Added normalfunc calcs

src/experiments/algorithms/distribution.py

@@ -0,0 +1,37 @@
+import numpy as np
+import matplotlib.pyplot as plt
+import math
+
+mean = 125
+sd = 20
+N = 255
+f = np.zeros(N, dtype=np.longdouble)
+
+def calcNormalFunc(mean, sd, len):
+    f = np.zeros(len, dtype=np.longdouble)
+    
+    # calculate PDF
+    for x in range(len):
+        exp = (-(x - mean) ** 2)/(2 * sd ** 2)
+        f[x] = 1 / math.sqrt(2 * np.pi * sd** 20 ) * (math.exp(exp))
+
+    # normalize PDF
+    max = np.amax(f)
+    min = np.amin(f)
+    for x in range(len):
+        f[x] = (f[x] - min) / (max - min)
+
+    return f
+
+f = calcNormalFunc(mean, sd, N)
+
+plt.title("PDF: $\mu = %d$, $\sigma = %d$"%(mean, sd))    
+plt.plot(f)
+plt.grid()
+plt.xlim(0, 255)
+plt.ylim(0, 1.05)
+plt.show()
+
+print("Score at 1 sigma: %f"%f[sd+mean])
+print("Score at 2 sigma: %f"%f[2*sd+mean])
+print("Score at 3 sigma: %f"%f[3*sd+mean])

src/helpers/statistics.py

@@ -1,4 +1,5 @@
 import numpy as np
+import math
 
 def imgStats(img):
     mean = np.zeros(3)
@@ -6,4 +7,20 @@ def imgStats(img):
     for i in range(img.shape[2]):
         mean[i] = np.mean(img[:, :, i])
         std[i] = np.std(img[:, :, i])
-    return mean, std
+    return mean, std
+
+def calcNormalFunc(mean, sd, len):
+    f = np.zeros(len, dtype=np.longdouble)
+    
+    # calculate PDF
+    for x in range(len):
+        exp = (-(x - mean) ** 2)/(2 * sd ** 2)
+        f[x] = 1 / math.sqrt(2 * np.pi * sd** 20 ) * (math.exp(exp))
+
+    # normalize PDF
+    max = np.amax(f)
+    min = np.amin(f)
+    for x in range(len):
+        f[x] = (f[x] - min) / (max - min)
+
+    return f