electricsheepafrica/nigerian_energy_and_utilities_energy_theft_anomalies

# Nigerian Energy & Utilities – Energy Theft & Anomalies

Anomaly events (tamper, bypass, reversed energy, spikes, zero consumption) with z-scores and flags.

- **[category]** Smart Meters & Consumption
- **[rows]** ~80,000
- **[formats]** CSV + Parquet (snappy)
- **[geography]** Nigeria (DisCos, substations, plants)

## Schema

| column | dtype |

|---|---| | event_id | object | | meter_id | object | | disco | object | | timestamp | object | | anomaly_type | object | | expected_kwh | float64 | | observed_kwh | float64 | | delta_kwh | float64 | | zscore | float64 | | flagged | bool | | resolution_status | object | | lat | float64 | | lon | float64 |

## Usage

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

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