Suggestion for page /03.01-descriptives.html

[robmoore]

Thanks for your work on translating the book to use Python! I've found it useful.

I wanted to suggest an alternative to the current implementation for calculating "blowouts".

Instead of using numpy's where function, like this:

afl_margins['blowouts'] = np.where(afl_margins['afl.margins'] > 50, True, False)
afl_margins.head()

you could use a pure pandas approach and drop the redundant np.where like so:

afl_margins['blowouts'] = afl_margins['afl.margins'] > 50
afl_margins.head()

The where function is redundant since the evaluation of afl_margins['afl.margins'] > 50 already returns a Boolean value.

[ethanweed]

Thanks for the encouragement, Rob!

I like your suggestion. As I have been going back over the first chapters that I translated, I have been trying to streamline and consolidate things, e.g. only using pingouin to calculate stats when possible, rather than a hodgepodge of numpy and scipy and statmodels. Your suggestion fits in with this approach, so I’ll probably add it. /Ethan

On 28 May 2023, at 23.00, Rob Moore @.***> wrote:

Thanks for your work on translating the book to use Python! I've found it useful.

I wanted to suggest an alternative to the current implementation for calculating "blowouts".

Instead of using numpy's where function, like this:

afl_margins['blowouts'] = np.where(afl_margins['afl.margins'] > 50, True, False) afl_margins.head() you could use a more pandas-like approach and drop np.where like so:

afl_margins['blowouts'] = afl_margins['afl.margins'] > 50 afl_margins.head() — Reply to this email directly, view it on GitHub https://github.com/ethanweed/pythonbook/issues/16, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACKG7GQL5EUZLU2NQWPFRXDXIO4H7ANCNFSM6AAAAAAYSCCMVY. You are receiving this because you are subscribed to this thread.