// GET: Classifier
public ActionResult ClassifyPre(string className, int classID)
{
var classModel = new ClassificationChoicePre {
choiceName = className, choiceID = classID
};
return(View(classModel));
}//End of ClassifyPre
public ActionResult ResultsPre(ClassificationChoicePre model)
{
ML_DBEntities dbML = new ML_DBEntities();
//Calculate Prior Probabilities
var priorProbabilityM = Convert.ToDouble(
(from c in dbML.pre96
where c.Metric == "Percent Malignant"
select new { c.Value }).Single().Value);
var priorProbabilityB = Convert.ToDouble(
(from c in dbML.pre96
where c.Metric == "Percent Benign"
select new { c.Value }).Single().Value);
// Pull Benign Values From the Database
string[,] allkeys = new string[, ] {
{ "Avg Clump Thickness", "stdev Clump Thickness" },
{ "Avg Uniformity Shape", "stdev Uniformity Shape" },
{ "Avg Uniformity Size", "stdev Uniformity Size" },
{ "Avg Bland Chromatin", "stdev Bland Chromatin" },
{ "Avg Marginal Adhesion", "stdev Marginal Adhesion" },
{ "Avg Normal Nucleoli", "stdev Normal Nucleoli" },
{ "Avg Single Epithelial Cell Size", "stdev Single Epithelial Cell Size" },
{ "Avg Bare Nuclei", "stdev Bare Nuclei" },
{ "Avg Mitosis", "stdev Mitosis" }
};
double[][] numbers = new double[allkeys.Length][];
for (int i = 0; i < 9; i++)
{
double[] statsValues = new double[2];
string key1 = allkeys[i, 0].ToString();
string key2 = allkeys[i, 1].ToString();
statsValues[0] = Convert.ToDouble(
(from c in dbML.pre96
where c.Metric == key1
select new { c.Value }).Single().Value);
statsValues[1] = Convert.ToDouble(
(from c in dbML.pre96
where c.Metric == key2
select new { c.Value }).Single().Value);
numbers[i] = statsValues;
}
// Pull Malignant Values From the Database
string[,] allkeysM = new string[, ] {
{ "Avg Clump Thickness M", "stdev Clump Thickness M" },
{ "Avg Uniformity Shape M", "stdev Uniformity Shape M" },
{ "Avg Uniformity Size M", "stdev Uniformity Size M" },
{ "Avg Bland Chromatin M", "stdev Bland Chromatin M" },
{ "Avg Marginal Adhesion M", "stdev Marginal Adhesion M" },
{ "Avg Normal Nucleoli M", "stdev Normal Nucleoli M" },
{ "Avg Single Epithelial Cell Size M", "stdev Single Epithelial Cell Size M" },
{ "Avg Bare Nuclei M", "stdev Bare Nuclei M" },
{ "Avg Mitosis M", "stdev Mitosis M" }
};
double[][] numbersM = new double[allkeysM.Length][];
for (int i = 0; i < 9; i++)
{
double[] statsValues = new double[2];
string key1 = allkeysM[i, 0].ToString();
string key2 = allkeysM[i, 1].ToString();
statsValues[0] = Convert.ToDouble(
(from c in dbML.pre96
where c.Metric == key1
select new { c.Value }).Single().Value);
statsValues[1] = Convert.ToDouble(
(from c in dbML.pre96
where c.Metric == key2
select new { c.Value }).Single().Value);
numbersM[i] = statsValues;
}
//Calculate Likelihoods
var diagnosisLikelihoodB =
calculateProbability(model.clumpThickness, numbers[0]) *
calculateProbability(model.uniformityShape, numbers[1]) *
calculateProbability(model.uniformitySize, numbers[2]) *
calculateProbability(model.blandChromatin, numbers[3]) *
calculateProbability(model.marginalAdhesion, numbers[4]) *
calculateProbability(model.normalNucleoli, numbers[5]) *
calculateProbability(model.singleEpithelial, numbers[6]) *
calculateProbability(model.bareNuclei, numbers[7]) *
calculateProbability(model.mitosis, numbers[8]);
var diagnosisLikelihoodM =
calculateProbability(model.clumpThickness, numbersM[0]) *
calculateProbability(model.uniformityShape, numbersM[1]) *
calculateProbability(model.uniformitySize, numbersM[2]) *
calculateProbability(model.blandChromatin, numbersM[3]) *
calculateProbability(model.marginalAdhesion, numbersM[4]) *
calculateProbability(model.normalNucleoli, numbersM[5]) *
calculateProbability(model.singleEpithelial, numbersM[6]) *
calculateProbability(model.bareNuclei, numbersM[7]) *
calculateProbability(model.mitosis, numbersM[8]);
//Compare p(X|B)p(B) to p(C|M)p(M)
if ((diagnosisLikelihoodB * priorProbabilityB) >= (diagnosisLikelihoodM * priorProbabilityM))
{
ViewBag.diagnosis = "Diagnosis: Benign";
ViewBag.probabilityB = "The Likelihood Maximum of a Benign Tumor: " + (diagnosisLikelihoodB * priorProbabilityB).ToString();
ViewBag.probabilityM = "The Likelihood Maximum of a Malignant Tumor: " + (diagnosisLikelihoodM * priorProbabilityM).ToString();
Highcharts chart = new Highcharts("chart")
.InitChart(new Chart {
DefaultSeriesType = ChartTypes.Bar
})
.SetTitle(new Title {
Text = "Patient Data Comparison"
})
.SetSubtitle(new Subtitle {
Text = "Data for Patient " + model.patientID
})
.SetXAxis(new XAxis
{
Categories = new[] { "Clump Thickness", "Uniformity Size", "Uniformity Shape", "Marginal Adhesion", "Single Epithelial Cell Size",
"Bare Nuclei",
"Bland Chromatin",
"Normal Nucleoli",
"Mitosis" },
Title = new XAxisTitle {
Text = "Cell Attributes"
}
})
.SetYAxis(new YAxis
{
Min = 0,
Title = new YAxisTitle
{
Text = "Value 1-10",
Align = AxisTitleAligns.High
}
})
.SetTooltip(new Tooltip {
Formatter = "function() { return ''+ this.series.name +': '+ this.y; }"
})
.SetPlotOptions(new PlotOptions
{
Bar = new PlotOptionsBar
{
DataLabels = new PlotOptionsBarDataLabels {
Enabled = false
}
}
})
.SetCredits(new Credits {
Enabled = false
})
.SetSeries(new[]
{
new Series {
Name = "Average Benign Patient", Data = new Data(new object[] { numbers[0][0], numbers[1][0], numbers[2][0], numbers[3][0], numbers[4][0], numbers[5][0],
numbers[6][0], numbers[7][0], numbers[8][0] })
},
new Series {
Name = "Average Malignant Patient", Data = new Data(new object[] { numbersM[0][0], numbersM[1][0], numbersM[2][0], numbersM[3][0], numbersM[4][0], numbersM[5][0],
numbersM[6][0], numbersM[7][0], numbersM[8][0] })
},
new Series {
Name = "Patient " + model.patientID.ToString(), Data = new Data(new object[] { model.clumpThickness, model.uniformitySize, model.uniformityShape, model.marginalAdhesion,
model.singleEpithelial, model.bareNuclei, model.blandChromatin, model.normalNucleoli, model.mitosis })
}
});
return(View(chart));
}
else
{
ViewBag.diagnosis = "Diagnosis: Malignant";
ViewBag.probabilityB = "The Likelihood Maximum of a Benign Tumor: " + (diagnosisLikelihoodB * priorProbabilityB).ToString();
ViewBag.probabilityM = "The Likelihood Maximum of a Malignant Tumor: " + (diagnosisLikelihoodM * priorProbabilityM).ToString();
Highcharts chart = new Highcharts("chart")
.InitChart(new Chart {
DefaultSeriesType = ChartTypes.Bar
})
.SetTitle(new Title {
Text = "Patient Data Comparison"
})
.SetSubtitle(new Subtitle {
Text = "Data for Patient " + model.patientID
})
.SetXAxis(new XAxis
{
Categories = new[] { "Clump Thickness", "Uniformity Size", "Uniformity Shape", "Marginal Adhesion", "Single Epithelial Cell Size",
"Bare Nuclei",
"Bland Chromatin",
"Normal Nucleoli",
"Mitosis" },
Title = new XAxisTitle {
Text = "Cell Attributes"
}
})
.SetYAxis(new YAxis
{
Min = 0,
Title = new YAxisTitle
{
Text = "Value 1-10",
Align = AxisTitleAligns.High
}
})
.SetTooltip(new Tooltip {
Formatter = "function() { return ''+ this.series.name +': '+ this.y; }"
})
.SetPlotOptions(new PlotOptions
{
Bar = new PlotOptionsBar
{
DataLabels = new PlotOptionsBarDataLabels {
Enabled = false
}
}
})
.SetCredits(new Credits {
Enabled = false
})
.SetSeries(new[]
{
new Series {
Name = "Average Benign Patient", Data = new Data(new object[] { numbers[0][0], numbers[1][0], numbers[2][0], numbers[3][0], numbers[4][0], numbers[5][0],
numbers[6][0], numbers[7][0], numbers[8][0] })
},
new Series {
Name = "Average Malignant Patient", Data = new Data(new object[] { numbersM[0][0], numbersM[1][0], numbersM[2][0], numbersM[3][0], numbersM[4][0], numbersM[5][0],
numbersM[6][0], numbersM[7][0], numbersM[8][0] })
},
new Series {
Name = "Patient " + model.patientID.ToString(), Data = new Data(new object[] { model.clumpThickness, model.uniformitySize, model.uniformityShape, model.marginalAdhesion,
model.singleEpithelial, model.bareNuclei, model.blandChromatin, model.normalNucleoli, model.mitosis })
}
});
return(View(chart));
}
}
