creates koffka adelson automatically for different ppd and returns th…

…e stimulus and mask

stimuli/illusions/koffka_adelson_auto_including_mask.py

@@ -0,0 +1,111 @@
+import numpy as np
+from stimuli.utils import resize_array, write_array_to_image
+import PIL as pil
+import matplotlib.pyplot as plt
+import matplotlib.image as mpimg
+
+# choose values by hand 
+value_1 = 255
+value_2 = 200
+value_3 = 165
+value_4 = 130
+
+# choose ppd 
+ppd = 64
+
+def ini_matrix(ppd):
+    a = []
+    for i in range(ppd):
+        a.append(np.zeros(ppd))
+    return(a)
+
+def get_mask(arr_list, ppd):
+    mask = np.array(ini_matrix(ppd))
+    target1 = arr_list[0]
+    target2 = arr_list[1]
+
+
+    for i in range(int(ppd/16)):
+        y_diff = target1[2] - target1[0] +1
+        for j in range(y_diff):
+            y = target1[0]
+            x = target1[1]
+            mask[y+j][x+i] = 1 
+
+    for i in range(int(ppd/16)):
+        y_diff = target2[2] - target2[0] +1
+        for j in range(y_diff):
+            y = target2[0]
+            x = target2[3]
+            mask[y+j][x+i] = 2
+
+    plt.imshow(mask)
+    return(mask)
+
+
+def get_array(num, row):
+    arr = np.array([row])
+    for i in range(num):
+        arr_add = np.array([row])
+        arr = (np.concatenate((arr,arr_add)))
+    return arr
+
+def koffka_mask():
+    arr_1 = np.array([int(ppd/2),int(ppd *(3/4)),int(ppd *(3/4)),int(ppd *(3/4)) ])
+
+    arr_2 = np.array([int(ppd *(5/16)),int(ppd *(5/16)),int(ppd *(9/16)),int(ppd *(5/16))])
+    arr_list = [arr_1,arr_2]
+
+    mask = get_mask(arr_list,ppd)
+
+    return(mask)
+
+
+def koffka_adelson_auto():
+
+    a = np.ones((ppd,1), dtype = int)*value_4
+
+    b = np.ones((ppd,1), dtype = int)*value_4
+    x = int(ppd/8)
+    y = int(ppd*(11/16))
+    b[x:y] = value_2
+
+    c = np.ones((ppd,1), dtype = int)*value_4
+    x1 = int(ppd/8)
+    y1 = int(ppd*(5/16))
+    c[x1:y1] = value_2
+    x2 = int(ppd/2)
+    y2 = int(ppd*(11/16))
+    c[x2:y2] = value_2
+
+    d = np.ones((ppd,1), dtype = int)*value_1
+    x1 = int(ppd*(5/16))
+    y1 = int(ppd/2)
+    d[x1:y1] = value_3
+    x2 = int(ppd*(11/16))
+    y2 = int(ppd*(7/8))
+    d[x2:y2] = value_3
+
+    e = np.ones((ppd,1), dtype = int)*value_1
+    x1 = x1
+    y1 = y2
+    e[x1:y1] = value_3
+
+    f = np.ones((ppd,1), dtype = int)*value_1
+
+    arr = np.concatenate((get_array(int(ppd*(3/16)),a),
+                         get_array(int(ppd*(3/16)),b),
+                         get_array(int(ppd*(2/16)),c),
+                         get_array(int(ppd*(2/16)),d),
+                         get_array(int(ppd*(3/16)),e),
+                         get_array(int(ppd*(3/16)),f)))
+
+    arr = np.fliplr(np.rot90((arr),3))
+    return(arr)
+
+if __name__ == '__main__':
+    #test
+    #plt.imshow(koffka_adelson_auto())
+    #koffka_mask()
+
+