Support date-time binning in histograms

[mkcor]

When date-time durations are expressed with a numerical value, plotly's unit is the millisecond (ms).
This is different in R, where the POSIXt class uses seconds, and the Date class uses days.

This PR offers the corresponding conversions for the width (size) of (x) bins in histograms.

See these successful ggplotly conversions when dates are of class POSIXt:
https://plot.ly/~marianne2/336/deaths-by-police-across-time-by-region/,
when dates are of class Date:
https://plot.ly/~marianne2/358/deaths-by-police-across-time-by-region/

/cc @chriddyp @alexcjohnson @jackparmer

[mkcor]

New images can be seen here: 18709da
(Corresponding to b290026)

[alexcjohnson]

whitespace?

[mkcor]

Yes, actually: indentation. This is the convention / automatically added in RStudio.
Both conventions (A: preserving indentation within a block and B: no indentation on empty lines) make sense, I once found a great thread on the topic on StackOverflow... Let me see if I can find it again.

[alexcjohnson]

Aside from the 3 lines with added whitespace, looks great!

[mkcor]

@alexcjohnson This one, I think: http://stackoverflow.com/questions/2727988/python-indentation-in-empty-lines
I like all these points made by all the people... I really like that A is copy-pastable in an interpreter.
And most R users are RStudio users.

[alexcjohnson]

If that's a standard R convention I guess we can keep it in this repo. In the rest of our code though we have always deleted trailing whitespace. I see the argument about python interpreters, but I think this has too much weight behind it and it keeps diffs much cleaner to insist on no trailing whitespace.

[mkcor]

Yes, I know. You probably even have a git alias for forcing trailing whitespace removal upon every commit, don't you? ;)

Anyway, it's a cultural thing and I wasn't trying to contaminate the rest of our code base. ;)

[alexcjohnson]

@mkcor Even farther back than git - I have my editor set to strip trailing whitespace on every save. So I personally would have to override my editor settings (or just use RStudio) in order to work in this repo, or use something like http://stackoverflow.com/questions/3515597/git-add-only-non-whitespace-changes (wow, people really get riled up about this issue!).

Anyway, the only reason I care about this is having to look at diffs with whitespace-only changes. As long as all contributors here agree to the same convention (looks like the blank lines were added by @chriddyp ) it's fine.