Skip to content
New issue

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

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

CI: Drop numpy 1.6 support #7954

Merged
merged 1 commit into from
Aug 12, 2014
Merged

CI: Drop numpy 1.6 support #7954

merged 1 commit into from
Aug 12, 2014

Conversation

cpcloud
Copy link
Member

@cpcloud cpcloud commented Aug 7, 2014

closes #7711

@cpcloud
Copy link
Member Author

cpcloud commented Aug 7, 2014

well that was cathartic

@cpcloud cpcloud added the Compat label Aug 7, 2014
@cpcloud cpcloud added this to the 0.15.0 milestone Aug 7, 2014
@cpcloud cpcloud self-assigned this Aug 7, 2014
@cpcloud cpcloud added CI labels Aug 7, 2014
@jreback
Copy link
Contributor

jreback commented Aug 7, 2014

fyi, your 2.6/2.7_locale build are prob taking forever as they have to compie numpy (until the wheel is there)

@cpcloud
Copy link
Member Author

cpcloud commented Aug 7, 2014

yeah i know .. there's some fails too

@cpcloud
Copy link
Member Author

cpcloud commented Aug 7, 2014

i think we might need to require 1.7.1, there's an issue with type coercion in ufuncs in 1.7.0 not totally sure what the issue is taking a look now

@cpcloud
Copy link
Member Author

cpcloud commented Aug 7, 2014

arg nope 1.7.1 still fails

@@ -658,11 +658,10 @@ def test_constructor_dtype_datetime64(self):
expected = Series(values2, dates)

# numpy < 1.7 is very odd about astyping
Copy link
Contributor

Choose a reason for hiding this comment

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

this line too

@jreback
Copy link
Contributor

jreback commented Aug 7, 2014

ok wheels are up for 1.7.0/1.7.1 on both 2.6 and 2.7_LOCALE

enjoy!

@cpcloud
Copy link
Member Author

cpcloud commented Aug 9, 2014

I'm going to close this until i make some more progress

@cpcloud cpcloud closed this Aug 9, 2014
@cpcloud cpcloud reopened this Aug 9, 2014
@cpcloud
Copy link
Member Author

cpcloud commented Aug 9, 2014

whoops wrong pr

@cpcloud
Copy link
Member Author

cpcloud commented Aug 11, 2014

This will pass but there's a weird bug where

with tm.assertRaises(ValueError):
    o.item()

will fail but

self.assertRaises(ValueError, o.item)

will pass

there might be a weird py26 bug in this decorator ... let me see if i can't dig around and fix it

@cpcloud
Copy link
Member Author

cpcloud commented Aug 12, 2014

@jreback @jorisvandenbossche Any comments? The assert raises issue only shows up on Travis, even when I use the exact same deps.

@jreback
Copy link
Contributor

jreback commented Aug 12, 2014

yeh I think their might be something in the decorator

I always use:

self.assertRaises(ValueError, lambda : o.item())

from pandas.version import version as __version__
from pandas.info import __doc__


if LooseVersion(_np_version) < '1.7.0':
Copy link
Contributor

Choose a reason for hiding this comment

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

not tested anywhere I assume (as we are not running builds like this), but give a test manually

Copy link
Member Author

Choose a reason for hiding this comment

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

tested this

not totally sure how you'd even get past the setup.py but this will raise if somehow you do

@jreback
Copy link
Contributor

jreback commented Aug 12, 2014

need to update install.rst and big fat warning in v0.15.0 and highlites (also may need to update else where in docs)

@cpcloud
Copy link
Member Author

cpcloud commented Aug 12, 2014

Cool. Will do.

You actually don't need a lambda in those cases because the function is already a no argument callable. A bit easier to read IMO

@cpcloud
Copy link
Member Author

cpcloud commented Aug 12, 2014

@jreback good to go?

@cpcloud
Copy link
Member Author

cpcloud commented Aug 12, 2014

i'll squash if you watn

.. warning::

pandas will no longer support compatibility with NumPy versions < 1.7.0. If
you want to use the latest versions of pandas, please upgrade to NumPy >=
Copy link
Contributor

Choose a reason for hiding this comment

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

add this to highlites too (I know its sort of duplicated but that's ok)

Copy link
Member Author

Choose a reason for hiding this comment

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

ok

Copy link
Member Author

Choose a reason for hiding this comment

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

i just moved it up to right above the "Highlights include"

@jorisvandenbossche
Copy link
Member

version is also mentioned in the readme

@jreback
Copy link
Contributor

jreback commented Aug 12, 2014

minor comments (I would squash). ok otherwise

@jorisvandenbossche

@cpcloud
Copy link
Member Author

cpcloud commented Aug 12, 2014

alright if this next build passes going to merge

@jorisvandenbossche thanks for the heads up about the readme

@jreback
Copy link
Contributor

jreback commented Aug 12, 2014

ok sure

@cpcloud
Copy link
Member Author

cpcloud commented Aug 12, 2014

this is passing on my fork, travis is going to take forever to get to this i guess can merge tomorrow

@jreback
Copy link
Contributor

jreback commented Aug 12, 2014

if passing then ok to merge

@cpcloud cpcloud changed the title WIP: Drop numpy 1.6 support CI: Drop numpy 1.6 support Aug 12, 2014
cpcloud added a commit that referenced this pull request Aug 12, 2014
@cpcloud cpcloud merged commit 534784b into pandas-dev:master Aug 12, 2014
@cpcloud cpcloud deleted the drop-np16 branch August 12, 2014 21:59
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CI Continuous Integration Compat pandas objects compatability with Numpy or Python functions
Projects
None yet
Development

Successfully merging this pull request may close these issues.

COMPAT: Consider dropping numpy 1.6 support
3 participants