added code for create image tiles

Author: ishaan95

README.md

@@ -1,2 +1,6 @@
 # MapDatasetGenerator
 Generate and load dataset of road network maps.
+
+# Quick start
+* Use requirements file to install pip packages.
+* Run `maps.py` script. It will generate many PIL Image objects. 

data/input/sf_layered.txt

@@ -0,0 +1,92 @@
+from torch.utils.data import Dataset
+from PIL import Image
+import torch.nn.functional as F
+import os
+from tqdm.notebook import tqdm
+import torch
+import numpy as np
+import random
+
+# Credit - Bahar
+class MapReader:
+    def __init__(self, filename):
+        f = open(filename, 'r')
+        data = f.read().split()
+        self.size = (int(data[0]), int(data[1]))
+        self.data = [[int(data[i * self.size[1] + j + 2]) for j in range(self.size[1])] for i in range(self.size[0])]
+
+# Converter object that can return character for OpenStreetMap layer group
+# Credit - Bahar
+class LayerToCharConverter:
+    def __init__(self, char_groups=[[0], [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]]):
+        self.char_groups = char_groups
+        self.char_size = len(self.char_groups)
+
+    def get_char(self, layer):
+        for i in range(self.char_size):
+            if layer in self.char_groups[i]:
+                return i
+        raise Exception("layer not available in char groups: {}".format(layer))
+
+multi_layer_converter = LayerToCharConverter([[0], [1, 2, 3, 4, 5], [6, 7], [8, 9, 10, 11, 12]])
+single_layer_converter = LayerToCharConverter([[0], [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]])
+
+# New Cell
+# Credit - Bahar
+# Modifications - Ishaan
+
+class MapsDataset(Dataset):
+    def __init__(self, window_size, step_size, converter):
+        self.char_size = converter.char_size
+        self.converter = converter
+        self.window_size = window_size
+        self.step_size = step_size
+        self.samples = []
+        self.block_size = self.window_size[0] * self.window_size[1] - 1
+
+    def __len__(self):
+        return len(self.samples)
+
+    def __getitem__(self, idx):
+        sample = (self.samples[idx], self.samples[idx], self.samples[idx])
+        return torch.from_numpy(np.array(sample)).unsqueeze(0)
+        #flat = torch.from_numpy(np.array(sample)).view(-1)
+        #flat = flat[self.perm].float()
+        #return flat
+    
+    def add(self, mapReader):
+        for i in range(0, mapReader.size[0] - self.window_size[0] + 1, self.step_size):
+            for j in range(0, mapReader.size[1] - self.window_size[1] + 1, self.step_size):
+                self.samples.append([[
+                    (self.converter.get_char(mapReader.data[i + x][j + y]) / (len(self.converter.char_groups) - 1)) * -2 + 1
+                    for y in range(self.window_size[1])]
+                    for x in range(self.window_size[0])])
+    
+    def shuffle(self):
+        random.shuffle(self.samples)
+
+    def get_train_test(self, train_ratio):
+        train_dataset = MapsDataset(self.window_size, self.step_size, self.converter)
+        test_dataset = MapsDataset(self.window_size, self.step_size, self.converter)
+        train_size = int(len(self.samples) * train_ratio)
+        train_dataset.samples = self.samples[:train_size]
+        test_dataset.samples = self.samples[train_size:]
+        return train_dataset, test_dataset
+
+# Some utility functions
+def img_to_tensor(im):
+  return torch.tensor(np.array(im.convert('RGB'))/255).permute(2, 0, 1).unsqueeze(0) * 2 - 1
+
+def tensor_to_image(t):
+  return Image.fromarray(np.array(((t.squeeze().permute(1, 2, 0)+1)/2).clip(0, 1)*255).astype(np.uint8))
+
+def gather(consts: torch.Tensor, t: torch.Tensor):
+    """Gather consts for $t$ and reshape to feature map shape"""
+    c = consts.gather(-1, t)
+    return c.reshape(-1, 1, 1, 1)
+
+def map_img_to_tensor(im):
+  return torch.tensor(np.array(im.convert('L'))/255).unsqueeze(0) * 2 - 1
+
+def map_tensor_to_image(t):
+  return Image.fromarray(np.array(((t.squeeze()+1)/2).clip(0, 1)*255).astype(np.uint8))

maps.py

@@ -0,0 +1,5 @@
+matplotlib==3.5.3
+numpy==1.23.3
+Pillow==9.2.0
+torch==1.12.1
+tqdm==4.64.1