ENH: Rename get_dummies to more inclusive language

[davidcavazos]

Feature Type

• Adding new functionality to pandas

• Changing existing functionality in pandas

• Removing existing functionality in pandas

Problem Description

The word "dummy" from the pd.get_dummies function can be offensive to some people and should be renamed.

It's marked as a word that should not be used by Google's inclusive language word list.

Feature Description

A good alternative name could be renaming it to pd.get_indicator_variables, which would also be more explicit on what it does.

Alternative Solutions

Alternatively, pd.get_one_hot or pd.get_one_hot_encoded could also be an option familiar to Machine Learning practitioners.

Google trends show "indicator variable" and "one-hot encoding" to be similarly popular, with "indicator variable" being slightly more popular.

Additional Context

No response

[MarcoGorelli]

Hey @davidcavazos

I'd also be favourable to renaming it, but this was discussed in #35724 and at the time rejected

I think there isn't a settled alternative to dummy, and until there is no change should be made. One the world converges into something that mostly stops using dummy variable, then that should be adopted. More or less how the master->main change worked in pandas.

[MarcoGorelli]

Closing for now then, if the world does converge on an alternative which would allow for #35724 to be reconsidered then we can go with that

[davidcavazos]

The world won't converge on an alternative if we don't start doing the change. GitHub's master -> main only happened because they took the initiative.

[TheNeuralBit]

My takeaway from the previous discussion is that adding a separate get_indicators (or some other agreed upon alternative) would be amenable. From there we could either:

• Deprecate and ultimately remove get_dummies, or
• Prefer get_indicators in documentation to nudge users there

It seems the former was rejected, but the latter could be acceptable. Could we pursue that approach?

(Also, would you prefer for us to continue discussion in #35274?)

[TheNeuralBit]

A non-Google reference for "dummy" being non-inclusive: https://itconnect.uw.edu/guides-by-topic/identity-diversity-inclusion//inclusive-language-guide/

The origin of the word, “dummy,” is a person who cannot speak. Because the use of this word is often negatively associated with a disability, implying a person is worthless, ineffective or incapable, an alternative word should be used.

[MarcoGorelli]

A non-Google reference for "dummy" being non-inclusive: https://itconnect.uw.edu/guides-by-topic/identity-diversity-inclusion//inclusive-language-guide/

The origin of the word, “dummy,” is a person who cannot speak. Because the use of this word is often negatively associated with a disability, implying a person is worthless, ineffective or incapable, an alternative word should be used.

That's a nice reference, thanks!

(Also, would you prefer for us to continue discussion in #35274?)

Yes please, let's keep the discussion in one place - perhaps post the reference you linked above there?

[davidcavazos]

Thanks @TheNeuralBit. I think that either get_indicators (as per the previous discussion), or get_indicator_variables (longer, but more explicit), would work. For consistency with other pandas functions with usually shorter names, I think get_indicators might be more suitable.

I would personally create the new name and mark get_dummies as deprecated (but still usable), and finally remove it completely in a future release.

[davidcavazos]

The University of Delaware also has a similar (although shorter) list including to remove "dummy value":

http://www1.udel.edu/itwebdev/help/dei.html

[davidcavazos]

And here are some more examples:

• University of Guelph
• Research and Education Networks
• University of California

And even the National Institute of Standards and Technology

[davidcavazos]

This document also mentions how it causes harm:

“Dummy” and similar terms stigmatize mental disabilities. The alternatives are clearer.

[MarcoGorelli]

Thanks @davidcavazos , appreciate the references - could you post them in #35724 please so we keep the discussion in one place?

[davidcavazos]

Sure, I just posted a summary of the key points.