public NaiveBayesClassifierOld(TableFixedData data)
{
_data = data;
var doubleConverter = new DoubleConverter();
_distribution = new IDistribution[data.ClassesValue.Length, data.Attributes.Length];
for (int index = 0; index < data.Attributes.Length; index++)
{
if (data.Attributes[index] == TableData.ClassAttributeName)
{
var column = data.GetColumn <int>(index);
_classesProbablityDistribution = new CategoricalDistribution(column, data.ClassesValue.Length);
}
else
{
var isColumnNumeric = data[0, index].IsNumeric();
if (isColumnNumeric)
{
var values = GetDataPerClass <double>(data, index, doubleConverter);
for (int classIndex = 0; classIndex < data.ClassesValue.Length; classIndex++)
{
_distribution[classIndex, index] = new GaussianDistribution(values[classIndex]);
}
}
else
{
var values = GetDataPerClass <string>(data, index);
for (int classIndex = 0; classIndex < data.ClassesValue.Length; classIndex++)
{
var categoryData = values[classIndex].Select(item => data.GetSymbol(item, index)).ToArray();
_distribution[classIndex, index] = new CategoricalDistribution(categoryData, categoryData.Length);
}
}
}
}
}
public static DataSample[] ToSample(TableFixedData tableFixedData)
{
var samples = new DataSample[tableFixedData.Count];
for (int rowIndex = 0; rowIndex < tableFixedData.Count; rowIndex++)
{
var currentSample = new DataSample
{
DataPoints = new DataPoint[tableFixedData.Attributes.Length - 1],
ClassId = tableFixedData.Class(rowIndex)
};
int dataPointIndex = 0;
for (int columnIndex = 0; columnIndex < tableFixedData.Attributes.Length; columnIndex++)
{
if (tableFixedData.Attributes[columnIndex] != TableData.ClassAttributeName)
{
var value = tableFixedData[rowIndex, columnIndex];
var dataPoint = new DataPoint
{
ColumnId = columnIndex,
Value =
tableFixedData.IsDiscreteColumn(columnIndex)
? Convert.ToDouble(tableFixedData.GetSymbol(value.ToString(), columnIndex))
: Convert.ToDouble(value)
};
currentSample.DataPoints[dataPointIndex] = dataPoint;
dataPointIndex++;
}
}
samples[rowIndex] = currentSample;
}
return(samples);
}
public string Compute(IDataRow datarow)
{
var probabilities = new double[_data.ClassesValue.Length];
var attributes = datarow.Attributes.ToArray();
var doubleConverter = new DoubleConverter();
var maxProbabilityIndex = 0;
for (int index = 0; index < probabilities.Length; index++)
{
probabilities[index] = 1;
for (int columnIndex = 0; columnIndex < _data.Attributes.Length; columnIndex++)
{
if (attributes[columnIndex] == TableData.ClassAttributeName)
{
probabilities[index] = _classesProbablityDistribution.GetLogProbability(index);
continue;
}
var value = datarow[attributes[columnIndex]];
if (!value.IsNumeric())
{
probabilities[index] = probabilities[index] + _distribution[index, columnIndex].GetLogProbability(_data.GetSymbol((string)value, columnIndex));
}
else
{
probabilities[index] = probabilities[index] + _distribution[index, columnIndex].GetLogProbability(doubleConverter.Convert(value));
}
}
if (probabilities[maxProbabilityIndex] < probabilities[index])
{
maxProbabilityIndex = index;
}
}
return(_data.ClassesValue[maxProbabilityIndex]);
}
