leo-senate-model

Leo combines polls with other information to predict how many Senate races Democrats and Republicans will win in 2014.

Leo uses:

Historical results from the Open Elections project, CQ Press Voting and Elections Collection, Dave Leip’s Atlas of U.S. Presidential Elections, and the Federal Election Commission.

Fund-raising data from the Federal Election Commission. Biographical data from Project Vote Smart and candidate websites.

Polls aggregated by Pollster, the Roper Center for Public Opinion Research, the U.S. Officials’ Job Approval Ratings project, Polling Report, Gallup, Talking Points Memo, The Argo Journal, Real Clear Politics and FiveThirtyEight.

Brief instructions

Please make sure the following R packages have been installed: gam, gtools, lubridate, maps, RJSONIO, gdata, plotrix, zoo
Change directory to the top-level working directory of this Git repository.
Run: Rscript master-public.R
Prediction output can then be found in the data-publisher/public/_big_assets/ subdirectory.

Sample output data in `/data-publisher/public/_big_assets`

histogram.tsv

dem	chance
44	0
45	0
46	0
47	0.05
48	0.25
49	0.15
50	0.3
51	0.15

matchups.tsv

statePostal	probability	mean	scale	office	dem	rep
AK	0.0675992536643285	13.3371042222312	6.37706218633251	ak-senate-class-ii-2014	mark-begich	joe-miller
AK	0.672480368777626	-0.369330147923164	6.40862522137136	ak-senate-class-ii-2014	mark-begich	daniel-s-sullivan
AK	0.259920377558046	-1.42864965262853	6.43282051331378	ak-senate-class-ii-2014	mark-begich	mead-treadwell
AL	1	-92.2586242817344	13.0668255800746	al-senate-class-ii-2014	NA	jeff-sessions
AR	1	-0.588349828909989	5.43497477566511	ar-senate-class-ii-2014	mark-pryor	thomas-cotton-1
CO	1	2.75130137738909	5.32999951709124	co-senate-class-ii-2014	mark-udall	cory-gardner
DE	0.0485586168626867	22.1226995091985	12.2918575405219	de-senate-class-ii-2014	christopher-a-coons	tom-kovach
DE	0.951441383137313	24.462299974418	12.2976359697371	de-senate-class-ii-2014	christopher-a-coons	christine-o'donnell

nyt-ratings-today.json

{
  "date": "2014-04-22T13:19:49-0400",
  "data": [
    {
      "dem-pct": 49.428,
      "state": "AK",
      "office": "ak-senate-class-ii-2014",
      "fips": "2",
      "nyt-rating": 4
    },
    {
      "dem-pct": 4.1438E-10,
      "state": "AL",
      "office": "al-senate-class-ii-2014",
      "fips": "1",
      "nyt-rating": 7
    },
    {
      "dem-pct": 45.693,
      "state": "AR",
      "office": "ar-senate-class-ii-2014",
      "fips": "5",
      "nyt-rating": 4
    }
  ]
}

parameters.json

{
  "continuingSeats": {
    "dem": 34,
    "rep": 30
  },
  "nationalErrorScale": 0.40797,
  "upsetProb": 40,
  "houseEffect": 3.5,
  "mt": 85,
  "mtNum": "17",
  "mtDenom": "20",
  "aan": "an"
}

senate-likelihood-all.tsv

office	date	dem
ak-senate-class-ii-2014	1-Jan-14
ak-senate-class-ii-2014	25-Feb-14	51.4921956208518
ak-senate-class-ii-2014	22-Apr-14	49.4276015976854
al-senate-class-ii-2014	1-Jan-14
al-senate-class-ii-2014	25-Feb-14	9.22057393777991e-10
al-senate-class-ii-2014	22-Apr-14	4.14383203099536e-10

senate-likelihood.tsv

date	dem
1-Jan-14	50
25-Feb-14	35
22-Apr-14	55

slug-lookup.tsv

slug	name
mark-begich	Mark	Begich
william-bryk	William	Bryk
john-jaramillo	John	Jaramillo
joe-miller	Joe	Miller
sarah-palin	Sarah	Palin
daniel-s-sullivan	Dan	Sullivan
kathleen-tonn	Kathleen	Tonn
mead-treadwell	Mead	Treadwell

jinhyeokjang/leo-senate-model