CLN, DOC: Remove unused parameters

[Michael-E-Rose]

• closes #xxxx
• tests added / passed
• passes git diff upstream/master | flake8 --diff
• remove unused parameters out and dtype from cumulated functions (cummin, cumsum, cumprod, cummax)

This also affects the documentation. Check for example: http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.cumsum.html, where the two parameters are mentioned, but not explained.

[jorisvandenbossche]

@gfyoung This is OK? I think dtype and out are already checked in validate_cum_func (https://github.com/pydata/pandas/blob/23eb483d17ce41c0fe41b0bfa72c90df82151598/pandas/compat/numpy/function.py#L148)

[gfyoung]

I would like to say "LGTM", but unfortunately, Travis disagrees. You will have to edit validate_cum_func and replace the two arguments with *args for this to work with older numpy versions.

You probably should also then do some pre-processing on axis and skipna to make sure that they aren't being given numpy parameters (which should then be validated while axis and skipna are given their default values).

Please look at pandas/compat/numpy/function.py and pandas/util/validators.py to see how the validation system works as well as examples.

@Michael-E-Rose : Sorry if this is more complicated than you had expected! You unfortunately walked into a compatibility minefield with these functions. Let me know if you have any questions!

[Michael-E-Rose]

My nosetests run 0 tests - quite useless.

Being in root, I did as explained here, but all I get is

/usr/local/lib/python2.7/dist-packages/nose/importer.py:94: FutureWarning: The pandas.rpy module is deprecated and will be removed in a future version. We refer to external packages like rpy2. 
See here for a guide on how to port your code to rpy2: http://pandas.pydata.org/pandas-docs/stable/r_interface.html
  mod = load_module(part_fqname, fh, filename, desc)

----------------------------------------------------------------------
Ran 0 tests in 0.023s

OK

[jreback]

you need to make sure that you build in-place and run them there. Not sure what root is in this context, are you actually running as root?

e.g. something like

git clone https:://www.github.com/pydata/pandas.git
cd pandas
python setup.py build_ext --inplace
nosetests -A 'not slow' pandas

[Michael-E-Rose]

Root of the repo.

But my build process has quite a long output as well. Mostly warnings about outdated stuff etc.

[gfyoung]

@Michael-E-Rose : Because this is actually a numpy compat issue, I would suggest building in a separate environment (e.g. conda or virtualenv) so that you can swap out the various numpy versions (e.g. 1.11.0, 1.10.4, etc.) for testing. Then just do the following in your top pandas directory:

> python setup.py install
> cd .. // or anywhere else outside of your `pandas` directory
> python -c "import pandas as pd;pd.test(raise_warnings=())"

[jreback]

@Michael-E-Rose can you rebase / update

[Michael-E-Rose]

No sorry. This is too complicated for me.

On 30 June 2016 at 12:20, Jeff Reback [email protected] wrote:

@Michael-E-Rose https://github.com/Michael-E-Rose can you rebase /
update

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#13167 (comment), or mute
the thread
https://github.com/notifications/unsubscribe/AGxGdtd6g5qhvsv_8lU07_c1gvv-nv2pks5qQ5hcgaJpZM4Id0Lk
.

Michael E. Rose / PhD student

University of Cape Town | African Institute of Financial Markets and Risk
Management
michael-e-ro.se/

[gfyoung]

@Michael-E-Rose : I can take a look since I have waded through those muddy waters that are compat with numpy. If that's alright, you can close this.

[Michael-E-Rose]

It's alright, thanks. I close this now.