DOC: Flesh out the R comparison section of docs (GH3980)

[HeardACat]

Some additions to #3980.

I've done some more comparisons to the plyr functions located here. More specifically aaply, alply don't really have much to do with pandas (its just list comprehension). and dlply is probably more suited to be in statsmodel example rather than pandas.

[ghost]

Thanks @chappers , this is useful stuff. But the PR violates a bunch of git-related guidelines
explained in CONTRIBUTING.md:

• You merged master into your PR branch.
• You developed the PR on top of a pandas commit from 6 months ago, that's about 2500 commits
  back in the fog of time.

Both of these do nasy things to the git history and so the devs follow a bunch of rules on how to use git to avoid that.

Please fix it yourself or let us know if you need git help. It's fine if you do, we all did at some point.
We value your work, we want your help making pandas better, but please follow the guidelines or ask
for help if you need it.

[jtratner]

@chappers - easiest is just to copy paste your changes to a separate file, checkout current master, and then paste them back on. Since it's docs and a relatively new section of the docs, you should be fine.

[HeardACat]

Yeah...I don't have any idea what I'm doing. So I'm suppose to update a branch to upstream/master and then run git rebase on it? And then submit a PR?

[ghost]

No, that would have prevented this, now you need to fix the existing branch.

upddate your master branch to point at current master on github, then
checkout your PR branch. then:

[chappers-r-compare] % git reset master    # reset head, your "starting point", without changing the working copy
Unstaged changes after reset:
M   doc/source/comparison_with_r.rst
[chappers-r-compare] % ga doc/source/comparison_with_r.rst
[chappers-r-compare] % gc -m "DOC: your description here"
# and finally force push to clobber the current PR branch on github with the repaired one
git push -f

[HeardACat]

Thanks a lot.

[ghost]

nicely done.

[jorisvandenbossche]

Thanks for this work!
I add some comments inline.

[jorisvandenbossche]

Can you import numpy as np? (and maybe also pandas as pd)

[jreback]

since this is only docs, as soon as @jorisvandenbossche , @y-p ok...then can merge it in

[ghost]

I haven't run sphinx, but assuming it's ok... merge away.

[jorisvandenbossche]

I ran Sphinx with this, and its OK. I cannot judge everything in the PR (the R code), but it seems OK.

[jreback]

ok will merge and can check how

[jreback]

@chappers

docs look good

http://pandas.pydata.org/pandas-docs/dev/comparison_with_r.html

feel free to add more :)