Some refactoring, some TODOs

Pipfile

@@ -14,4 +14,4 @@ scipy = "==1.4.1"
 wandb = "==0.8.36"
 
 [requires]
-python_full_version = "3.8.2"
+python_version = "3.8"

digits.py

@@ -10,10 +10,7 @@
 
 
 def main():
-	if not os.path.isfile('data/mnist_py3_deskewed.pkl.gz'):
-		x = input('Deskewed data not found, generate now? (y/n): ')
-		if x.lower() == 'y':
-			deskew_data()
+	check_deskewed_files_exist()
 
 	if os.path.isfile('networks/network.json'):
 		x = input('Trained network found, train new network anyways? (y/n: ')
@@ -24,7 +21,8 @@ def main():
 			new_dir = os.path.join(archive_dir, '1')
 			os.rename(current_dir, new_dir)
 
-	# TODO: Refactor these functions to make sense
+	# REVIEW: Pass the loaded data as a global
+	#  so that it can be used across runs?
 	training, validation, test = load_data()
 	wandb.init(project='digits')
 	n = train([784, 128, 10], 0.008, 0.2, 0.05, training, validation)
@@ -40,5 +38,13 @@ def gui():
 	run_gui(n)
 
 
+def check_deskewed_files_exist():
+	# TODO: Check for all parts of deskewed dataset
+	if not os.path.isfile('data/mnist_py3_deskewed.pkl.gz'):
+		x = input('Deskewed data not found, generate now? (y/n): ')
+		if x.lower() == 'y':
+			deskew_data()
+
+
 if __name__ == '__main__':
 	main()

src/gui.py

@@ -5,7 +5,7 @@
 
 from globals import current_dir
 from preprocessor import deskew_image, dots_to_image
-from utils import draw_digit, save_digit
+from utils import save_digit
 
 
 class InputGUI:
@@ -26,6 +26,10 @@ def __init__(self, root, n=None):
 		self.label = tk.Label(root, text=desc)
 		self.label.grid(column=1, row=0, columnspan=3)
 
+		# TODO: Add text field(s) to display prediction and confidence
+		# REVIEW: Perhaps allow user to enter correct answer and save
+		#  as additional test data?
+		#  Perhaps make new training data out of it?
 		self.canvas = tk.Canvas(
 			self.root, width=self.size, height=self.size,
 			highlightthickness=2, highlightbackground='black'