32

Is there any commonly used Pandas DataFrame column naming convention? Is PEP8 recommended here (ex. instance variables)?

Concious that lots of data is loaded from external sources with headers but I'm curious what is the correct approach when I have to name/rename the columns on my own?

Improve this question

3 Answers 3

Reset to default
30

Some people tend to use snake_case (lower case with underscores) so that they can access the column using period like this df.my_column

I tend to always access columns using the df['my_column'] syntax because it avoids confusion with DataFrame methods and properties, and it easier to extend to slices and fancy indexing, so the snake case is not necessary.

In short, I think you should use whatever is clearest to a potential reader.

Improve this answer
Sign up to request clarification or add additional context in comments.

Comments

10

One more thing to keep in mind, if your application also makes use of relational databases - I would recommend that you keep your Pandas naming conventions consistent with the column names of your relational database tables.

Improve this answer

1 Comment

e.g. in the case of PostgreSQL and most(?) others, that means all lowercase letters (case sensitivity for columns in databases requires a lot of extra quotes in queries). I ran into this headache when I started using a database after I wrote a ton a pandas code. Seems the cleanest way forward is to go back and change all my pandas code to use lowercase column names.
3

There is no clear guidance from pandas founding fathers and the choice is really between already mentioned snake vs Pascal (camel) case, or df['my_column'] vs df['MyColumn'], and is a matter of preference. A lot of R packages use snake case for dataframes. I'd say snake case is more readable while Pascal case require fewer characters.

Improve this answer

Comments

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.