Normalize Multiple Features

[nick0kcin]

Trying to normalize multiple features with the Multi-Field Formula tool with the following expression (that throws an error):

([_CurrentField_] - Min([_CurrentField_])) / (Max([_CurrentField_]) - Min([_CurrentField_]))

Is there a way to make this work?

[MichalM]

@nick0kcin 

The Min and Max functions in the formula tool (and multi-field formula) will only return result from a list of values not from a referenced variable

Min(10, 20, 30)

In order to calculate, you will need to use the Summarize tool and append these to your data

[nick0kcin]

That's unfortunate, thanks for the information.

[MichalM]

There might be different ways of approaching it. Could you share sample of your data and a bit more context around what you're hoping to achieve?

[nick0kcin]

The operation itself is simple to do in the Python tool for multiple columns (I have hundreds which I'd like to apply the transformation to).

For example: 

https://community.alteryx.com/t5/Alteryx-Designer-Discussions/How-to-do-Feature-Normalization-in-Alt...

I was hoping to save a step for having to select all of the numerical columns in the data first by using the Multi-Field Formula tool's numeric filter and dynamically performing the calculation there.

[MichalM · Accepted solution]

@nick0kcin 

You can use the Dynamic Select to do the same and then a combination of Transpose and Crosstab to make this dynamic. Example attached.

[nick0kcin]

Perfect, thanks.

[nick0kcin]

Slight revision to the equation posted in the example for completeness. Should be: (([Value]-[Min]))/([Max]-[Min])