Fixing 'names' doc for 'read_csv'

[bessarabov]

• closes #xxxx
• tests added / passed
• passes black pandas
• passes git diff upstream/master -u -- "*.py" | flake8 --diff
• whatsnew entry

I think there is a bug in the docs. This Jupyter notebook shows that the code does not work as expected from the docs https://gist.github.com/bessarabov/5e42bbb359fb86aaa87aed6c12a95bc9

So I've fixed the doc to match behaviour.

[alimcmaster1]

‘If the file contains a header row, then you should explicitly pass ‘’header=0’’ to override the column names.’ - I think that would make more sense?

[bessarabov]

Thank you. I've rechecked who the system is working (I've added the comment with the details) and I have updated the text as you have suggested.

[bessarabov]

I made one more python notebook to check how the code works when there is no header in csv:

• pandas read_csv no-header.ipynb
• pandas read_csv with-header.ipynb

The current doc text is:

names : array-like, optional
List of column names to use. If file contains no header row, then you should explicitly pass
header=None. Duplicates in this list are not allowed.

As you can see from pandas read_csv no-header.ipynb

• [pandas read_csv with-header.ipynb] there is no need to explicitly pass header=None the code works without header=None as it does with header=None

So I think there should be no such text in the doc.

But when there is a header in csv file you must specify header=0 to make the code work as expected — pandas read_csv with-header.ipynb

[alimcmaster1]

Sorry minor typo from me should be ‘file contains’

[bessarabov]

Oops. Fixed, thank you.

[alimcmaster1]

LGTM

[jreback]

thanks @bessarabov