From ef562590d1105a87653480efc1c5b831e8f14387 Mon Sep 17 00:00:00 2001
From: HombreLaser <sebastian-440@live.com>
Date: Fri, 20 Oct 2023 15:17:53 -0600
Subject: Fix neural network

---
 dataset_utils.py  | 10 ++++++++--
 main.py           | 17 +++++++++++++++++
 neural_network.py | 25 +++++++++++++++++--------
 utils.py          |  8 +++-----
 4 files changed, 45 insertions(+), 15 deletions(-)
 create mode 100644 main.py

diff --git a/dataset_utils.py b/dataset_utils.py
index fda75d2..927a78a 100644
--- a/dataset_utils.py
+++ b/dataset_utils.py
@@ -1,4 +1,4 @@
-import numpy
+import numpy as np
 from PIL import Image
 from pathlib import Path
 
@@ -16,8 +16,14 @@ class DatasetUtils:
                 image = Image.open(str(file)).convert(mode='L')
                 # Return the image's pixel values as an array alongside
                 # the character that it represents.
-                yield (dir.name, numpy.asarray(image).ravel())
+                yield (dir.name, np.asarray(image))
 
+    """
+    Get an image from the dataset.
+    """
+    def get_image(self, path: str):
+        return np.asarray(Image.open(f"{self.data_path}/{path}").convert('1'))
+                
     """Search for a file in the dataset."""
     def search(self, filename: str):
         pass
diff --git a/main.py b/main.py
new file mode 100644
index 0000000..9ca0a13
--- /dev/null
+++ b/main.py
@@ -0,0 +1,17 @@
+from PIL import Image
+from dataset_utils import DatasetUtils
+from neural_network import NeuralNetwork
+
+LEARNING_RATE = .5
+INPUT_RESOLUTION = 278
+
+def main():
+    data = DatasetUtils()
+    image = data.get_image('А/5a2f3c19c27bb.png')
+    neural_network = NeuralNetwork(LEARNING_RATE, INPUT_RESOLUTION)
+
+    print(neural_network.guess(image))
+
+
+if __name__ == "__main__":
+    main()
diff --git a/neural_network.py b/neural_network.py
index 90fd304..68592f5 100644
--- a/neural_network.py
+++ b/neural_network.py
@@ -1,6 +1,7 @@
 import numpy as np
 from scipy.special import expit
 from utils import random_array
+from dataset_utils import DatasetUtils
 
 CYRILLIC_ALPHABET = ['I', 'А', 'Б', 'В', 'Г', 'Д', 'Е', 'Ë', 'Ж', 'З',
                      'И', 'Й', 'К', 'Л', 'М', 'Н', 'О', 'П', 'Р', 'С',
@@ -11,11 +12,10 @@ CYRILLIC_ALPHABET = ['I', 'А', 'Б', 'В', 'Г', 'Д', 'Е', 'Ë', 'Ж', 'З',
 class NeuralNetwork:
     def __init__(self, learning_rate: float, input_resolution: int) -> None:
         self.learning_rate = learning_rate
-        self.output_layer_size = len(CYRILLIC_ALPHABET)
-        self.input_layer_size = input_resolution ** 2
-        self.hidden_layer_size = round((self.input_layer_size + self.output_layer_size) / 2)
-        self._hidden_weights = random_array(self.hidden_layer_size, self.input_layer_size)
-        self._output_weights = random_array(self.output_layer_size, self.hidden_layer_size)
+        self.input_layer_size = input_resolution
+        self.hidden_layer_size = len(CYRILLIC_ALPHABET)
+        self._hidden_weights = np.random.rand(input_resolution, self.hidden_layer_size)
+        self._output_weights = np.random.rand(input_resolution)
 
 
     """
@@ -36,8 +36,10 @@ class NeuralNetwork:
     Guess the letter contained in the image file pointed by
     input_image (a path).
     """
-    def guess(self, input_image: str) -> str:
-        pass
+    def guess(self, input_image: np.array) -> str:
+        output_layer = self._predict(input_image)
+
+        return self.out(output_layer)
 
     """
     Save the weights to a csv file.
@@ -58,7 +60,7 @@ class NeuralNetwork:
         hidden_layer_inputs = np.dot(input_layer, self._hidden_weights)
         hidden_layer_outputs = self._get_layer_output(hidden_layer_inputs)
         output_layer_inputs = np.dot(self._output_weights, hidden_layer_outputs)
-
+        
         # The output layer outputs. (Final output of the neural network).
         return self._get_layer_output(output_layer_inputs)
 
@@ -67,3 +69,10 @@ class NeuralNetwork:
     """
     def _get_layer_output(self, layer: np.array):
         return expit(layer)
+
+    """
+    Get the result from a sigmoid matrix (the index with the highest chance
+    of being the correct answer).
+    """
+    def out(self, output_layer: np.array):
+        return CYRILLIC_ALPHABET[np.argmax(np.transpose(output_layer))]
diff --git a/utils.py b/utils.py
index d068023..a4f6cd3 100644
--- a/utils.py
+++ b/utils.py
@@ -1,6 +1,4 @@
-import numpy
-from random import random
+import numpy as np
 
-def random_array(rows: int, columns: int):
-    random_matrix = [[random() for x in range(rows)] for y in range(columns)]
-    return numpy.asarray(random_matrix)
+def random_array(length: int):
+    return np.random.rand(length, 1)
-- 
cgit v1.2.3