mlCommom

pkg/ml/mlCommon.go

@@ -19,7 +19,7 @@ func GetCategory(num int) string {
 	}
 }
 
-//Activation function - two class classifier.
+// Activation function - two class classifier.
 func BinaryClassifier(z float64) int {
 	if z >= 0.5 {
 		return 1
@@ -28,12 +28,12 @@ func BinaryClassifier(z float64) int {
 	}
 }
 
-//Sigmoid function.
+// Sigmoid function.
 func Sigmoid(z float64) float64 {
 	return 1.0 / (1 + math.Exp(-1*z))
 }
 
-//Net input (z.)
+// Net input (z.)
 func NetInput(w []float64, intercept float64, x []float64) float64 {
 	var z float64 = 0
 
@@ -44,7 +44,7 @@ func NetInput(w []float64, intercept float64, x []float64) float64 {
 	return z + intercept
 }
 
-// Standard Scaler.
+// Standard Scaler. done for normalization
 func StdScaler(x []float64, mu []float64, std []float64) []float64 {
 	scaledX := []float64{}
 
@@ -58,10 +58,10 @@ func StdScaler(x []float64, mu []float64, std []float64) []float64 {
 // Offline values of means , standard deviation from standard scaling and weights.
 func ModelParameters() ([]float64, float64, []float64, []float64) {
 	// Weights from logistic regression
-	f, errWt := os.Open(constants.WeightsFile)
+	f, errWt := os.Open(constants.WeightsFile) //weights.txt
 
 	if errWt != nil {
-		logrus.Info(errWt)
+		logrus.Info(errWt) //break
 	}
 
 	defer f.Close()