> ## Documentation Index
> Fetch the complete documentation index at: https://nixtla-old-docs.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

> Download and evaluate the M4 dataset.

# M4 dataset

> M4 meta information

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L68" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### Other

> ```text theme={null}
>  Other (seasonality:int=1, horizon:int=8, freq:str='D', name:str='Other',
>         n_ts:int=5000, included_groups:Tuple=('Weekly', 'Daily',
>         'Hourly'))
> ```

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L59" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### Hourly

> ```text theme={null}
>  Hourly (seasonality:int=24, horizon:int=48, freq:str='H',
>          name:str='Hourly', n_ts:int=414)
> ```

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L51" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### Daily

> ```text theme={null}
>  Daily (seasonality:int=1, horizon:int=14, freq:str='D', name:str='Daily',
>         n_ts:int=4227)
> ```

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L43" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### Weekly

> ```text theme={null}
>  Weekly (seasonality:int=1, horizon:int=13, freq:str='W',
>          name:str='Weekly', n_ts:int=359)
> ```

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L35" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### Monthly

> ```text theme={null}
>  Monthly (seasonality:int=12, horizon:int=18, freq:str='M',
>           name:str='Monthly', n_ts:int=48000)
> ```

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L27" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### Quarterly

> ```text theme={null}
>  Quarterly (seasonality:int=4, horizon:int=8, freq:str='Q',
>             name:str='Quarterly', n_ts:int=24000)
> ```

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L19" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### Yearly

> ```text theme={null}
>  Yearly (seasonality:int=1, horizon:int=6, freq:str='Y',
>          name:str='Yearly', n_ts:int=23000)
> ```

## Download data class

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L81" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### M4

> ```text theme={null}
>  M4 (source_url:str='https://raw.githubusercontent.com/Mcompetitions/M4-
>      methods/master/Dataset/', naive2_forecast_url:str='https://github.com
>      /Nixtla/m4-forecasts/raw/master/forecasts/submission-Naive2.zip')
> ```

```python theme={null}
group = 'Hourly'
await M4.async_download('data', group=group)
df, *_ = M4.load(directory='data', group=group)
n_series = len(np.unique(df.unique_id.values))
display_str  = f'Group: {group} '
display_str += f'n_series: {n_series}'
print(display_str)
```

## Evaluation class

***

<a href="https://github.com/Nixtla/datasetsforecast/blob/main/datasetsforecast/m4.py#L230" target="_blank" style={{ float: "right", fontSize: "smaller" }}>source</a>

### M4Evaluation

> ```text theme={null}
>  M4Evaluation ()
> ```

*Initialize self. See help(type(self)) for accurate signature.*

### URL-based evaluation

The method `evaluate` from the class
[`M4Evaluation`](https://Nixtla.github.io/datasetsforecast/m4.html#m4evaluation)
can receive a url of a [benchmark uploaded to the M4
competiton](https://github.com/Mcompetitions/M4-methods/tree/master/Point%20Forecasts).

The results compared to the on-the-fly evaluation were obtained from the
[official
evaluation](https://github.com/Mcompetitions/M4-methods/blob/master/Evaluation%20and%20Ranks.xlsx).

```python theme={null}
from fastcore.test import test_close
```

```python theme={null}
esrnn_url = 'https://github.com/Nixtla/m4-forecasts/raw/master/forecasts/submission-118.zip'
esrnn_evaluation = M4Evaluation.evaluate('data', 'Hourly', esrnn_url)
# Test of the same evaluation as the original one
test_close(esrnn_evaluation['SMAPE'].item(), 9.328, eps=1e-3)
test_close(esrnn_evaluation['MASE'].item(), 0.893, eps=1e-3)
test_close(esrnn_evaluation['OWA'].item(), 0.440, eps=1e-3)
esrnn_evaluation
```

### Numpy-based evaluation

Also the method `evaluate` can recevie a numpy array of forecasts.

```python theme={null}
fforma_url = 'https://github.com/Nixtla/m4-forecasts/raw/master/forecasts/submission-245.zip'
fforma_forecasts = M4Evaluation.load_benchmark('data', 'Hourly', fforma_url)
fforma_evaluation = M4Evaluation.evaluate('data', 'Hourly', fforma_forecasts)
# Test of the same evaluation as the original one
test_close(fforma_evaluation['SMAPE'].item(), 11.506, eps=1e-3)
test_close(fforma_evaluation['MASE'].item(), 0.819, eps=1e-3)
test_close(fforma_evaluation['OWA'].item(), 0.484, eps=1e-3)
fforma_evaluation
```