---
metadata:
  - name: generator
    content: Diplodoc Platform v5.50.3
alternate:
  - https://catboost.ai/docs/en/concepts/input-data_baseline.md
---
> **Documentation Index:** Fetch the complete configuration index at https://catboost.ai/docs/en/llms.txt

# Baseline

#### Contains

Baseline values for objects in the dataset.

#### Specification

- List the baseline value for each object on a new line.
- Use a tab as the delimiter to separate the columns on a line.
- Use the RawFormulaVal prediction type of the [model values](https://catboost.ai/docs/en/concepts/output-data_model-value-output.md) as the baseline. Other prediction types are not supported.

#### Row format

- The first line contains the description of data in the corresponding column.

  Format:

  {% list tabs %}

    - Non multiclassification methods


      ```
          RawFormulaVal[<\t><other tab-separated columns>]
      ```

      `other tab-separated columns` can be set for auxiliary data (for example, for SampleId).

    - Multiclassification methods

      ```
          RawFormulaVal:Class=<class_name_1><\t>...<\t>RawFormulaVal:Class=<class_name_N>[<\t><other tab-separated columns>]
      ```

      `class_name_1`...`class_name_N` are the names of corresponding classes. The order must match the one specified in the `--classes-count` parameter and in the model.

  {% endlist %}

- All the other lines contain baseline values for the objects of the input dataset. Specify the baseline value for the `i`-th object on the line numbered `i+1`.

  Format:

  {% list tabs %}

    - Non multiclassification methods

      ```
          <value>[<\t><other tab-separated data>]
      ```

    - Multiclassification methods

      ```
          <value for class_name_1><\t>...<\t><value for class_name_N>[<\t><other tab-separated data]
      ```

      `class_name_1`...`class_name_N` are the names of corresponding classes.

  {% endlist %}


#### Example

```
RawFormulaVal
-2.036136239767074585e-01
-2.093112945556640625e+00
-1.437631964683532715e+00
9.149890393018722534e-02
```