> For the complete documentation index, see [llms.txt](https://docs.seldon.ai/alibi-detect/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.seldon.ai/alibi-detect/api-reference/utils/data.md).

# alibi\_detect.utils.data

## `Bunch`

*Inherits from:* `dict`

Container object for internal datasets

Dictionary-like object that exposes its keys as attributes.

### Constructor

```python
Bunch(self, **kwargs)
```

## Functions

### `create_outlier_batch`

```python
create_outlier_batch(data: numpy.ndarray, target: numpy.ndarray, n_samples: int, perc_outlier: int) -> Union[alibi_detect.utils.data.Bunch, Tuple[numpy.ndarray, numpy.ndarray]]
```

Create a batch with a defined percentage of outliers.

| Name           | Type            | Default | Description |
| -------------- | --------------- | ------- | ----------- |
| `data`         | `numpy.ndarray` |         |             |
| `target`       | `numpy.ndarray` |         |             |
| `n_samples`    | `int`           |         |             |
| `perc_outlier` | `int`           |         |             |

**Returns**

* Type: `Union[alibi_detect.utils.data.Bunch, Tuple[numpy.ndarray, numpy.ndarray]]`

### `sample_df`

```python
sample_df(df: pandas.core.frame.DataFrame, n: int)
```

Sample n instances from the dataframe df.

| Name | Type                          | Default | Description |
| ---- | ----------------------------- | ------- | ----------- |
| `df` | `pandas.core.frame.DataFrame` |         |             |
| `n`  | `int`                         |         |             |