atleast_1d(jnp.asarray(d.scale)) + else: + raise TypeError( + "cuthbert EnKF requires Gaussian observation distributions. " + "Expected LinearGaussianObservation, GaussianObservation, or a " + "callable returning Normal, Independent(Normal), or " + f"MultivariateNormal; got {type(d).__name__}." + ) + if scale.size == 1 and obs_dim > 1: + scale = jnp.full((obs_dim,), scale[0]) + return jnp.diag(scale) + + +def _check_state_independent_noise( + chol_R_at_x0: jax.Array, probe_dist_at_x1: dist.Distribution, obs_dim: int +) -> None: + """Raise if the observation noise covariance varies with state. + + cuthbert's EnKF API resolves ``chol_R`` once per step (before the ensemble + update), so it may depend on time/controls but NOT on the latent state. + Silently using a state-dependent scale would freeze it at the probe state + and misrepresent the noise. Detect this under concrete evaluation and raise. + Under JAX tracing, the comparison yields a tracer and we skip the check — + the constraint is documented in the filter docstring. + """ + chol_R_at_x1 = _extract_gaussian_chol(probe_dist_at_x1, obs_dim) + try: + equal = bool( + jnp.asarray(chol_R_at_x0).shape == jnp.asarray(chol_R_at_x1).shape + and jnp.allclose(chol_R_at_x0, chol_R_at_x1) + ) + except jax.errors.TracerBoolConversionError: + return + if not equal: + raise ValueError( + "cuthbert EnKF requires state-independent observation noise, but " + "the observation scale changes with the latent state (heteroscedastic " + "noise). The EnKF API resolves chol_R once per step before the " + "ensemble update, so a state-dependent scale cannot be honoured. " + "Either make the noise depend only on time/controls, or use a " + "particle filter (PFConfig)." + ) + + def _config_to_filter_kwargs(config: BaseFilterConfig) -> dict: """Build filter_kwargs dict from config dataclass.""" kwargs = dict(config.extra_filter_kwargs) @@ -259,6 +310,19 @@ def _cuthbert_filter_enkf(dynamics: DynamicalModel, filter_kwargs: dict | None = state_dim = dynamics.state_dim obs_dim = dynamics.observation_dim + obs_model = dynamics.observation_model + if not isinstance(obs_model, (LinearGaussianObservation, GaussianObservation)): + probe_u = jnp.zeros(()) + probe_t = jnp.zeros(()) + try: + probe_d0 = obs_model(jnp.zeros((state_dim,)), probe_u, probe_t) + probe_d1 = obs_model(jnp.ones((state_dim,)), probe_u, probe_t) + except Exception: + probe_d0 = probe_d1 = None + if probe_d0 is not None and probe_d1 is not None: + chol0 = _extract_gaussian_chol(probe_d0, obs_dim) + _check_state_independent_noise(chol0, probe_d1, obs_dim) + def init_sample(key, mi: CuthbertInputs): return jnp.atleast_1d(jnp.asarray(dynamics.initial_condition.sample(key))) @@ -304,30 +368,13 @@ def observation_fn(x): return observation_fn, chol_R, y else: - probe_x = jnp.zeros((state_dim,), dtype=y.dtype) - probe_dist = obs_model(probe_x, mi.u, mi.time) - - if isinstance(probe_dist, dist.MultivariateNormal): - chol_R = jnp.asarray(probe_dist.scale_tril) - elif isinstance(probe_dist, dist.Normal): - scale = jnp.atleast_1d(jnp.asarray(probe_dist.scale)) - if scale.size == 1 and obs_dim > 1: - scale = jnp.full((obs_dim,), scale[0]) - chol_R = jnp.diag(scale) - elif isinstance(probe_dist, dist.Independent) and isinstance( - probe_dist.base_dist, dist.Normal - ): - scale = jnp.atleast_1d(jnp.asarray(probe_dist.base_dist.scale)) - if scale.size == 1 and obs_dim > 1: - scale = jnp.full((obs_dim,), scale[0]) - chol_R = jnp.diag(scale) - else: - raise TypeError( - "cuthbert EnKF requires Gaussian observation distributions. " - "Expected LinearGaussianObservation, GaussianObservation, or a " - "callable returning Normal, Independent(Normal), or " - f"MultivariateNormal; got {type(probe_dist).__name__}." - ) + probe_x0 = jnp.zeros((state_dim,), dtype=y.dtype) + probe_x1 = jnp.ones((state_dim,), dtype=y.dtype) + probe_dist = obs_model(probe_x0, mi.u, mi.time) + chol_R = _extract_gaussian_chol(probe_dist, obs_dim) + _check_state_independent_noise( + chol_R, obs_model(probe_x1, mi.u, mi.time), obs_dim + ) def observation_fn(x): edist = obs_model(x, mi.u, mi.time) diff --git a/tests/test_filters.py b/tests/test_filters.py index ae51cdf1..d9d60f6b 100644 --- a/tests/test_filters.py +++ b/tests/test_filters.py @@ -205,6 +205,33 @@ def test_cuthbert_enkf_accepts_callable_independent_normal_observation(): assert all(d.event_shape == (dynamics.state_dim,) for d in filtered_dists) +def test_cuthbert_enkf_rejects_state_dependent_observation_noise(): + obs_times = jnp.arange(start=0.0, stop=4.0, step=1.0) + obs_values = jnp.zeros((len(obs_times), 2)) + dynamics = DynamicalModel( + initial_condition=dist.MultivariateNormal( + loc=jnp.zeros(2), covariance_matrix=jnp.eye(2) + ), + state_evolution=lambda x, u, t_now, t_next: dist.MultivariateNormal( + loc=0.9 * x, covariance_matrix=0.1 * jnp.eye(2) + ), + observation_model=lambda x, u, t: dist.Independent( + dist.Normal(loc=x, scale=0.1 + 0.5 * jnp.abs(x)), 1 + ), + ) + + with pytest.raises(ValueError, match="state-independent"): + with trace(), seed(rng_seed=jr.PRNGKey(1)): + run_cuthbert_discrete_filter( + "f", + dynamics, + EnKFConfig(n_particles=16, filter_source="cuthbert"), + key=jr.PRNGKey(2), + obs_times=obs_times, + obs_values=obs_values, + ) + + def test_cuthbert_enkf_records_filtered_gaussian_sites(): obs_times, obs_values = _make_discrete_lti_data() substituted = numpyro.handlers.substitute( From c57f7459bcd1b46eed3950917054efa1f2184864 Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Wed, 22 Apr 2026 16:14:29 -0400 Subject: [PATCH 07/11] Lint --- .../inference/integrations/cuthbert/discrete.py | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/dynestyx/inference/integrations/cuthbert/discrete.py b/dynestyx/inference/integrations/cuthbert/discrete.py index a30fe6d7..22a60ab8 100644 --- a/dynestyx/inference/integrations/cuthbert/discrete.py +++ b/dynestyx/inference/integrations/cuthbert/discrete.py @@ -1,3 +1,4 @@ +import warnings from typing import NamedTuple import jax @@ -14,6 +15,7 @@ stop_gradient_decorator, systematic, ) +from numpyro.distributions import Distribution from dynestyx.inference.filter_configs import ( BaseFilterConfig, @@ -315,10 +317,16 @@ def _cuthbert_filter_enkf(dynamics: DynamicalModel, filter_kwargs: dict | None = probe_u = jnp.zeros(()) probe_t = jnp.zeros(()) try: - probe_d0 = obs_model(jnp.zeros((state_dim,)), probe_u, probe_t) - probe_d1 = obs_model(jnp.ones((state_dim,)), probe_u, probe_t) + probe_d0: Distribution | None = obs_model( + jnp.zeros((state_dim,)), probe_u, probe_t + ) + probe_d1: Distribution | None = obs_model( + jnp.ones((state_dim,)), probe_u, probe_t + ) except Exception: - probe_d0 = probe_d1 = None + warnings.warn( + "Failed to probe observation model for state-independent noise check. Please ensure the observation model is state-independent." + ) if probe_d0 is not None and probe_d1 is not None: chol0 = _extract_gaussian_chol(probe_d0, obs_dim) _check_state_independent_noise(chol0, probe_d1, obs_dim) From 782c4edf096c7f14990bbcd5fbeb8046c5e48a1d Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Wed, 22 Apr 2026 16:16:11 -0400 Subject: [PATCH 08/11] Update EKF description --- docs/api_reference/public/inference/filter_configs.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/api_reference/public/inference/filter_configs.md b/docs/api_reference/public/inference/filter_configs.md index 2c4d83be..a0a0a521 100644 --- a/docs/api_reference/public/inference/filter_configs.md +++ b/docs/api_reference/public/inference/filter_configs.md @@ -8,7 +8,7 @@ The single `Filter()` handler is directed to the appropriate filtering algorithm |----------------------------|---------------------|-------------------| | `KFConfig` | Discrete | Linear-Gaussian dynamics and linear-Gaussian observations (exact & optimal). | | `EnKFConfig` | Discrete | Nonlinear or expensive models with Gaussian observations; cuthbert-backed and a good general-purpose default. *(default)* | -| `EKFConfig` | Discrete | Nonlinear, differentiable Gaussian dynamics, nonlinear but differentiable Gaussian observations (approximate). | +| `EKFConfig` | Discrete | Nonlinear, differentiable Gaussian dynamics, nonlinear (and with `cuthbert`, non-Gaussian) but differentiable observations (approximate). | | `UKFConfig` | Discrete | Nonlinear, differentiable Gaussian dynamics, nonlinear but differentiable Gaussian observations (approximate). Generally more accurate, but slower than `EKFConfig`. | | `PFConfig` | Discrete | Applicable for arbitrary state-space models, but quite expensive and noisy estimates (asymptotically exact in the limit of infinite particles, approximate in practice). | | `HMMConfig` | Discrete (HMM) | Finite discrete latent state space (exact & optimal). | From 4787fc8e049624de4fd7603ca2814ee0f72d8459 Mon Sep 17 00:00:00 2001 From: Matthew Levine Date: Wed, 22 Apr 2026 16:19:12 -0400 Subject: [PATCH 09/11] re-run profile likelihood with better plot axes --- ...discrete_time_lti_profile_likelihood.ipynb | 42 +++++++++---------- 1 file changed, 20 insertions(+), 22 deletions(-) diff --git a/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb b/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb index ed2c2d71..22c1d263 100644 --- a/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb +++ b/docs/deep_dives/discrete_time_lti_profile_likelihood.ipynb @@ -260,13 +260,13 @@ "text": [ "Filter | max profile ll | time (s)\n", "------------|-----------------|--------\n", - "KF (Cuthbert) | -912.7115 | 1.805\n", - "KF (Dynamax) | -912.7115 | 0.915\n", - "EKF (Cuthbert) | -912.7115 | 2.249\n", - "EKF (Dynamax) | -912.7115 | 0.694\n", - "EnKF (Cuthbert) | -916.0390 | 1.562\n", - "UKF (Dynamax) | -912.7115 | 0.784\n", - "PF (Cuthbert) | -912.7143 | 10.546\n" + "KF (Cuthbert) | -912.7115 | 1.723\n", + "KF (Dynamax) | -912.7115 | 0.843\n", + "EKF (Cuthbert) | -912.7115 | 2.028\n", + "EKF (Dynamax) | -912.7115 | 0.684\n", + "EnKF (Cuthbert) | -916.0390 | 1.333\n", + "UKF (Dynamax) | -912.7115 | 0.717\n", + "PF (Cuthbert) | -912.7143 | 7.158\n" ] } ], @@ -303,12 +303,12 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 15, "metadata": {}, "outputs": [ { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -324,7 +324,6 @@ "times = [timings[n] for n in names]\n", "ax.bar(names, times, color=[\"C0\", \"C1\", \"C2\", \"C3\", \"C4\", \"C5\"])\n", "ax.set_ylabel(\"Time (s)\")\n", - "ax.set_ylim([0, 2])\n", "ax.set_title(\"Profile likelihood run time by filter\")\n", "plt.tight_layout()\n", "plt.show()" @@ -418,13 +417,13 @@ "text": [ "Filter | max score | time (s)\n", "------------|-----------------|--------\n", - "KF (Cuthbert) | 182.1674 | 3.272\n", - "KF (Dynamax) | 182.1674 | 2.088\n", - "EKF (Cuthbert) | 182.1674 | 2.855\n", - "EKF (Dynamax) | 182.1674 | 2.117\n", - "EnKF (Cuthbert) | 183.2659 | 2.371\n", - "UKF (Dynamax) | 182.1674 | 2.245\n", - "PF (Cuthbert) | 201.7599 | 7.428\n" + "KF (Cuthbert) | 182.1674 | 2.147\n", + "KF (Dynamax) | 182.1674 | 2.152\n", + "EKF (Cuthbert) | 182.1674 | 2.614\n", + "EKF (Dynamax) | 182.1674 | 2.008\n", + "EnKF (Cuthbert) | 183.2659 | 2.245\n", + "UKF (Dynamax) | 182.1674 | 2.160\n", + "PF (Cuthbert) | 201.7599 | 7.097\n" ] } ], @@ -461,12 +460,12 @@ }, { "cell_type": "code", - "execution_count": 11, + "execution_count": null, "metadata": {}, "outputs": [ { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -482,7 +481,6 @@ "times = [timings[n] for n in names]\n", "ax.bar(names, times, color=[\"C0\", \"C1\", \"C2\", \"C3\", \"C4\", \"C5\"])\n", "ax.set_ylabel(\"Time (s)\")\n", - "ax.set_ylim([0, 2])\n", "ax.set_title(\"Profile likelihood run time by filter\")\n", "plt.tight_layout()\n", "plt.show()" @@ -529,7 +527,7 @@ ], "metadata": { "kernelspec": { - "display_name": ".venv", + "display_name": ".venv (3.12.11)", "language": "python", "name": "python3" }, @@ -543,7 +541,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.12.10" + "version": "3.12.11" } }, "nbformat": 4, From 7e3c3fcf7016bb277a06a9d93e5c3800308b7764 Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Thu, 23 Apr 2026 09:04:11 -0400 Subject: [PATCH 10/11] Move cuthbert indexing shenanigans --- dynestyx/inference/filters.py | 14 +--- .../integrations/cuthbert/discrete.py | 74 +++++++++++-------- tests/test_filters.py | 38 ++++++++++ ...erarchical_simulator_discretizer_smokes.py | 20 +++++ 4 files changed, 103 insertions(+), 43 deletions(-) diff --git a/dynestyx/inference/filters.py b/dynestyx/inference/filters.py index 56b6ffe3..2184730e 100644 --- a/dynestyx/inference/filters.py +++ b/dynestyx/inference/filters.py @@ -113,24 +113,14 @@ def _cuthbert_states_to_dists( if isinstance(config, PFConfig): particles = states.particles log_weights = states.log_weights - # cuthbert includes an init step at index 0; align with dynestyx T convention. - particles = particles[ - (slice(None),) * len(plate_shapes) + (slice(1, None), ...) - ] - log_weights = log_weights[ - (slice(None),) * len(plate_shapes) + (slice(1, None), ...) - ] return _particle_to_batched_dists( particles, log_weights, plate_shapes=plate_shapes, ) - # Kalman / Taylor-KF variants expose mean/chol_cov and include init at index 0. - mean = states.mean[(slice(None),) * len(plate_shapes) + (slice(1, None), ...)] - chol_cov = states.chol_cov[ - (slice(None),) * len(plate_shapes) + (slice(1, None), ...) - ] + mean = states.mean + chol_cov = states.chol_cov cov = covariance_from_cholesky(chol_cov) t_len = _time_len_from_array(mean, plate_shapes) return [ diff --git a/dynestyx/inference/integrations/cuthbert/discrete.py b/dynestyx/inference/integrations/cuthbert/discrete.py index 22a60ab8..1361415a 100644 --- a/dynestyx/inference/integrations/cuthbert/discrete.py +++ b/dynestyx/inference/integrations/cuthbert/discrete.py @@ -120,6 +120,22 @@ def _config_to_filter_kwargs(config: BaseFilterConfig) -> dict: return kwargs +def _drop_cuthbert_dummy_step(states, *, obs_len: int): + """Drop cuthbert's leading dummy state from every time-indexed leaf.""" + raw_len = obs_len + 1 + + def _drop_if_time_leaf(leaf): + shape = getattr(leaf, "shape", None) + ndim = getattr(leaf, "ndim", None) + if ndim is None and shape is not None: + ndim = len(shape) + if shape is not None and ndim is not None and ndim > 0 and shape[0] == raw_len: + return leaf[1:] + return leaf + + return jax.tree.map(_drop_if_time_leaf, states) + + def compute_cuthbert_filter( dynamics: DynamicalModel, filter_config: BaseFilterConfig, @@ -133,18 +149,18 @@ def compute_cuthbert_filter( """Pure-JAX cuthbert filter computation (no numpyro side-effects). Returns: - tuple: (marginal_loglik, states) where states is the raw cuthbert filter state. + tuple: (marginal_loglik, states) where states are aligned to obs_times. """ filter_kwargs = _config_to_filter_kwargs(filter_config) ys = obs_values - T1 = int(ys.shape[0]) + obs_len = int(ys.shape[0]) times = obs_times if ctrl_values is None: control_dim = dynamics.control_dim - ctrl_values = jnp.zeros((T1, control_dim), dtype=ys.dtype) - elif ctrl_values.shape[0] > T1: + ctrl_values = jnp.zeros((obs_len, control_dim), dtype=ys.dtype) + elif ctrl_values.shape[0] > obs_len: inds = jnp.searchsorted(ctrl_times, times, side="left") ctrl_values = ctrl_values[inds] @@ -162,7 +178,7 @@ def compute_cuthbert_filter( u_prev=jnp.concatenate([dummy_u, u_prev], axis=0), time=jnp.concatenate([dummy_time, times], axis=0), time_prev=jnp.concatenate([dummy_time, time_prev], axis=0), - is_first_step=jnp.arange(T1 + 1) == 1, + is_first_step=jnp.arange(obs_len + 1) == 1, ) if isinstance(filter_config, PFConfig): @@ -189,8 +205,9 @@ def compute_cuthbert_filter( "Expected KFConfig, EKFConfig, EnKFConfig, PFConfig." ) - states = cuthbert_filter(filter_obj, cuthbert_inputs, parallel=False, key=key) - marginal_loglik = states.log_normalizing_constant[-1] + raw_states = cuthbert_filter(filter_obj, cuthbert_inputs, parallel=False, key=key) + marginal_loglik = raw_states.log_normalizing_constant[-1] + states = _drop_cuthbert_dummy_step(raw_states, obs_len=obs_len) return marginal_loglik, states @@ -211,8 +228,8 @@ def run_discrete_filter( Returns: list[dist.Distribution]: Filtered state distributions at each obs time. """ - T1 = int(obs_values.shape[0]) - if T1 == 0: + obs_len = int(obs_values.shape[0]) + if obs_len == 0: return [] marginal_loglik, states = compute_cuthbert_filter( @@ -231,14 +248,14 @@ def run_discrete_filter( if isinstance(filter_config, PFConfig): _add_sites_pf(name, states, record_kwargs) - particles = states.particles[1:] + particles = states.particles if particles.ndim == 2: particles = particles[..., None] - return particles_to_delta_mixtures(particles, states.log_weights[1:]) + return particles_to_delta_mixtures(particles, states.log_weights) else: _add_sites_gaussian_filter(name, states, record_kwargs) - mean = states.mean[1:] - chol_cov = states.chol_cov[1:] + mean = states.mean + chol_cov = states.chol_cov cov = covariance_from_cholesky(chol_cov) return [ dist.MultivariateNormal(mean[i], covariance_matrix=cov[i]) @@ -563,15 +580,12 @@ def log_potential(x): def _add_sites_pf( name: str, states: particle_filter.ParticleFilterState, record_kwargs: dict ): - # Strip the init entry (index 0) — cuthbert output is (T+1, ...), - # we want (T, ...) matching dynestyx's convention where output[0] - # is the filtered state after observing y_0. - log_weights = states.log_weights[1:] - particles = states.particles[1:] + log_weights = states.log_weights + particles = states.particles if particles.ndim == 2: particles = particles[..., None] max_elems = record_kwargs["record_max_elems"] - t1, n_particles, state_dim = particles.shape + t_len, n_particles, state_dim = particles.shape add_particles = _should_record_field( record_kwargs["record_filtered_particles"], particles.shape, max_elems @@ -580,15 +594,15 @@ def _add_sites_pf( record_kwargs["record_filtered_log_weights"], log_weights.shape, max_elems ) add_mean = _should_record_field( - record_kwargs["record_filtered_states_mean"], (t1, state_dim), max_elems + record_kwargs["record_filtered_states_mean"], (t_len, state_dim), max_elems ) add_filtered_states_cov = _should_record_field( record_kwargs["record_filtered_states_cov"], - (t1, state_dim, state_dim), + (t_len, state_dim, state_dim), max_elems, ) add_filtered_states_cov_diag = _should_record_field( - record_kwargs["record_filtered_states_cov_diag"], (t1, state_dim), max_elems + record_kwargs["record_filtered_states_cov_diag"], (t_len, state_dim), max_elems ) need_filtered_means = ( @@ -596,8 +610,8 @@ def _add_sites_pf( ) if need_filtered_means: - w = jax.nn.softmax(log_weights, axis=1)[..., None] # (T+1, n_particles, 1) - filtered_means = jnp.sum(particles * w, axis=1) # (T+1, state_dim) + w = jax.nn.softmax(log_weights, axis=1)[..., None] # (T, n_particles, 1) + filtered_means = jnp.sum(particles * w, axis=1) # (T, state_dim) if add_filtered_states_cov or add_filtered_states_cov_diag: second_mom = jnp.einsum( @@ -626,11 +640,9 @@ def _add_sites_gaussian_filter( record_kwargs: dict, ): max_elems = record_kwargs["record_max_elems"] - # Strip the init entry (index 0) — cuthbert output is (T+1, ...), - # we want (T, ...) matching dynestyx's convention. - mean = states.mean[1:] - chol_cov = states.chol_cov[1:] - t1, state_dim, _ = chol_cov.shape + mean = states.mean + chol_cov = states.chol_cov + t_len, state_dim, _ = chol_cov.shape add_mean = _should_record_field( record_kwargs["record_filtered_states_mean"], mean.shape, max_elems @@ -642,11 +654,11 @@ def _add_sites_gaussian_filter( ) add_filtered_states_cov = _should_record_field( record_kwargs["record_filtered_states_cov"], - (t1, state_dim, state_dim), + (t_len, state_dim, state_dim), max_elems, ) add_filtered_states_cov_diag = _should_record_field( - record_kwargs["record_filtered_states_cov_diag"], (t1, state_dim), max_elems + record_kwargs["record_filtered_states_cov_diag"], (t_len, state_dim), max_elems ) if add_mean: diff --git a/tests/test_filters.py b/tests/test_filters.py index d9d60f6b..026f3bb4 100644 --- a/tests/test_filters.py +++ b/tests/test_filters.py @@ -9,11 +9,15 @@ import dynestyx as dsx from dynestyx.inference.filter_configs import ( ContinuousTimeDPFConfig, + EKFConfig, EnKFConfig, KFConfig, PFConfig, ) from dynestyx.inference.filters import Filter +from dynestyx.inference.integrations.cuthbert.discrete import ( + compute_cuthbert_filter, +) from dynestyx.inference.integrations.cuthbert.discrete import ( run_discrete_filter as run_cuthbert_discrete_filter, ) @@ -139,6 +143,40 @@ def _make_discrete_lti_dynamics(alpha=0.35): "filter_config", [ KFConfig(filter_source="cuthbert"), + EKFConfig(filter_source="cuthbert"), + EnKFConfig(n_particles=16, filter_source="cuthbert"), + PFConfig(n_particles=16, filter_source="cuthbert"), + ], +) +def test_compute_cuthbert_filter_returns_observation_aligned_states(filter_config): + obs_times, obs_values = _make_discrete_lti_data() + dynamics = _make_discrete_lti_dynamics() + + marginal_loglik, states = compute_cuthbert_filter( + dynamics, + filter_config, + key=jr.PRNGKey(2), + obs_times=obs_times, + obs_values=obs_values, + ) + + assert jnp.ndim(marginal_loglik) == 0 + assert states.log_normalizing_constant.shape[0] == len(obs_times) + assert states.model_inputs.y.shape[0] == len(obs_times) + + if isinstance(filter_config, PFConfig): + assert states.particles.shape[0] == len(obs_times) + assert states.log_weights.shape[0] == len(obs_times) + else: + assert states.mean.shape[0] == len(obs_times) + assert states.chol_cov.shape[0] == len(obs_times) + + +@pytest.mark.parametrize( + "filter_config", + [ + KFConfig(filter_source="cuthbert"), + EKFConfig(filter_source="cuthbert"), EnKFConfig(n_particles=16, filter_source="cuthbert"), PFConfig(n_particles=16, filter_source="cuthbert"), ], diff --git a/tests/test_hierarchical_simulator_discretizer_smokes.py b/tests/test_hierarchical_simulator_discretizer_smokes.py index 5fb08b38..5d5332a4 100644 --- a/tests/test_hierarchical_simulator_discretizer_smokes.py +++ b/tests/test_hierarchical_simulator_discretizer_smokes.py @@ -352,6 +352,26 @@ def test_plate_rollout_discrete_gaussian_pf_hmm(): assert tr["f_predicted_states"]["value"].shape[:3] == (2, 1, len(predict_times)) +def test_plate_rollout_discrete_cuthbert_kf_keeps_filtered_time_alignment(): + obs_times = jnp.arange(4.0) + predict_times = jnp.arange(6.0) + obs_gaussian = _make_obs_values((2, len(obs_times), 1)) + + with DiscreteTimeSimulator(): + with Filter(filter_config=KFConfig(filter_source="cuthbert")): + with trace() as tr, seed(rng_seed=jr.PRNGKey(18)): + _plate_discrete_lti_model( + obs_times=obs_times, + obs_values=obs_gaussian, + predict_times=predict_times, + M=2, + ) + + assert tr["f_predicted_times"]["value"].shape == (2, 1, len(predict_times)) + assert tr["f_predicted_states"]["value"].shape[:3] == (2, 1, len(predict_times)) + assert jnp.array_equal(tr["f_predicted_times"]["value"][0, 0], predict_times) + + def test_plate_rollout_continuous_gaussian_and_dpf(): obs_times = jnp.linspace(0.0, 0.3, 4) predict_times = jnp.linspace(0.0, 0.5, 6) From 8c7406216effcc150d2b2995000c90cb27da51ee Mon Sep 17 00:00:00 2001 From: Dan Waxman Date: Thu, 23 Apr 2026 09:13:02 -0400 Subject: [PATCH 11/11] Make state-dependency check a helper function --- .../integrations/cuthbert/discrete.py | 43 ++++++++++++------- 1 file changed, 27 insertions(+), 16 deletions(-) diff --git a/dynestyx/inference/integrations/cuthbert/discrete.py b/dynestyx/inference/integrations/cuthbert/discrete.py index 1361415a..109ef919 100644 --- a/dynestyx/inference/integrations/cuthbert/discrete.py +++ b/dynestyx/inference/integrations/cuthbert/discrete.py @@ -100,6 +100,30 @@ def _check_state_independent_noise( ) +def _probe_state_independent_observation_noise( + obs_model, *, state_dim: int, obs_dim: int +) -> None: + """Probe custom observation callables for state-dependent noise.""" + probe_u = jnp.zeros(()) + probe_t = jnp.zeros(()) + try: + probe_d0: Distribution | None = obs_model( + jnp.zeros((state_dim,)), probe_u, probe_t + ) + probe_d1: Distribution | None = obs_model( + jnp.ones((state_dim,)), probe_u, probe_t + ) + except Exception: + warnings.warn( + "Failed to probe observation model for state-independent noise check. Please ensure the observation model is state-independent." + ) + return + + if probe_d0 is not None and probe_d1 is not None: + chol0 = _extract_gaussian_chol(probe_d0, obs_dim) + _check_state_independent_noise(chol0, probe_d1, obs_dim) + + def _config_to_filter_kwargs(config: BaseFilterConfig) -> dict: """Build filter_kwargs dict from config dataclass.""" kwargs = dict(config.extra_filter_kwargs) @@ -331,22 +355,9 @@ def _cuthbert_filter_enkf(dynamics: DynamicalModel, filter_kwargs: dict | None = obs_model = dynamics.observation_model if not isinstance(obs_model, (LinearGaussianObservation, GaussianObservation)): - probe_u = jnp.zeros(()) - probe_t = jnp.zeros(()) - try: - probe_d0: Distribution | None = obs_model( - jnp.zeros((state_dim,)), probe_u, probe_t - ) - probe_d1: Distribution | None = obs_model( - jnp.ones((state_dim,)), probe_u, probe_t - ) - except Exception: - warnings.warn( - "Failed to probe observation model for state-independent noise check. Please ensure the observation model is state-independent." - ) - if probe_d0 is not None and probe_d1 is not None: - chol0 = _extract_gaussian_chol(probe_d0, obs_dim) - _check_state_independent_noise(chol0, probe_d1, obs_dim) + _probe_state_independent_observation_noise( + obs_model, state_dim=state_dim, obs_dim=obs_dim + ) def init_sample(key, mi: CuthbertInputs): return jnp.atleast_1d(jnp.asarray(dynamics.initial_condition.sample(key)))