Skip to content

Conversation

mdhaber
Copy link
Contributor

@mdhaber mdhaber commented Feb 25, 2022

Reference issue

closes gh-10096

What does this implement/fix?

This adds a literature reference for stats.circstd.
It also adds comments indicating the equations from the reference that correspond with each line of code, includes the definition of the statistic in the notes, and demonstrates that the property originally mentioned in the notes (similar to linear standard deviation in the small angle limit) in the example.

image

Additional information

A separate PR addressing gh-5747 will add the reference for stats.circvar.

@mdhaber mdhaber added scipy.stats Documentation Issues related to the SciPy documentation. Also check https://github.com/scipy/scipy.org labels Feb 25, 2022
@mdhaber mdhaber requested a review from josef-pkt February 25, 2022 07:54
[skip actions] [skip azp]
[skip actions] [skip azp]
@mdhaber
Copy link
Contributor Author

mdhaber commented Mar 2, 2022

@JoKeyser how does this look to you?

Copy link
Member

@tupui tupui left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, just a suggestion to add the DOI.

(In general we are missing a tutorial on these circ and geom stats.)

[skip actions] [skip azp]

Co-authored-by: Pamphile Roy <roy.pamphile@gmail.com>
@mdhaber
Copy link
Contributor Author

mdhaber commented Mar 4, 2022

(In general we are missing a tutorial on these circ and geom stats.)

I think that they should not be advertised with a tutorial in their current state. For instance, the calculation of stats.circvar is incorrect according to every reference we've found, which I'll fix when this merges. There is also very little here. We could also consider deprecating it all and just relying on astropy for this stuff. I'm conflicted.

@tupui
Copy link
Member

tupui commented Mar 4, 2022

(In general we are missing a tutorial on these circ and geom stats.)

I think that they should not be advertised with a tutorial in their current state. For instance, the calculation of stats.circvar is incorrect according to every reference we've found, which I'll fix when this merges. There is also very little here. We could also consider deprecating it all and just relying on astropy for this stuff. I'm conflicted.

Mmmm yes I also saw that astropy had this. But to me it seems like they did it because we lacked such methods. We can raise this on an issue/mailing list.

@tupui
Copy link
Member

tupui commented Mar 4, 2022

Doc renders good, merging. Thanks again Matt.

@tupui tupui merged commit 8773a0a into scipy:main Mar 4, 2022
@JoKeyser
Copy link

JoKeyser commented Mar 4, 2022

For the record: @mdhaber thank you, from me as well, looks great.

@mdhaber
Copy link
Contributor Author

mdhaber commented Mar 4, 2022

Great, thanks @JoKeyser @tupui!

@mdhaber mdhaber added this to the 1.9.0 milestone Jun 18, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Documentation Issues related to the SciPy documentation. Also check https://github.com/scipy/scipy.org scipy.stats
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Add literature reference for circstd (and circvar?)
3 participants