Imputer

Imputation estimator for completing missing values

Type

ml-estimator

Class

fire.nodes.ml.NodeImputer

Fields

Name	Title	Description
inputCol	Input Column	Column containing inputs
outputCol	Output Column	Output column names
strategy	Strategy	Imputer strategy

Details

Imputer Node Details

The Imputer Node is used to complete missing values in a dataset. It takes in a DataFrame and transforms it to another DataFrame by filling the missing values in the input columns.

It takes in the parameters inputCols, outputCols and strategy, which are used for input column names, output column names and the imputation strategy respectively.

Input Parameters

INPUT COLUMNS : Select the required columns for imputation.

OUTPUT COLUMNS : The name of the output columns after imputation.

STRATEGY : The imputation strategy can be mean, median or mode.

Examples

Imputer Node Example

Consider the below Imputer output for the age and income columns with strategy mean.

id age income

0 20 50000

1 null 60000

2 30 null

The imputed Dataframe will be

id age income

0 20 50000

1 25 60000

2 30 58333.3333

In this example, the input columns are age and income and the output columns are also age and income. The imputer fills the missing values in the input columns with the mean of the column. Here the mean of age column is 25 and mean of income column is 58333.3333