public static ConditionedDistribution SeparateGenderCalculation(Row[] allData, List <List <int> > matches)
{
ConditionedDistribution toReturn = new ConditionedDistribution();
var eidToRow = new Dictionary <int, Row>();
foreach (var row in allData)
{
eidToRow[row.EnterpriseID] = row;
}
long[] givenMatch = new long[3];
foreach (var set in matches)
{
for (int i = 0; i < set.Count; i++)
{
for (int j = i + 1; j < set.Count; j++)
{
var ri = eidToRow[set[i]];
var rj = eidToRow[set[j]];
if (ri.GENDER == "" || rj.GENDER == "")
{
givenMatch[0]++;
}
else if (ri.GENDER == rj.GENDER)
{
givenMatch[1]++;
}
else
{
givenMatch[2]++;
}
}
}
}
long blank = allData.Count(r => r.GENDER == "");
long male = allData.Count(r => r.GENDER == "M");
long female = allData.Count(r => r.GENDER == "F");
long[] givenNoMatch = new long[3];
givenNoMatch[0] = blank * (blank - 1) / 2 + blank * (male + female) - givenMatch[0];
givenNoMatch[1] = male * (male - 1) / 2 + female * (female - 1) / 2 - givenMatch[1];
givenNoMatch[2] = male * female - givenMatch[2];
toReturn.GivenMatch = CountsToDistribution(givenMatch);
toReturn.GivenNoMatch = CountsToDistribution(givenNoMatch);
return(toReturn);
}
public static ConditionedDistribution ComputeConditionedDistribution(List <List <int> > knownMatches, RowMatchObject fieldMatches, int totalRowCount)
{
int largestDistance = fieldMatches.Matches.MatchArray.Max(l => l.Max(e => e.Distance));
var distanceCounts = DistanceCounts(fieldMatches, totalRowCount, largestDistance);
var distanceCountsGivenMatched = DistanceCountsGivenMatched(knownMatches, fieldMatches, largestDistance);
long[] distanceCountsGivenUnmatched = new long[largestDistance + 2];
for (int i = 0; i < largestDistance + 2; i++)
{
distanceCountsGivenUnmatched[i] = distanceCounts[i] - distanceCountsGivenMatched[i];
}
ConditionedDistribution toReturn = new ConditionedDistribution();
toReturn.GivenMatch = CountsToDistribution(distanceCountsGivenMatched);
toReturn.GivenNoMatch = CountsToDistribution(distanceCountsGivenUnmatched);
return(toReturn);
}