Working with the Clinton State Dept Email Dumps in R (Part 1: Graphs)

2016-02-20 – 12:06
Posted in R
Tagged post

I put this together after experimenting with `ggplot2` and `ggnetwork` earlier this week. The changes I made added `svgPanZoom` into the mix. Consequently, it has a widget in it, so it was just easier to embed the full R markdown HTML into an `iframe` than to try to extract the content piecemeal into WP.

You can bust the `iframe` via .

Read on to see how to grab some JSON, create edge list, do some basic graph stats with `igraph` and generate an interactive visualization with `ggplot2` and `svgPanZoom`.

4 Comments

- Steven L. Johnson (@StevenLJohnson)
- Posted 2016-02-21 at 07:14
- Permalink
- Reply
How are cc’s and bcc’s handled? Could that be the reason for the results you show for:

filter(from != “Hillary Clinton” & to != “Hillary Clinton”)
- - hrbrmstr
  - Posted 2016-02-21 at 08:01
  - Permalink
  - Reply
  Good q. Until I get the PDFs (which should have this info) I won’t know. The WSJ JSON is woefully scant.
- ryan
- Posted 2016-02-25 at 16:10
- Permalink
- Reply
I believe the orphaned mini-graphs are indeed cases where Hillary is CC’d. See document C05791081 as an example.
- - hrbrmstr
  - Posted 2016-02-25 at 18:38
  - Permalink
  - Reply
  Aye. I’m leaning towards CC or BCC for them now. I have the PDFs and converted them to (cleaned) ASCII. Next post will be working with those and prbly a bit more graphs.

One Trackback/Pingback

By Working with the Clinton State Dept Email Dumps in R (Part 1: Graphs) – Mubashir Qasim on 20 Feb 2016 at 10:21 pm

[…] article was first published on R – rud.is, and kindly contributed to […]

rud.is