diff --git a/.DS_Store b/.DS_Store
new file mode 100644
index 0000000..32baf48
Binary files /dev/null and b/.DS_Store differ
diff --git a/test.py b/test.py
deleted file mode 100644
index b7579ad..0000000
--- a/test.py
+++ /dev/null
@@ -1,13 +0,0 @@
-from utils.one_hot_encoder import HotEncoder
-import  numpy as np
-import torch
-import skimage
-from torch.nn.functional import one_hot
-img__= skimage.io.imread("/Users/madhav/DataSets/data_semantics/training/semantic_rgb/000035_10.tif")
-encoder= HotEncoder(dir= '/Users/madhav/DataSets/data_semantics/training/semantic_rgb', extension="tif", is_binary= False, n_classes= 30)
-color_dict= encoder.gen_colors()
-onehottt= encoder.PerPixelClassMatrix(img__)
-encoded= encoder.HotEncode(onehottt)
-ch= encoder.channelSplit(encoded)
-for _ in range(30):
-    skimage.io.imsave(arr= ch[_].reshape(375,1242), fname= str(_)+".jpeg")
diff --git a/utils/one_hot_encoder.py b/utils/one_hot_encoder.py
index 323b438..b70416e 100644
--- a/utils/one_hot_encoder.py
+++ b/utils/one_hot_encoder.py
@@ -51,6 +51,8 @@ def PerPixelClassMatrix(self, Image):
                 class_map[x][y]= self.color[clr]
         return class_map
     def HotEncode(self, class_map):
+        """Inputs the class map of the image and returns the hot encoded
+        N channel output. Each channel represents the presence of a class per pixel"""
         assert isinstance(class_map, np.ndarray), "Class map has to be an ndarray and not {}".format(type(class_map))
         shape_= class_map.shape
         encoded= np.zeros(shape= (shape_[0], shape_[1], self.n_classes), dtype=float)