📊 Ontario Fuel Price — Data Visualization

A two-part data visualization project analyzing Ontario's retail fuel prices from 1990–2023,
spanning statistical charts, time series analysis, and geospatial mapping.

Assignment 1 — Statistical Charts · Assignment 2 — Time Series & Maps · Tech Stack · Setup

---

🗂 Dataset

Dataset	Source	Description
energy-price.csv	Ontario Open Data Portal	Weekly retail fuel prices (cents/litre) across 14 Ontario cities, 1990–present
volcano_db.csv	Plotly Datasets	Global volcano database with coordinates and country
earthquake.txt	Natural Resources Canada	Canadian earthquake records with magnitude and coordinates

The fuel price dataset covers Regular Unleaded Gasoline, diesel, auto propane, and compressed natural gas across cities including Ottawa, Toronto, Windsor, Sudbury, Thunder Bay, and more.

---

📁 Assignment 1 — Statistical Charts

Exploring the distribution and comparison of Regular Unleaded Gasoline prices across Ontario cities in 2023.

Q1 · Bar Chart — Mean Prices by City (2023)

Compares mean gasoline prices across 14 Ontario cities, sorted ascending by price.

# Key technique: filter → aggregate → sort → plot
mean_price = df[(df['Fuel Type'] == 'Regular Unleaded Gasoline') &
                (df['Date'].str.contains('2023'))][city].mean()
sorted_cities = sorted(mean_prices_dict, key=lambda x: mean_prices_dict[x])
plt.bar(sorted_cities, [mean_prices_dict[city] for city in sorted_cities])

---

Q2 · Dot Plot — Mean Prices by City (2023)

Same data as Q1, reframed as a horizontal dot plot — a cleaner alternative for reading exact values across many categories.

plt.scatter([mean_prices_dict[city] for city in sorted_cities], sorted_cities, marker='o')

---

Q3 · Histogram — Toronto West Price Distribution (2023)

Distribution of weekly prices in Toronto West/Ouest using 15 bins, revealing price clustering patterns throughout the year.

plt.hist(toronto_west_gas_2023, bins=15, edgecolor='black')

---

Q4 · ECDF — Normalized Cumulative Distribution (2023)

Empirical Cumulative Distribution Function showing the proportion of weeks at or below each price point in Toronto West/Ouest.

sorted_prices = np.sort(toronto_west_gas_2023)
ecdf = np.arange(1, len(sorted_prices) + 1) / len(sorted_prices)
plt.plot(sorted_prices, ecdf, marker='.', linestyle='none')

---

Q5 · Boxplot — Price Spread Across Cities (2021–Present)

Horizontal boxplots displaying the spread, median, and IQR of gasoline prices across all 14 cities from 2021 onwards.

plt.boxplot([filtered_data[city] for city in cities], labels=cities, vert=False, showfliers=False)

---

📁 Assignment 2 — Time Series & Geospatial Mapping

Moving beyond static distributions into temporal trends and geographic data.

Q1 · Time Series + Moving Average — Ottawa (2000–2023)

Plots weekly Ottawa gasoline prices from 2000–2023, smoothed with a Simple Moving Average (window = 40 weeks) to reveal the long-term price trend beneath short-term volatility.

df_filtered['SMA_40'] = df_filtered['Ottawa'].rolling(window=40).mean()

plt.plot(df_filtered['Date'], df_filtered['Ottawa'], label='Original')
plt.plot(df_filtered['Date'], df_filtered['SMA_40'], label='SMA 40', color='red')

What the trend shows: A steady climb from ~55¢/L in 2000, a spike around 2008, a sharp drop in 2020, and a steep rise into 2022.

---

Q2 · Geospatial Map — Canadian Volcanoes & Earthquakes

A layered map of Canada visualizing:

• 🔺 Volcanoes — filtered from global dataset, plotted as red triangles
• 🟠 Earthquakes — color-coded by magnitude (OrRd colormap), sourced from NRCan bulletin

canada.plot(ax=ax, color='lightgrey')
gdf_volcanoes.plot(ax=ax, color='red', marker='^', label='Volcanoes')
gdf_earthquakes.plot(ax=ax, column='Magnitude', cmap='OrRd', legend=True, markersize=10)

---

🛠 Tech Stack

Tool	Purpose
pandas	Data loading, filtering, aggregation, datetime handling
matplotlib	All chart types — bar, dot, histogram, ECDF, boxplot, time series
numpy	ECDF computation, array sorting
geopandas	Geospatial data handling and Canada map rendering
Jupyter Notebook	Interactive development environment

---

⚙️ Setup

# Clone the repository
git clone https://github.com/YOUR_USERNAME/ontario-fuel-visualization.git
cd ontario-fuel-visualization

# Install dependencies
pip install pandas matplotlib numpy geopandas

# Launch Jupyter
jupyter notebook

Place the datasets in a data/ folder before running:

project/
├── data/
│   ├── energy-price.csv
│   ├── volcano_db.csv
│   └── earthquake.txt
├── assignment1.ipynb
├── assignment2.ipynb
└── README.md

---

⚠️ Known Warnings

Warning	Cause	Resolution
SettingWithCopyWarning	Modifying a filtered DataFrame slice	Use .copy() after filtering or .loc[] for assignment
FutureWarning (GeoPandas)	geopandas.datasets deprecated in v1.0	Download shapefile directly from Natural Earth

---

Mobina Tooranisama · Data Visualization · Wilfrid Laurier University