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

# Utils | MLForecast

```python theme={null}
from fastcore.test import test_eq, test_fail
from nbdev import show_doc
```

### `generate_daily_series`

```python theme={null}
generate_daily_series(n_series, min_length=50, max_length=500, n_static_features=0, equal_ends=False, static_as_categorical=True, with_trend=False, seed=0, engine='pandas')
```

Generate Synthetic Panel Series.

**Parameters:**

| Name | Type | Description | Default |
| - | - | - | - |
| `n_series` | <code>[int](#int)</code> | Number of series for synthetic panel. | *required* |
| `min_length` | <code>int, default=50</code> | Minimum length of synthetic panel's series. | <code>50</code> |
| `max_length` | <code>int, default=500</code> | Maximum length of synthetic panel's series. | <code>500</code> |
| `n_static_features` | <code>int, default=0</code> | Number of static exogenous variables for synthetic panel's series. | <code>0</code> |
| `equal_ends` | <code>bool, default=False</code> | Series should end in the same date stamp `ds`. | <code>False</code> |
| `static_as_categorical` | <code>bool, default=True</code> | Static features should have a categorical data type. | <code>True</code> |
| `with_trend` | <code>bool, default=False</code> | Series should have a (positive) trend. | <code>False</code> |
| `seed` | <code>int, default=0</code> | Random seed used for generating the data. | <code>0</code> |
| `engine` | <code>str, default='pandas'</code> | Output Dataframe type. | <code>'pandas'</code> |

**Returns:**

| Type | Description |
| - | - |
| <code>[DataFrame](#utilsforecast.compat.DataFrame)</code> | pandas or polars DataFrame: Synthetic panel with columns \[`unique_id`, `ds`, `y`] and exogenous features. |

Generate 20 series with lengths between 100 and 1,000.

```python theme={null}
n_series = 20
min_length = 100
max_length = 1000

series = generate_daily_series(n_series, min_length, max_length)
series
```

| | unique\_id | ds | y |
| - | - | - | - |
| 0 | id\_00 | 2000-01-01 | 0.395863 |
| 1 | id\_00 | 2000-01-02 | 1.264447 |
| 2 | id\_00 | 2000-01-03 | 2.284022 |
| 3 | id\_00 | 2000-01-04 | 3.462798 |
| 4 | id\_00 | 2000-01-05 | 4.035518 |
| ... | ... | ... | ... |
| 12446 | id\_19 | 2002-03-11 | 0.309275 |
| 12447 | id\_19 | 2002-03-12 | 1.189464 |
| 12448 | id\_19 | 2002-03-13 | 2.325032 |
| 12449 | id\_19 | 2002-03-14 | 3.333198 |
| 12450 | id\_19 | 2002-03-15 | 4.306117 |

We can also add static features to each serie (these can be things like
product\_id or store\_id). Only the first static feature (`static_0`) is
relevant to the target.

```python theme={null}
n_static_features = 2

series_with_statics = generate_daily_series(n_series, min_length, max_length, n_static_features)
series_with_statics
```

| | unique\_id | ds | y | static\_0 | static\_1 |
| - | - | - | - | - | - |
| 0 | id\_00 | 2000-01-01 | 7.521388 | 18 | 10 |
| 1 | id\_00 | 2000-01-02 | 24.024502 | 18 | 10 |
| 2 | id\_00 | 2000-01-03 | 43.396423 | 18 | 10 |
| 3 | id\_00 | 2000-01-04 | 65.793168 | 18 | 10 |
| 4 | id\_00 | 2000-01-05 | 76.674843 | 18 | 10 |
| ... | ... | ... | ... | ... | ... |
| 12446 | id\_19 | 2002-03-11 | 27.834771 | 89 | 42 |
| 12447 | id\_19 | 2002-03-12 | 107.051746 | 89 | 42 |
| 12448 | id\_19 | 2002-03-13 | 209.252845 | 89 | 42 |
| 12449 | id\_19 | 2002-03-14 | 299.987801 | 89 | 42 |
| 12450 | id\_19 | 2002-03-15 | 387.550536 | 89 | 42 |

```python theme={null}
for i in range(n_static_features):
    assert all(series_with_statics.groupby('unique_id')[f'static_{i}'].nunique() == 1)
```

If `equal_ends=False` (the default) then every serie has a different end
date.

```python theme={null}
assert series_with_statics.groupby('unique_id')['ds'].max().nunique() > 1
```

We can have all of them end at the same date by specifying
`equal_ends=True`.

```python theme={null}
series_equal_ends = generate_daily_series(n_series, min_length, max_length, equal_ends=True)

assert series_equal_ends.groupby('unique_id')['ds'].max().nunique() == 1
```

***

### `generate_prices_for_series`

```python theme={null}
generate_prices_for_series(series, horizon=7, seed=0)
```

```python theme={null}
series_for_prices = generate_daily_series(20, n_static_features=2, equal_ends=True)
series_for_prices.rename(columns={'static_1': 'product_id'}, inplace=True)
prices_catalog = generate_prices_for_series(series_for_prices, horizon=7)
prices_catalog
```

| | ds | unique\_id | price |
| - | - | - | - |
| 0 | 2000-10-05 | id\_00 | 0.548814 |
| 1 | 2000-10-06 | id\_00 | 0.715189 |
| 2 | 2000-10-07 | id\_00 | 0.602763 |
| 3 | 2000-10-08 | id\_00 | 0.544883 |
| 4 | 2000-10-09 | id\_00 | 0.423655 |
| ... | ... | ... | ... |
| 5009 | 2001-05-17 | id\_19 | 0.288027 |
| 5010 | 2001-05-18 | id\_19 | 0.846305 |
| 5011 | 2001-05-19 | id\_19 | 0.791284 |
| 5012 | 2001-05-20 | id\_19 | 0.578636 |
| 5013 | 2001-05-21 | id\_19 | 0.288589 |

```python theme={null}
test_eq(set(prices_catalog['unique_id']), set(series_for_prices['unique_id']))
test_fail(lambda: generate_prices_for_series(series), contains='equal ends')
```

***

### `PredictionIntervals`

```python theme={null}
PredictionIntervals(n_windows=2, h=1, method='conformal_distribution', scale_estimator=None)
```

Class for storing prediction intervals metadata information.


## Related topics

- [Core | MLForecast](/mlforecast/core.html.md)
- [Feature Engineering | UtilsForecast](/utilsforecast/feature_engineering.html.md)
- [Utils | DatasetsForecast](/datasetsforecast/utils.html.md)
- [Electricity Load Forecast | MLForecast](/mlforecast/docs/tutorials/electricity_load_forecasting.html.md)
- [MLForecast](/mlforecast/forecast.html.md)