Imputation Tool Error

Hi there,

The imputation tool is reducing the number of columns in my data set. I am feeding in 100+ columns but only looking to impute 5 of them. The output of the tool has < 100 columns which seem to be dropped at random. Unless I am using the tool incorrectly, can you please advise on this?

Thanks,
David

Common Use Cases

Accepted answers

JessicaS

Hi @dcleere,

I was able to take a look here. Under the hood, the tool has a regex statement to deselect the imputed value indicator field and the secondary imputed values field when the below settings are checked.

The regex inside the dynamic select tool is looking for fields that contain _ followed by a string containing either mp or nd (the fields created by the tool have suffix of either _Indicator or _ImputedValue)

It just so happens that your field names match the regex written with _INDUST.

To Navigate around, rename the fields to a name that does not contain 'IN' after an _ such as INDUST_ERNINGS_AGE or select output 'imputed value indicator.'

@MarqueeCrew's suggestion is a great workaround as well.

I will alert our development team about this issue.

Thanks,

All comments

JessicaS

HI @dcleere,

I've moved your post over to data preparation and blending. Can you post a sample workflow for us to take a look at ?

Thanks,

MarqueeCrew

@dcleere,

If the imputation is simple, I prefer to use the multi-field formula tool.

As an example, to remove NULL values from string variables and make them blank

IIF(IsNull([_CurrentField_]),'',[_CurrentField_])

or to make numeric variables that are NULL as zero (0)

IIF(IsNull([_CurrentField_]),0,[_CurrentField_])

In the configuration, you choose the type of variables to operate on and then select the fields. You can either create "NEW" variables or update them directly.

Cheers,

Mark

dcleere

Hi Jessica,

Please find sample workflow attached. It is mock data and the earlier steps in the workflow mock what is contained in the workflow.

- Select unqiue on Col1

- Impute the last three columns in the data set

You will see the number of columns drops from 113 to 111 following the imputatation tool.

Thanks,

David

Imputation_Defect.yxmd

Quick Links

This months top contributors

atcodedog05 19598

Qiu 15867

binu_acs 15708

MarqueeCrew 13708

apathetichell 13703