Docs/guide manual reference

[adolgert]

This splits documentation into a guide, a manual, and a reference.

• The guide now has expanded text with a gentle-ish introduction.
• The manual part is pretty much the same as what was there. Needs work still.
• The reference is organized, and the docstrings now include the standard form for Documenter.

This uses Literate.jl for one of the guide's examples. Works OK.

[slwu89]

@adolgert this looks good to me. I made some changes in d2c89d4. I will describe the major changes I did below:

• in docs/src/distributions.md: changed the wording of how Fleck is introduced to explicitly bring up process vs event-oriented simulations; anyone who knows what JuliaSim is will know those two terms and instantly grasp (hopefully) what Fleck is doing. I also added exaplanation of the CDF/Survival there because otherwise the survival term in the marginal firing probability comes out of nowhere
• docs/src/guide.md: for some reason the mosquito vs bat metaphor took a while for me to grasp; changed to a metaphor very obviously in the exact same domain. Hopefully it helps.
• docs/src/simple_board.jl: moved the key type out of the functions so it's only defined once (it's currently defined twice in the code, a bit brittle).

I did not touch develop.md or vas.md as they look like they are still in progress. Thanks for keeping delay.md around; I'm not sure what to do with it, although I am very proud of that figure I made at the top. Perhaps it will live nicely with #45?

Can I suggest that distributions.md move to using Documenter/Literate to produce the graphs? I want to include plots of the CDF/survival and it would be nice to have it all programatically generated from code.

Other things:

• Can I be added as coauthor/contributor on the package?
• Your contact email is probably not still the uw.edu one
• Do you want to link to your counting processes paper somewhere for the adventurous reader? https://arxiv.org/abs/1610.03939

[adolgert]

Your edits are great, and your suggestions are, too. I'll merge this PR because it's big and then do the suggestions.