Feature Selection With Correlation

Output

Pass the input DataFrame to output and in table format correlation value between target and feature columns displayed.

Type

transform

Class

fire.nodes.featureselection.FeatureSelectionWithCorrelation

Fields

Name

Title

Description

targetCol

TargetCol

featureCols

Feature Columns

Details

Feature Selection With Correlation Node Details

This node enables us to discover the possible correlation which can exist between a defined target column and the feature column’s.

Input Parameters

  • OUTPUT STORAGE LEVEL : Keep this as DEFAULT.

  • TARGETCOL : Select the target variable for which we want to explore the existence of a relationship.

  • FEATURE COLUMNS :

  • Available : A list of numeric feature columns derived from the input dataframe schema.

  • Selected : A list of columns for whom the node will compute correlational values against the TARGETCOL.

Examples

Feature Selection With Correlation Node Example

For a given dataframe having the below housing schema:

price  | bathrms | stories| bedrooms|
Double | Double  | Double | Double  |
-------------------------------------

We can select the Target Columns as price and explore the correlation which exists between the target column and the feature columns of bathrms, stories and bedrooms.

This will yield two output sections:

  • A Correlation table showing the correlation value between the target column price and the feature columns &

  • Output of the input dataframe in tabular format