---
toc_priority: 250
---

# categoricalInformationValue {#categoricalinformationvalue}

Calculates the value of `(P(tag = 1) - P(tag = 0))(log(P(tag = 1)) - log(P(tag = 0)))` for each category.

``` sql
categoricalInformationValue(category1, category2, ..., tag)
```

The result indicates how a discrete (categorical) feature `[category1, category2, ...]` contribute to a learning model which predicting the value of `tag`.