Index To String

Maps a column of indices back to a new column of corresponding string values. The index-string mapping is either from the ML attributes of the input column, or from user-supplied labels

Type

ml-estimator

Class

fire.nodes.etl.NodeIndexToString

Fields

Name	Title	Description
inputCol	Input Column	Column containing label indices
outputCol	Output Column	Output column name
labels	Labels Column	eg: yes,no

Details

Index To String Node Details

The Index To String Node is used to map a column of indices back to a new column of corresponding string values. The index-string mapping is either from the ML attributes of the input column, or from user-supplied labels. It takes in a DataFrame and transforms it to another DataFrame by adding a new column containing the corresponding string values of the input column of label indices.

It takes in the parameters inputCol, outputCol and labels, which are used for input column name, output column name and user-supplied labels respectively.

Input Parameters

INPUT COLUMN : Select the required column of label indices for mapping

OUTPUT COLUMN : The name of the output column after mapping

LABELS : The user-supplied labels for the input column of label indices

Examples

Index To String Node Example

Consider the below Index To String output for the label column

id label string_label

0 0 dog

1 1 cat

2 2 bird

In this example, the input column is label and the output column is string_label. The index to string node maps the label column of indices back to a new column of corresponding string values. The labels provided are “dog”, “cat” and “bird” and the index 0 corresponds to “dog”, index 1 corresponds to “cat” and index 2 corresponds to “bird”