electricsheepafrica/nigerian_energy_and_utilities_renewable_forecasts

# Nigerian Energy & Utilities – Renewable Forecasts

Forecast vs actual for solar/wind sites across horizons with MAPE.

- **[category]** Renewable & Environmental
- **[rows]** ~150,000
- **[formats]** CSV + Parquet (snappy)
- **[geography]** Nigeria (DisCos, substations, plants)

## Schema

| column | dtype |

|---|---| | timestamp | object | | site_id | object | | type | object | | horizon_h | int64 | | forecast_mw | float64 | | actual_mw | float64 | | error_mw | float64 | | mape_pct | float64 |

## Usage

```python
import pandas as pd
df = pd.read_parquet('data/nigerian_energy_and_utilities_renewable_forecasts/nigerian_energy_and_utilities_renewable_forecasts.parquet')
df.head()
```

```python
from datasets import load_dataset
ds = load_dataset('electricsheepafrica/nigerian_energy_and_utilities_renewable_forecasts')
ds
```

## Notes

- Data generated with Nigeria-specific parameters (DisCos, tariff bands, 50 Hz grid)
- Time-of-use shapes and seasonal/weather effects included where applicable
- Values are internally consistent (e.g., kWh ~ kW*h; voltage/current ~ power)