Perlkonig.Condorcet 0.2.0

.NET implementation of various Condorcet voting algorithms

Install-Package Perlkonig.Condorcet -Version 0.2.0
dotnet add package Perlkonig.Condorcet --version 0.2.0
<PackageReference Include="Perlkonig.Condorcet" Version="0.2.0" />
For projects that support PackageReference, copy this XML node into the project file to reference the package.
paket add Perlkonig.Condorcet --version 0.2.0
The NuGet Team does not provide support for this client. Please contact its maintainers for support.

Condorcet Library

A C# library that implements various Condorcet voting algorithms.

Hosted by GitHub

Ballots

All the algorithms use the same ballot: an dictionary whose key is the candidate name (of type T, which must implement IComparable) and whose value is the rank (which must be an unsigned integer). Typically, the absolute values of the ranks is irrelevant. All that matters is their sequence. For example, you could rank your first choice as 100 as long as your next choice was something like 110 and so on down the line. But this might differ by algorithm. Traditionally you indicate your first choice with a 1 and go from there. You do not need to vote for all candidates. Skipping a candidate simply means they're at the very bottom of your list.

Here's an example of a ballot ranking five candidates in the order ACBED:

Dictionary<char, uint> b1 = new Dictionary<char, uint>
{
    {'A', 1},
    {'C', 2},
    {'B', 3},
    {'E', 4},
    {'D', 5}
};

Available Algorithms

For all algorithms, the constructor takes the list of candidates, after which you can use the AddBallot method. You can then call the Rank method to get a list of candidates in order, winner first.

Schulze

A description of the method can be found on Wikipedia.

using System.Collections.Generic;
using Condorcet;

HashSet<char> candidates = new HashSet<char> {'A', 'B', 'C', 'D', 'E'};

//Build the ballot
Dictionary<char, uint> ballot = new Dictionary<char, uint>
{
    {'A', 1},
    {'C', 2},
    {'B', 3},
    {'E', 4},
    {'D', 5}
};

//Construct the Schulze object with the list of candidates
Schulze<char> s = new Schulze<char>(candidates);

//Assume 50 people all voted the same way
s.AddBallot(ballot, 50);

//Get the rankings
char[] ranked = s.Rank();

//Because everybody voted unanimously, you should get the array ['A', 'C', 'B', 'E', 'D'].
//See the tests for more examples.

Ranked Pairs

A description of the method can be found on Wikipedia.

using System.Collections.Generic;
using Condorcet;

HashSet<string> candidates = new HashSet<string> {"Memphis", "Nashville", "Chattanooga", "Knoxville"};

Dictionary<string, uint> b1 = new Dictionary<string, uint>
{
    {"Chattanooga", 1},
    {"Knoxville", 2},
    {"Nashville", 3},
    {"Memphis", 4}
};

RankedPair<string> s = new RankedPair<string>(candidates);
s.AddBallot(b1, 42);
string[] expected = new string[] {"Chattanooga", "Knoxville", "Nashville", "Memphis"};
Assert.True(expected.SequenceEqual(s.Rank()));

//Because everybody voted unanimously, you should get the array ["Chattanooga", "Knoxville", "Nashville", "Memphis"].
//See the tests for more examples.

TODO

  • I'm a novice programmer. Any suggestions to improve efficiency or readability would be warmly welcomed.
  • I'm always looking for more test cases. I need to add code for ties.

Condorcet Library

A C# library that implements various Condorcet voting algorithms.

Hosted by GitHub

Ballots

All the algorithms use the same ballot: an dictionary whose key is the candidate name (of type T, which must implement IComparable) and whose value is the rank (which must be an unsigned integer). Typically, the absolute values of the ranks is irrelevant. All that matters is their sequence. For example, you could rank your first choice as 100 as long as your next choice was something like 110 and so on down the line. But this might differ by algorithm. Traditionally you indicate your first choice with a 1 and go from there. You do not need to vote for all candidates. Skipping a candidate simply means they're at the very bottom of your list.

Here's an example of a ballot ranking five candidates in the order ACBED:

Dictionary<char, uint> b1 = new Dictionary<char, uint>
{
    {'A', 1},
    {'C', 2},
    {'B', 3},
    {'E', 4},
    {'D', 5}
};

Available Algorithms

For all algorithms, the constructor takes the list of candidates, after which you can use the AddBallot method. You can then call the Rank method to get a list of candidates in order, winner first.

Schulze

A description of the method can be found on Wikipedia.

using System.Collections.Generic;
using Condorcet;

HashSet<char> candidates = new HashSet<char> {'A', 'B', 'C', 'D', 'E'};

//Build the ballot
Dictionary<char, uint> ballot = new Dictionary<char, uint>
{
    {'A', 1},
    {'C', 2},
    {'B', 3},
    {'E', 4},
    {'D', 5}
};

//Construct the Schulze object with the list of candidates
Schulze<char> s = new Schulze<char>(candidates);

//Assume 50 people all voted the same way
s.AddBallot(ballot, 50);

//Get the rankings
char[] ranked = s.Rank();

//Because everybody voted unanimously, you should get the array ['A', 'C', 'B', 'E', 'D'].
//See the tests for more examples.

Ranked Pairs

A description of the method can be found on Wikipedia.

using System.Collections.Generic;
using Condorcet;

HashSet<string> candidates = new HashSet<string> {"Memphis", "Nashville", "Chattanooga", "Knoxville"};

Dictionary<string, uint> b1 = new Dictionary<string, uint>
{
    {"Chattanooga", 1},
    {"Knoxville", 2},
    {"Nashville", 3},
    {"Memphis", 4}
};

RankedPair<string> s = new RankedPair<string>(candidates);
s.AddBallot(b1, 42);
string[] expected = new string[] {"Chattanooga", "Knoxville", "Nashville", "Memphis"};
Assert.True(expected.SequenceEqual(s.Rank()));

//Because everybody voted unanimously, you should get the array ["Chattanooga", "Knoxville", "Nashville", "Memphis"].
//See the tests for more examples.

TODO

  • I'm a novice programmer. Any suggestions to improve efficiency or readability would be warmly welcomed.
  • I'm always looking for more test cases. I need to add code for ties.

Release Notes

Added RankedPair algorithm

Dependencies

This package has no dependencies.

This package is not used by any popular GitHub repositories.

Version History

Version Downloads Last updated
0.2.0 178 12/14/2018
0.1.0 137 12/13/2018