Skip to content
Standalone tool for estimating Bayesian posterior loss from election tally for ballot-polling audit.
HTML Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.gitignore
LICENSE
README.md
bptool.py
bptool_api_docs.html
bptool_api_docs.pdf
test.csv

README.md

2018-bptool

Standalone tool for estimating Bayesian posterior loss from election tally for ballot-polling audit.

Usage

bptool.py can be run with python3 in two ways: one for single-county ballot-polling audits, and one for multi-county ballot-poling audits.

Single County

If the election only has one county enter tallies as numbers e.g.

python3 bptool.py 20000 40 60 10

where the first number is the total number of votes cast, and the following numbers are the number of votes seen for each candidate in the auditing done so far.

Multiple Counties

If the election spans multiple counties, it might be easier to pass in the sample tallies as a csv file

python3 bptool.py --path_to_csv test.csv

Additional inforomation on usage and optional arguments can be view via:

python3 bptool.py -h

Output

The output for a single-county ballot-polling audit might look like (for the above command python bptool.py 200000 40 60 10):

```Candidate name           	 Estimated probability of winning a full recount
``` 2                        	   97.86 %  
``` 1                        	    2.14 %  
``` 3                        	    0.00 %  

## Importing bptool.py
The module ``bptool.py`` may also be imported into other python code,.  See the
code for guidance.
You can’t perform that action at this time.