Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Already on GitHub? Sign in to your account

stats: deprecate and remove alias for special functions #2741

Closed
josef-pkt opened this Issue Aug 18, 2013 · 5 comments

Comments

Projects
None yet
4 participants
Member

josef-pkt commented Aug 18, 2013

label: cleanup

I think we can deprecate and remove alias for special functions that are defined in scipy.stats.stats

zprob = special.ndtr
ksprob = special.kolmogorov
fprob = special.fdtrc
plus betai
and chisqprob

using the corresponding stats.distributions functions is much more explicit
zprob = stats.norm.sf
...
only ksprob might be not obvious, but I doubt anyone uses that one directly

If a user wants to save on the overhead of stats.distributions, then they can use the scipy.special functions directly.

There might still be a few cases where these functions are used in scipy.stats IIRC.

(Aside: I was thinking of adding some of these aliases to statsmodels, when I don't want to import scipy.stats and half of scipy with it when I only need scipy.special. But convenience and readability have won out so far.
This argument doesn't apply when we are already in scipy.stats)

Owner

rgommers commented Aug 19, 2013

+1 for cleaning up stats further.

Owner

rgommers commented Dec 30, 2013

It looks to me like betai (functionally) and chisqprob (name clarity) still provide something useful. Should we keep those?

@rgommers rgommers closed this in #3243 Jan 31, 2014

Contributor

larsmans commented Jul 24, 2015

(Sorry, late to the game)

+1 for keeping chisqprob. I would never have found that function by the name chdtrc.

Member

ewmoore commented Jul 24, 2015

Its other name in the stats namespace is stats.chi2.sf, if that makes a
difference.

On Fri, Jul 24, 2015 at 9:57 AM, Lars notifications@github.com wrote:

(Sorry, late to the game)

+1 for keeping chisqprob. I would never have found that function by the
name chdtrc.


Reply to this email directly or view it on GitHub
#2741 (comment).

Contributor

larsmans commented Jul 24, 2015

As @rgommers said over at #641, the function wasn't in the reference so I probably found it by git grep last time I used it. I guess it's ok to remove it then. (This does raise the issue that I always had with scipy.stats, that its method names are so short and cryptic...)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment