STV Vote Counting

Description

Functions in this package are designed to clean, validate, and count single transferable votes.

Author(s)

Jay Emerson (expand this)

See Also

Useful links:

• https://github.com/jayemerson/STV

• Report bugs at https://github.com/jayemerson/STV/issues

ballots data contains rankings of 44 candidates by 489 people. 10 rows are black, which will prompt an error. This data follows the structure of a 2016 Yale faculty senate election, but vote choices have been permuted and candidate names have been removed.

Description

ballots data contains rankings of 44 candidates by 489 people. 10 rows are black, which will prompt an error. This data follows the structure of a 2016 Yale faculty senate election, but vote choices have been permuted and candidate names have been removed.

Author(s)

John W. Emerson john.emerson@yale.edu

Tries to Address validateBallots() Warning(s).

Description

The cleanBallots() function tries to clean data for use in stv(). Some warnings from validateBallots() can only be addressed by the user.

Usage

cleanBallots(x, cand.names = NULL)

Arguments

Details

The cleanBallots() function assumes x contains rows and columns corresponding to ballots and candidates respectively. Formatting corrections are attempted in the following order:

1. If x is a matrix it is convered to a data.frame. Otherwise, user must provide a data.frame.

2. If x has non-numeric entries, user must convert data to numeric type.

3. If cand.names is specified, it is assigned as column names of x. All collumn names must be uniquely specified. x already has valid column names, no need to specify cand.names. If column names of x missing and cand.names not specified, returns error message.

4. Removes blank columns.

5. Removes blank and/or non-sequentially ranked rows.

Value

a data.frame compatible for stv() function.

Examples

data(ballots)
cballots <- cleanBallots(ballots)
validateBallots(cballots)

Implement STV Counting Systems

Description

Analyze a data frame of STV election ballot rankings and return the elected voters and details of the steps in the election counting.

Usage

stv(
  x,
  seats = 1,
  file = "",
  surplusMethod = "Cambridge",
  quotaMethod = "Droop",
  quotaFloor = TRUE
)

Arguments

Details

Revise the details: stv() first validates x by running the validateBallots() function. Once validation is complete, it implements the selected single transferable vote counting method. Each round of counting starts with idetification of active ballots. Then a quota is calculated (currently only supports Droop method floor(votes/(seats + 1)) + 1 and Hare method floor(votes/seats)) with or without the floor() rounding. A tally of each candidate's vote share is obtained using top choices of active ballots, where a ballot stays active until it runs out of marked choices or gets removed during surplus reallocation. If a candidate reaches the quota, she/he is elected and associated surplus ballots are reallocated (currently only supports Cambridge and Fractional methods). If multiple candidates reach the quota, all of them are elected and their surpluses are all reallocated. If no candidate reaches the quota, then the candidate with the minimum number of votes is eliminated. If multiple candidates tie for minimum number of votes, one of them is selected at random and eliminated. The process is repeated until all of the seats are filled or the number of candidates still in race equals the number of unfilled seats. In the later case, all of the active candidates are elected.

Value

The object returned is a list consisting of two components: a vector of the elected candidates, and a data frame with rows containing detailed results from each round of STV counting.

For every round of vote counting, a row of the detailed information contains: number of active ballots, seats to remaining to be filled, the current quota, the maximum and minimum votes obtained by each candidate, who was eliminated (if applicable), whether there was a tie for elimination (indicated by how many tied), who was elected (if applicable), the surplus if elected (or multiple surpluses if multiple candidates were elected), and each candidate's votes tally for that round.

Examples

data(ballots)
cballots <- cleanBallots(ballots)

set.seed(1)
result1 <- stv(cballots, seats = 4)
names(result1)
result1$elected

set.seed(4)
result2 <- stv(cballots, seats = 4)
result2$elected

## Not run: 
result3 <- stv(cballots, seats = 4, surplusMethod = "Fractional")
result3$elected

result4 <- stv(cballots, seats = 4, surplusMethod = "Fractional",
               quotaMethod = "Hare")
result4$elected

result5 <- stv(cballots, seats = 4, surplusMethod = "Cambridge",
               quotaMethod = "Hare")
result5$elected

## End(Not run)

Validates Input Data Format for stv().

Description

The function validateBallots validates data format for stv() input. It assumes that each row corresponds to a ballot and each column corresponds to a candidate. If input data is in the correct format, validateBallots returns string: "All tests passed. Please feel free to run stv() function." Otherwise, it prints a message corresponding to the formatting error.

Usage

validateBallots(x)

Arguments

Details

The validateBallots function should be run before stv() is called. Before any calculation, stv() calls this function and proceeds only if x passes all of the following tests.

1. If x is a data.frame.

2. If x has unique column names.

3. If x has numeric entries.

4. If x doesn't have any blank columns.

5. If x doesn't have any blank rows.

6. If x doesn't have any rows with non-sequential ranks.

Value

string: "All tests passed. Please feel free to run stv() function." if x passes all tests. Otherwise, a message corresponding to the problem is retunred. NOTE: Some of the warnings can be fixed using cleanBallots() function. Other issues must be fixed by the user.

Examples

data(ballots)
result <- try(validateBallots(ballots), silent=TRUE)
print(result)