Is this the fastest way to group in Pandas?

If you’d like to install the iPython shell, you can easily time your code using %timeit. After installing it, instead of typing python to launch the python interpreter, you would type ipython.

Then you can type your code exactly as you would type it in the normal interpreter (as you did above).

Then you can type, for example:

%timeit DF.groupby(['id1']).agg({'v1':'sum'})

This will accomplish exactly the same thing as what you’ve done, but if you’re using python a lot I find that this will save you significant typing time :).

Ipython has a lot of other nice features (like %paste, which I used to paste in your code and test this, or %run to run a script you’ve saved in a file), tab completion, etc.
http://ipython.org/

Leave a Comment