Hi! I found that the generated data was self-repeated, and did not contain the data in previous dates. It could be a mistake in the data generating function:
|
out = pd.DataFrame(np.column_stack([_add_shifts(dat[:,:-1], 1, n_cols) for i in range(shifts)]), |
Here, the
_add_shifts(dat[:,:-1], 1, n_cols) should be
_add_shifts(dat[:,:-1], i, n_cols) to get shifted rows.
Hi! I found that the generated data was self-repeated, and did not contain the data in previous dates. It could be a mistake in the data generating function:
nih-community/utils.py
Line 118 in 2446a1e
Here, the
_add_shifts(dat[:,:-1], 1, n_cols)should be_add_shifts(dat[:,:-1], i, n_cols)to get shifted rows.