pandas.util.testing no longer imported with pandas

[TomAugspurger]

0.25.3

In [1]: import pandas

In [2]: pandas.util.testing
Out[2]: <module 'pandas.util.testing' from '/private/tmp/pandas/util/testing.py'>

1.0

In [1]: import pandas

In [2]: pandas.util.testing
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-2-968e027d039f> in <module>
----> 1 pandas.util.testing

AttributeError: module 'pandas.util' has no attribute 'testing'

[jreback]

hmm it i don’t think is a problem
and wouldn’t do anything about it

[jorisvandenbossche]

It's perfect legitimate usage pattern of util.testing, and it's also easy to fix (we can still import it while hiding the warning)

[jreback]

It's perfect legitimate usage pattern of util.testing, and it's also easy to fix (we can still import it while hiding the warning)

missing the point - why should be supporting a non explicit import? i would simply break this

[jorisvandenbossche]

why should be supporting a non explicit import?

Because we have always supported this?

[jreback]

why should be supporting a non explicit import?

Because we have always supported this?

i disagree

it doesn’t matter if it’s easy

[jorisvandenbossche]

Because we have always supported this?

i disagree

I am assuming that your "I disagree" is about this being a good reason to keep supporting it? Because AFAIK, it's a fact that we have always supported this.

[jreback]

Because we have always supported this?

i disagree

I am assuming that your "I disagree" is about this being a good reason to keep supporting it? Because AFAIK, it's a fact that we have always supported this.

my point is that @TomAugspurger did enough on the current warning for an explicit import

an implicit import is not a reason to even have a depreciation

so -1 for fixing this and a hard break is fine

[TomAugspurger]

I don’t think it’ll be hard. Similar to the np deprecation.

…

On Jan 10, 2020, at 04:57, Jeff Reback ***@***.***> wrote:  Because we have always supported this? i disagree I am assuming that your "I disagree" is about this being a good reason to keep supporting it? Because AFAIK, it's a fact that we have always supported this. my point is that @TomAugspurger did enough on the current warning for an explicit import an implicit import is not a reason to even have a depreciation so -1 for fixing this and a hard break is fine — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.