Imputation Tool
Use Imputation to replace a specified value within 1 or more numeric data columns with another specified value.
A typical use case is to replace all NULL()
values with the average of the remaining values in the columns so the NULL() values do not affect the final outcome of your forecasting model.
Configure the Tool
Numeric Columns to Impute: Select the columns you want to update.
Value to Replace: Specify 1 of 2 choices.
Null(s)
Custom: Use the text box to enter a value.
Replace With: Specify 1 of 4 choices.
Average: Average of all values, not including the Incoming Value to Replace.
Median: Value separating the higher half from the lower half, or the middle value.
Mode: Value that appears most often.
Custom: Use the text box to enter a value.
Select the Output Options:
Include Imputed Value Indicator Column: Add a column to specify whether or not a value has been imputed.
Output Imputed Values as a Separate Column: Leave the original column to impute as-is and add a new column for the imputed value.
View the Output
The default output includes the original data stream with the imputed values embedded within the original data.
If you selected Include Imputed Value Indicator Column, there is an additional column next to each column selected for imputation with the prefix as the original field name and a suffix of
_Indicator
.If Output Imputed Values as a Separate Column is selected, there is an additional column next to each column selected for imputation with the prefix as the original field name and a suffix of
_ImputedValue
.