mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-09-29 05:00:47 +00:00
14 lines
425 B
Markdown
14 lines
425 B
Markdown
|
---
|
||
|
|
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`.
|