DOC: isin Example for .13 release notes

[TomAugspurger]

I also fixed a few typos in a separate commit. I can remove that commit if they're going to cause conflicts.

Also did I get the syntax correct for this link?

See :ref:Boolean Indexing<indexing.boolean> for more

[jreback]

this looks right, but I would add a reference, say indexing.isin (and maybe it a separate sub-section, where you put the series/frame cases). You can build the docs locally to see how it looks

[TomAugspurger]

Okay, I moved the Series and DataFrame isin sections of indexing.rst into their own subsection under Basics.

So pending a decision on how to handle #4421, which I don't have time to get to today, I think this is ready to go.

Off topic: @jreback were you aware that the to_datetime examples in What's New are failing? http://pandas.pydata.org/pandas-docs/dev/whatsnew.html#enhancements

I wanted to check before making an issue.

[jreback]

@TomAugspurger you mean the to_timedelta ones.....do you see that big warning about requiring numpy 1.7, unfortunately the build machine (which @changhiskhan / @wesm) run runs 1.6.2 :(.

timedeltas are so broken in 1.6 that only very limited ops can be done

[jreback]

can you also add a simple Series example? (series indexed by the same series)

[TomAugspurger]

I mean to switch the line you commented on to a Series.

You mean like

ser = Series(['one', 'one', 'two', 'three', 'two', 'one', 'six'])
ser.isin(['one', 'two'])
seri[ser.isin(['one', 'two'])]

?

[jreback]

no your example was fine, maybe something that uses integers or something? just a pure Series (before you do the DataFrame example)

In [2]: s = Series(np.arange(10),index=np.arange(10)[::-1],dtype='int64')

In [3]: s
Out[3]: 
9    0
8    1
7    2
6    3
5    4
4    5
3    6
2    7
1    8
0    9
dtype: int64

In [5]: s[s.isin([2,4])]
Out[5]: 
7    2
5    4
dtype: int64

[TomAugspurger]

Added the Series example.

[jreback]

perfect..thanks

[jreback]

@TomAugspurger can your rebase this and can get in

[TomAugspurger]

@jreback rebased… but the typo fixes I made in my second commit got tangled up with some changes someone (probably you) made to the HDF5 notes. Lines 35-109 my PR. Should I go back to before my rebase and do it again with just the changes in my first commit? There's a way to select which commits get applied right?

[jreback]

git rebase -i upstream/master should work

[TomAugspurger]

Ok, should be good to go.

[jreback]

gre8 thanks

[jreback]

@TomAugspurger @hayd I slightly changed the isin example in v0.13.0 (because it got moved and df was not defined correctly). Pls verify it and change if needed.

[TomAugspurger]

Looks good.