fl2000
OpenML dataset with id 477
No author found.
Full work available at URL: https://api.openml.org/data/v1/download/52589/fl2000.arff
Upload date: 28 September 2014
Dataset Characteristics
Number of classes: 5
Number of features: 16 (numeric: 14, symbolic: 2 and in total binary: 1 )
Number of instances: 67
Number of instances with missing values: 0
Number of missing values: 0
Author: Source: Unknown - Date unknown Please cite:
County data from the 2000 Presidential Election in Florida.
Compiled by Brett Presnell Department of Statistics, University of Florida
These data are derived from three sources, described below. As far as I am aware, you are free to use these data in any way that you see fit, though some acknowledgement is always nice.
The candidate vote counts are the final certified counts reported by the Florida Division of Elections. These were obtained from the NORC web site in the file Cert_results.csv. Note that these do NOT inculde the federal absentee votes (so that Gore's total vote is actually higher here than Bush's).
The undervote and overvote counts were extracted from the NORC ballot level data in the file aligned.txt. Since aligned.txt is too large to work with in R (or almost any other program) I used cut (a standard UNIX program) to extract just the columns I needed:
cut -f 2,9,10 -d"|" aligned.txt > tmp
Then I read the results into R and processed them there.
The technology and columns data were extracted from the Media Group data from the NORC web site. "Technology" is simply the type of voting machine used, and "columns" is 1 if the ballot listed the presidential candidates in a single column on a single page, and 2 if the presidential candidates were spread over two columns or two pages of the ballot.
These agree with some earlier data that I had obtained from the NY Times web site, except that in the media group data the PalmBeach county ballot (the famous butterfly ballot) was listed as having one column. I would definitely call this a two-column ballot, so that is the designation recorded here. At one time I thought that MiamiDade County also used a two-column ballot, but I was wrong (the ballot listed the candidates and parties in English and Spanish in opposing columns). Images of most of the ballots can be found on the New York Times web site: www.nytimes.com/images/2001/11/12/politics/recount/index_BALLOT.html
Information about the dataset
CLASSTYPE: nominal
CLASSINDEX: 2
