SciPy 2015 conference birds-of-a-feather session

[shoyer]

This issue is for discussing the organization of a birds-of-a-feather session for the SciPy 2015 conference. Please speak up if you're interested shaping this discussion (and especially if you'll be at the conference).

I'll be at SciPy and would definitely like to make sure this discussion happens. Here are a few possible topics:

Editable doc of the proceedings: http://tinyurl.com/futureofpandas

• What should pandas 1.0 look like? (Towards "pandas 1.0" #10000)
• Is the pandas DataFrame too complex? Is there room/need for a simpler DataFrame implementation? For more context: http://www.slideshare.net/wesm/dataframes-the-good-bad-and-ugly
• do we want out-of-core / parallel support (e.g. how to integrate dask)

[jreback]

myself and @cpcloud will be at SciPy as well.

[jorisvandenbossche]

I won't be at SciPy, but if there would be any possibility to in some way participate remotely, I would like to do that very much (but don't know if that would be possible in the BoF setup)

[shoyer]

Well, it certainly can't hurt to ask for help with remote participation. Even google hangouts works pretty well with a laptop plus an external mike/speaker.

[jorisvandenbossche]

cc @pydata/pandas

[jreback]

@shoyer did we register for a session, topic TBD?

[shoyer]

@jreback not yet, but will do before EOD today

[shoyer]

Submitted. @jreback you are listed as the backup moderator.

[wesm]

I'll be at the conference on the 8th and 9th (and more than happy to offer my $0.02) if we could do it one of those days

[jreback]

awesome @wesm !

[shoyer]

Great -- I just emailed the organizers

[kevindavenport]

Does https://github.com/ContinuumIO/blaze ever need to come into the fold of Pandas core?

[shoyer]

@kevindavenport dask (mentioned in the topics at the top) is part of the Blaze project, so this is definitely within scope for discussion. But Blaze definitely extends beyond the scope of pandas as currently envisioned (and frankly, we barely have to resources to maintain what we already have).

[bwillers]

Would love to be a fly on the wall via hangouts or something similar.

[sinhrks]

I hope I could go... It is nice if we can discuss about scikit-learn integration. Compared to statsmodels, sklearn and pandas is not integrated very well yet.

I understand scikit-learn partially supports DataFrame recently. OTOH I once prepared a POC to integrate sklearn to pandas accessor functionality.

• http://expandas.readthedocs.org/en/stable/sklearn.html

[shoyer]

This is now scheduled for 1-2PM on Thursday, 7/9/15 (US Central Time, UTC-06:00).

I will ask about remote options.

[jreback]

I updated the top section with the link: http://tinyurl.com/futureofpandas

for a shared doc of items to discuss at today's BOF at SciPy. pls feel free to add / edit.

@cpcloud @shoyer @wesm @jorisvandenbossche @TomAugspurger @sinhrks

[shoyer]

Checking on AV options... @jorisvandenbossche you're still interested in dialing in, yes?

[jorisvandenbossche]

yes, if possible. But no problem if it is too much problem! (or just being able to follow and possibly react via chat is also already nice)