make diagonalNetwork and radialNetwork responsive (sort-of)

[timelyportfolio]

This pull is a little more intrusive on the current behavior of diagonalNetwork and radialNetwork, but I think it wisely lets the viewBox attribute of SVG handle sizing and resizing. It seems the resize method already is a little broken, so I don't think this causes any more damage. Using this approach also insures that the diagram fits within the htmlwidget container div, so a R user does not need to fiddle with margins unless desired.

At a minimum, I think this should exist as an argument to toggle on or off, but I'm very comfortable making this the default behavior especially given @christophergandrud focus on simplicity.

[christophergandrud]

I really like it and don't think there is any need in principle to dedicate an argument to this.

One issue I just spotted and haven't had a chance to look into: If you knit the gh-pages .Rmd file you will see that it produces strange and undesirable behaviour.

[timelyportfolio]

Thanks so much for the feedback and bug discovery. I'll work out the kinks and hopefully have a working pull momentarily.

[timelyportfolio]

I believe I have worked out the kinks in 1a9d8f5. @christophergandrud, do you mind verifying?

[christophergandrud]

Nice, merging into master now.

[timelyportfolio]

oh great, what else for CRAN? anything you would like to implement? I'll keep going through the existing pulls.

[christophergandrud]

Generally, I'm a believer in relatively incremental CRAN updates to mimimise the risk of unanticipated code breaks at any given update. So I'm fine.

But if you think there is an important pull to merge, then lets do that.

[timelyportfolio]

#85 reminded me that the gh-pages site will demo the newest and greatest, which argues for a quicker CRAN submit. I am ready when you are.

[christophergandrud]

I'm also ready if you are. I'll make compile it now and send it up to CRAN.

[christophergandrud]

. . . submitted v0.2.4