Update app.py

app.py

@@ -106,12 +106,17 @@ def KNN_predict(train_features, train_labels, test_feature, K):
   
   ### get final prediction
   final_prediction = scipy.stats.mode(major_class).mode[0]
+  
+  ### get frequency of classes
+  class_freq = {}
+  for i in range(0,10):
+      class_freq['digit '+str(i)] = float(major_class.count(i)) / len(major_class)
 
   ### get neighbor images and save to local 
   neighbor_imgs =np.array(neighbor_imgs) 
   image_path = plot_digits(neighbor_imgs, images_per_row=6)
 
-  return final_prediction, image_path
+  return final_prediction, class_freq, image_path
 
 ### main function for gradio to call to classify image
 def call_our_KNN(test_image, K=7):
@@ -128,13 +133,17 @@ set_image = gr.inputs.Image(shape=(28, 28), image_mode='L')
 set_K = gr.inputs.Slider(1, 24, step=1, default=7)
 
 set_label = gr.outputs.Textbox(label="Predicted Digit")
+# define output as the single class text
+set_probability = gr.outputs.Label(num_top_classes=10, label="Predicted Class")
+
 set_out_images = gr.outputs.Image(label="Closest Neighbors")
 
 
+
 ### configure gradio, detailed can be found at https://www.gradio.app/docs/#i_slider
 interface = gr.Interface(fn=call_our_KNN, 
                          inputs=[set_image, set_K], 
-                         outputs=[set_label,set_out_images],
+                         outputs=[set_label,set_probability,set_out_images],
                          examples_per_page = 2,
                          examples = sample_images, 
                          title="CSCI4750/5750 Demo 1: Digit classification using KNN algorithm",