Skip to content

Conversation

preddy5
Copy link

@preddy5 preddy5 commented Oct 26, 2015

@MechCoder
Copy link
Member

Can you add tests?

@preddy5
Copy link
Author

preddy5 commented Oct 28, 2015

Ya, working on it

@giorgiop
Copy link
Contributor

Shall we update the docstring as well?

@ogrisel ogrisel changed the title issue 4931 Fix NearestNeighbors algorithm='auto' to work with all supported metrics by default Oct 28, 2015
@ogrisel
Copy link
Member

ogrisel commented Oct 28, 2015

Yes the docstring needs to be updated. I changed the title of this PR to make it more explicit. @pradyu1993 please add a [MRG] prefix to the title of this PR once you are done addressing the comments.

@ogrisel ogrisel changed the title Fix NearestNeighbors algorithm='auto' to work with all supported metrics by default [WIP] Fix NearestNeighbors algorithm='auto' to work with all supported metrics by default Oct 28, 2015
@preddy5 preddy5 changed the title [WIP] Fix NearestNeighbors algorithm='auto' to work with all supported metrics by default [MRG] Fix NearestNeighbors algorithm='auto' to work with all supported metrics by default Oct 29, 2015
@preddy5
Copy link
Author

preddy5 commented Nov 1, 2015

@MechCoder can you review the PR

@jnothman
Copy link
Member

jnothman commented Nov 1, 2015

Please also test with sparse matrix input

@preddy5 preddy5 force-pushed the iss4931 branch 2 times, most recently from 63ad4e4 to b969bf6 Compare November 2, 2015 05:31
@preddy5
Copy link
Author

preddy5 commented Nov 3, 2015

@jnothman Done

@jnothman
Copy link
Member

jnothman commented Nov 3, 2015

I intended that you would check that metrics which are supported for dense data but not for sparse data will still work in auto. Or is that covered in another test?

@preddy5
Copy link
Author

preddy5 commented Dec 7, 2015

@MechCoder can you review the PR

Xcsr = csr_matrix(X)
Valid_Metrics = dict(dense = ['precomputed', 'sqeuclidean',
'yule', 'cosine', 'correlation'],
sparse=PAIRWISE_DISTANCE_FUNCTIONS.keys())
Copy link
Member

Choose a reason for hiding this comment

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

You can do away with this and just check for VALID_METRICS['brute'] and VALID_METRICS_SPARSE['brute']. I think the other options are checked already.

@MechCoder
Copy link
Member

Thanks ! Just some minor comments. And then +1

But you might have to wait till after the holiday season for the second ^.^

@jnothman
Copy link
Member

@pradyu1993 are you still working on this?

@jnothman
Copy link
Member

jnothman commented Oct 8, 2016

This is up for grabs by a contributor as far as I'm concerned.

@jnothman jnothman added the Easy Well-defined and straightforward way to resolve label Oct 8, 2016
@amueller amueller added this to the 0.19 milestone Oct 11, 2016
@maniteja123
Copy link
Contributor

Hi @jnothman @amueller I can cherry pick this commit in case no one is working and do it as part of #7590 . Would it be fine ?

@jnothman
Copy link
Member

I'd rather this be patched separate to #7590...

On 11 October 2016 at 17:47, Maniteja Nandana notifications@github.com
wrote:

Hi @jnothman https://github.com/jnothman @amueller
https://github.com/amueller I can cherry pick this commit in case no
one is working and do it as part of #7590
#7590 . Would it be
fine ?


You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#5596 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAEz61XAlHHbexSm5CUvNcwMyYI2wRjbks5qyzDlgaJpZM4GVpFN
.

@amueller
Copy link
Member

👍 for small PRs, also they seem pretty unrelated (or I'm missing something, which is entirely plausible after reading like 5000 notifications this week)

@amueller
Copy link
Member

fixed by #9145.

@amueller amueller closed this Jun 19, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Easy Well-defined and straightforward way to resolve
Projects
None yet
Development

Successfully merging this pull request may close these issues.

7 participants