Two years of hourly energy data (2023–2024) explored with Python (pandas, matplotlib). Covers seasonal trends, daily load profiles, and year-over-year comparisons. Includes Jupyter notebook, Word report, and an interactive React dashboard with actionable savings insights.
├── electricity_data.csv # Synthetic dataset — 17,544 hourly readings
├── electricity_analysis.ipynb # Jupyter notebook (full walkthrough)
├── electricity_analysis_report.pdf # Technical Word report with embedded charts
├── fig1_overview.png # Overview dashboard figure
├── fig2_heatmap.png # Hour × month heatmap
├── fig3_seasonal.png # Seasonal load profiles
└── fig4_yoy.png # Year-over-year comparison
The dataset (electricity_data.csv) is a synthetic hourly time series generated to reflect realistic household electricity consumption patterns.
| Column | Description |
|---|---|
timestamp |
Hourly datetime (2023-01-01 00:00 → 2024-12-31 23:00) |
kwh |
Electricity consumed in that hour (kilowatt-hours) |
date |
Date component |
hour |
Hour of day (0–23) |
month |
Month number (1–12) |
month_name |
Abbreviated month name (Jan–Dec) |
year |
Year (2023 or 2024) |
day_of_week |
Full day name (Monday–Sunday) |
is_weekend |
Boolean — True for Saturday and Sunday |
season |
Season label (Spring / Summer / Fall / Winter) |
Modelled patterns include: seasonal HVAC demand, morning and evening routine peaks, weekend occupancy premium, always-on base load, and random appliance spikes.
| Finding | Value |
|---|---|
| Total consumption (2 years) | 6,487 kWh |
| Estimated total cost | $908 (at $0.14/kWh) |
| Daily average | 8.9 kWh |
| Peak demand hour | 19:00 (7 PM) |
| Highest-demand season | Summer (~30% of annual use) |
| Year-over-year change | +0.5% (2024 vs 2023) |
| Weekend vs weekday premium | ~15% more per hour |
| Figure | Description |
|---|---|
fig1_overview.png |
6-panel dashboard: monthly bars, KPI cards, seasonal donut, day-of-week bars, daily load curve |
fig2_heatmap.png |
Hour × month heatmap — identifies when each month's peak hours fall |
fig3_seasonal.png |
Four seasonal load profiles with within-season peak markers |
fig4_yoy.png |
Monthly delta bars and cumulative consumption curves for 2023 vs 2024 |
- All data is synthetic — generated to model realistic patterns, not collected from a real meter.
- Cost estimates assume a flat rate of $0.14/kWh. Actual rates vary by provider and location.
- The analysis was conducted using Python 3.10+, pandas, numpy, and matplotlib.