{"cells":[{"metadata":{},"cell_type":"markdown","source":"# Computing Conway's Game of Life\n## A machine learning and computer vision approach\n\n<img style=\"width:400px\" src=\"https://upload.wikimedia.org/wikipedia/commons/e/e5/Gospers_glider_gun.gif\">\n\n## So What are the rules?\n\n### Living Cells: \n* **Overpopulation**: if a living cell is surrounded by more than three living cells, it dies.\n\n* **Stasis**: if a living cell is surrounded by two or three living cells, it survives.\n\n* **Underpopulation**: if a living cell is surrounded by fewer than two living cells, it dies.\n\n### Dead Cells\n* **Reproduction**: if a dead cell is surrounded by exactly three cells, it becomes a live cell.\n\n\n"},{"metadata":{},"cell_type":"markdown","source":"### These rules can be simplified by the following:\n\ncell has 3 living neighbors -> alive\n\ncell has 2 living neighbors and is alive -> alive\n\notherwise -> dead"},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport matplotlib.pyplot as plt\nimport scipy\nimport scipy.sparse\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"root = '/kaggle/input/conways-reverse-game-of-life-2020'\n\ntrain_df = pd.read_csv(root + '/train.csv')\ntest_df = pd.read_csv(root + '/test.csv')\nsample_df = pd.read_csv(root + '/sample_submission.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_df.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_df.info()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"start = [col for col in train_df.columns if 'start' in col]\nstop  = [col for col in train_df.columns if 'stop' in col]\n\ndef plot_game(idx):\n    fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(5,3), dpi=200)\n    ax1.imshow(train_df.loc[idx,start].values.reshape(25,25))\n    ax2.imshow(train_df.loc[idx,stop].values.reshape(25,25))\n\n    fig.suptitle('Delta: {}'.format(train_df.loc[idx, \"delta\"]))\n\n    ax1.set_title('Start')\n    ax2.set_title('Stop')\n\n    ax1.set_xticks([0,5,10,15,20]);\n    ax1.set_yticks([0,5,10,15,20]);\n    ax2.set_xticks([0,5,10,15,20]);\n    ax2.set_yticks([0,5,10,15,20]);","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plot_game(4)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"plot_game(3)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Strategy\n\n## 1) Understand the problem. \n - Consider computational and functional limitations.\n - How is this different from standarm ML problem statements?\n\n## 2) Solve/Model the \"Forward\" problem.\n- We can easily compute the state of conway's game at time t+1 given the state at time t.\n- It should be feasible to train an ML model to find that function too right?\n\n## 3) Solve/Model the \"Reverse\" Problem.\n- There are many approaces to take. \n- It's likely that an optimized brute force or hard coded model would do the best. (100% accuracy is possible)\n- These methods are difficult to implement and this is a project for a Data Mining class after all.\n- Instead let's exploit the local precision of Convolutional neural nets to find a nice local minimimum. \n\nConway's game of life is built from a few deceptively simple rules which lead to immense complexity. "},{"metadata":{},"cell_type":"markdown","source":"# 1) Understanding the Problem\n\n\n## What is the general machine learning problem?\n\n![image.png](attachment:image.png)","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"![image.png](attachment:image.png)","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"# Neural Network \n\n<img style=\"width:400px\" src=\"https://miro.medium.com/max/1804/1*f9XlMlruW7TMF3EHbPDfYg.png\">"},{"metadata":{},"cell_type":"markdown","source":"# We hope there is a particular function underlying the relationship\n\n---------------------------------------------\n"},{"metadata":{},"cell_type":"markdown","source":"![Untitled%20Diagram%20%284%29.png](attachment:Untitled%20Diagram%20%284%29.png)","attachments":{"Untitled%20Diagram%20%284%29.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"# 2) The \"Forward\" problem."},{"metadata":{},"cell_type":"markdown","source":"# Dataset Configuration\n\n## Data Genereration\n\nRather than simply training on the dataset provided by kaggle, we can improve the learning rate by generating more board states.\n"},{"metadata":{"trusted":true},"cell_type":"code","source":"# Credit: This was inspired by 's getting started notebook\n\ndef update_cell(board,i,j):\n    neighbors = sum(board[k % len(board)][l % len(board)] \n                        for k in range(i-1,i+2)\n                            for l in range(j-1,j+2)\n                                if k != i or l != j)\n    return 1 if neighbors == 3 or (neighbors == 2 and board[i][j]) else 0\n\ndef step_forward(board):\n    return np.array([np.array([update_cell(board,i,j) for j in range(len(board))]) \n                                                         for i in range(len(board))])","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Sanity Check\n\n## step_forward function should update board in accordance w/ data set"},{"metadata":{"trusted":true},"cell_type":"code","source":"idx = 40\n\nd = train_df.iloc[idx,1]\nstart = train_df.iloc[idx,2:627].values.reshape(25,25)\nstop = train_df.iloc[idx,627:].values.reshape(25,25)\n\nprint('delta = {}'.format(d))\nplt.figure(figsize=(5,4), dpi=140)\nplt.imshow(start)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"pred = start\n\nfor _ in range(d):\n    pred = step_forward(pred)\n\nfig, (ax1, ax2) = plt.subplots(1, 2, figsize=(7,5), dpi=140)\nax1.set_title('kaggle dataset')\nax2.set_title('step_forward')\nax1.imshow(stop)\nax2.imshow(pred)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def generate_game(delta=1):\n    density = np.random.random() * 0.98 + 0.01\n    start   = scipy.sparse.random(25, 25, density=density, data_rvs=np.ones).toarray()\n    stop    = start\n    \n    for _ in range(delta):\n        stop = step_forward(stop)\n        \n    if np.count_nonzero(start) == 0: return generate_game()\n    if np.count_nonzero(stop) == 0: return generate_game()\n    \n    return delta,start,stop\n\ndef generate_games(n=10000,delta=1):\n    games = []\n    for i in range(n):\n        if (i + 1) % 1000 == 0: print('{} games created'.format(i+1))\n        games.append(generate_game(delta))\n    return games ","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"games = generate_games()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"markdown","source":"# Pytorch"},{"metadata":{"trusted":true},"cell_type":"code","source":"import torch\nimport torchvision\nimport torch.nn as nn\nimport torch.nn.functional as F\n\nfrom torchvision.transforms import ToTensor\nfrom torch.utils.data.dataloader import DataLoader\nfrom torch.utils.data import Dataset, random_split","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## CglDataset: custom dataset class"},{"metadata":{"trusted":true},"cell_type":"code","source":"class CglDataset(Dataset):\n    def __init__(self, data=[], transform=None):\n        self.data = data[:]\n        self.transform = transform\n\n    def __len__(self):\n        return len(self.data)\n\n    def __getitem__(self, n):\n        delta,start,stop = self.data[n]\n        if self.transform:\n            stop  = self.transform(stop).float()\n            start = self.transform(start).float()\n        return (delta, start, stop)\n    \n    def loadFromDataFrame(self, df):\n        for i in range(len(df)):\n            data = df.iloc[i]\n            delta = data[1]\n            start = data[2:627].values.reshape((25,25)).astype(np.float32)\n            stop = data[627:].values.reshape((25,25)).astype(np.float32)\n            self.data.append((delta, start, stop))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_ds = CglDataset(games, transform=ToTensor())\n#delta, start, stop = train_ds[3]\n#print(delta)\n\n#fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(7,5), dpi=140)\n#ax1.set_title('start')\n#ax2.set_title('stop')\n#ax1.imshow(start[0,:,:])\n#ax2.imshow(stop[0,:,:])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(len(train_ds))\ndelta_one = train_df[train_df['delta'] == 1]\ntrain_ds.loadFromDataFrame(delta_one)\nprint(len(train_ds))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Split into separate train and validation sets"},{"metadata":{"trusted":true},"cell_type":"code","source":"train_ds, val_ds = random_split(train_ds, [15000, 5113])\nlen(train_ds), len(val_ds)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"batch_size = 128\n\ntrain_loader = DataLoader(train_ds, batch_size, shuffle=True)\nval_loader = DataLoader(val_ds, batch_size)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"for delta, start, stop in train_loader:\n    print(stop.shape)\n    print(start.shape)\n    break","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Solving forward problem with Deep Learning\n\n## Feed Forward NN: CGL Neural Network class \nhidden layers: 2"},{"metadata":{},"cell_type":"markdown","source":"# Loss Function\n\n![image.png](attachment:image.png)","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{"trusted":true},"cell_type":"code","source":"class CglNNet(nn.Module):\n    def __init__(self, l1_size, l2_size):\n        super().__init__()\n        self.fc1 = nn.Linear(625, l1_size)\n        self.fc2 = nn.Linear(l1_size, l2_size)\n        self.fc3 = nn.Linear(l2_size, 625)\n        \n    def forward(self, x):\n        x = x.reshape(-1, 625)\n        x = self.fc1(x)\n        x = F.relu(x)\n        x = self.fc2(x)\n        x = F.relu(x)\n        x = self.fc3(x)\n        return x\n        \n    \n    @torch.no_grad()\n    def evaluate(self, loader, loss_fn):\n        loss, accur = [], []\n        for delta, start, stop in loader:\n            out = self(start)\n            target = stop.reshape(-1, 625)\n            loss.append(loss_fn(out, target))\n            accur.append(self.accuracy(out, target))\n        mean_loss  = torch.stack(loss).mean()\n        mean_accur = torch.stack(accur).mean()\n        return mean_loss, mean_accur\n    \n    def accuracy(self, outputs, expected):\n        res = torch.sum(self.cast_int(outputs) == self.cast_int(expected)).cpu().numpy() / np.prod(outputs.shape)\n        return torch.tensor(res)\n    \n    def cast_int(self, num):\n        return self.cast_bool(num).to(torch.int8)\n    \n    def cast_bool(self, x):\n        return (x > 0.5)\n    \n    @staticmethod\n    def weights_init(layer):\n        if isinstance(layer, (nn.Conv2d, nn.ConvTranspose2d)):\n            nn.init.kaiming_normal_(layer.weight)\n            nn.init.constant_(layer.bias, 0.1)\n","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# GPU config"},{"metadata":{"trusted":true},"cell_type":"code","source":"torch.cuda.is_available()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_default_device():\n    \"\"\"Pick GPU if available, else CPU\"\"\"\n    if torch.cuda.is_available():\n        return torch.device('cuda')\n    else:\n        return torch.device('cpu')\n    \n    \ndef to_device(data, device):\n    \"\"\"Move tensor(s) to chosen device\"\"\"\n    if isinstance(data, (list,tuple)):\n        return [to_device(x, device) for x in data]\n    return data.to(device, non_blocking=True)\n    \ndevice = get_default_device()\ndevice","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"class DeviceDataLoader():\n    \"\"\"Wrap a dataloader to move data to a device\"\"\"\n    def __init__(self, dl, device):\n        self.dl = dl\n        self.device = device\n        \n    def __iter__(self):\n        \"\"\"Yield a batch of data after moving it to device\"\"\"\n        for b in self.dl: \n            yield to_device(b, self.device)\n\n    def __len__(self):\n        \"\"\"Number of batches\"\"\"\n        return len(self.dl)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_loader = DeviceDataLoader(train_loader, device)\nval_loader = DeviceDataLoader(val_loader, device)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"layer1 = 1024\nlayer2 = 512\n\n# Neural Net\nmodel = CglNNet(layer1, layer2)\nto_device(model, device)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(model.fc1.weight.shape, model.fc1.bias.shape)\nprint(model.fc2.weight.shape, model.fc2.bias.shape)\nprint(model.fc3.weight.shape, model.fc3.bias.shape)\n\nmodel.fc1.weight","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def nn_fit(epochs, lr, model, train_loader, val_loader, opt_func=torch.optim.SGD, loss_fn=nn.MSELoss()):\n    \n    optimizer = opt_func(model.parameters(), lr)\n    history = []\n    \n    for e in range(epochs):\n        \n        # Training Phase \n        for delta, start, stop in train_loader:\n            out = model(start)\n            loss = loss_fn(out, stop.reshape(-1, 625))\n            loss.backward()\n            optimizer.step()\n            optimizer.zero_grad()\n        \n        # Validation phase\n        val_loss, val_accur = model.evaluate(val_loader, loss_fn)\n        print('Epoch [{} / {}], Validation Loss: {}, Validation Accuracy: {}'.format(e+1, epochs, val_loss, val_accur))\n        history.append((val_loss, val_accur))\n\n    return history","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"history = nn_fit(20, .05, model, train_loader, val_loader)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Convolutional Neural Network: CGL CNN class \n\n### Layers\n1) 2D Convolution with 9 Channel(s): kernel_size = 3, padding = 1, out = (9,n,25,25)\n\n2) ReLU Activation function to introduce non-linearity\n\n3) 2D Convolution with 1 Channel(s): kernel_size = 1, padding = 0, out = (1,n,25,25)\n\n4) Sigmoid function to compress outputs into (0,1) interval"},{"metadata":{"trusted":true},"cell_type":"code","source":"class CglCNN(nn.Module):\n    def __init__(self):\n        super().__init__() \n\n        self.conv1   = nn.Conv2d(1, 9, kernel_size=(3,3), padding=1, padding_mode='circular') #padding_mode='circular'\n        self.conv2   = nn.Conv2d(9, 1,  kernel_size=(1,1))\n        self.apply(self.weights_init)\n\n\n    def forward(self, x):\n        x = self.cast_inputs(x)\n\n        x = self.conv1(x)\n        x = F.relu(x)\n\n        x = self.conv2(x)\n        x = torch.sigmoid(x)\n        \n        return x\n    \n    @torch.no_grad()\n    def evaluate(self, loader, loss_fn):\n        loss, accur = [], []\n        for delta, start, stop in loader:\n            out = self(start)\n            target = stop\n            loss.append(loss_fn(out, target))\n            accur.append(self.accuracy(out, target))\n        mean_loss  = torch.stack(loss).mean()\n        mean_accur = torch.stack(accur).mean()\n        return mean_loss, mean_accur\n    \n    def accuracy(self, outputs, expected):\n        res = torch.sum(self.cast_int(outputs) == self.cast_int(expected)).cpu().numpy() / np.prod(outputs.shape)\n        return torch.tensor(res)\n    \n    def cast_int(self, num):\n        return self.cast_bool(num).to(torch.int8)\n    \n    def cast_bool(self, x):\n        return (x > 0.5)\n    \n    @staticmethod\n    def weights_init(layer):\n        if isinstance(layer, (nn.Conv2d, nn.ConvTranspose2d)):\n            nn.init.kaiming_normal_(layer.weight)\n            nn.init.constant_(layer.bias, 0.1)\n            \n    def cast_to_tensor(self, x):\n        if torch.is_tensor(x):\n            return x.to(torch.float32).to(device)\n        if isinstance(x, list):\n            x = np.array(x)\n        if isinstance(x, np.ndarray):\n            x = torch.from_numpy(x).to(torch.float32)\n            x = x.to(device)\n            return x  # x.shape = (42,3)\n        raise TypeError(f'{self.__class__.__name__}.cast_to_tensor() invalid type(x) = {type(x)}')\n\n\n    # Credit: \n    def cast_inputs(self, x):\n        x = self.cast_to_tensor(x)\n        if x.dim() == 1:             # single row from dataframe\n            x = x.view(1, 1, torch.sqrt(x.shape[0]), torch.sqrt(x.shape[0]))\n        elif x.dim() == 2:\n            if x.shape[0] == x.shape[1]:  # single 2d board\n                x = x.view(1, 1, x.shape[0], x.shape[1])\n            else: # rows of flattened boards\n                x = x.view(-1, 1, torch.sqrt(x.shape[1]), torch.sqrt(x.shape[1]))\n        elif x.dim() == 3:                                 # numpy  == (batch_size, height, width)\n            x = x.view(x.shape[0], 1, x.shape[1], x.shape[2])   # x.shape = (batch_size, channels, height, width)\n        elif x.dim() == 4:\n            pass  # already in (batch_size, channels, height, width) format, so do nothing\n        return x","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Convolutional Neural Net\ncnn_model = CglCNN()\nto_device(cnn_model, device)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def cnn_fit(epochs, lr, model, train_loader, val_loader, opt_func=torch.optim.SGD, loss_fn=nn.MSELoss()):\n    \n    optimizer = opt_func(model.parameters(), lr)\n    history = []\n    \n    for e in range(epochs):\n        \n        # Training Phase \n        for delta, start, stop in train_loader:\n            out = model(start)\n            loss = loss_fn(out, stop)\n            loss.backward()\n            optimizer.step()\n            optimizer.zero_grad()\n        \n        # Validation phase\n        val_loss, val_accur = model.evaluate(val_loader, loss_fn)\n        print('Epoch [{} / {}], Validation Loss: {}, Validation Accuracy: {}'.format(e+1, epochs, val_loss, val_accur))\n        history.append((val_loss, val_accur))\n\n    return history","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"history = cnn_fit(12, 0.5, cnn_model, train_loader, val_loader)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Predictions and Analysis"},{"metadata":{"trusted":true},"cell_type":"code","source":"losses, accuracies = zip(*history)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"\n## Validation Loss"},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.plot(losses)\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Validation Accuracy"},{"metadata":{"trusted":true},"cell_type":"code","source":"plt.plot(accuracies, color='#bca244')\nplt.show()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Computing future board state, delta = 1"},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_test_batch(delta=1):\n    df_delta = train_df[train_df['delta'] == delta]\n    \n    test_ds = CglDataset(transform=ToTensor())\n    test_ds.loadFromDataFrame(df_delta)\n    return test_ds\n\ndef plot_comparison(start, truth, pred):\n    fig, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(8,6), dpi=140)\n    ax1.set_title('start')\n    ax2.set_title('kaggle')\n    ax3.set_title('CNN pred')\n    ax1.imshow(start.cpu().reshape(25,25))\n    ax2.imshow(truth.cpu().reshape(25,25), cmap='gray')\n    ax3.imshow(pred.cpu().reshape(25,25), cmap='gray')\n    \ndef cnn_feed_forward(state, iters=1):\n    for i in range(iters): state = cnn_model.cast_int(cnn_model(state))\n    return state ","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"batch = get_test_batch()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"delta, start, stop = batch[20]\n\nmodel_pred = cnn_feed_forward(start)\nplot_comparison(start, stop, model_pred)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Computing future board state, delta = 4"},{"metadata":{"trusted":true},"cell_type":"code","source":"batch = get_test_batch(delta = 4)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"delta, start, stop = batch[0]\n\nmodel_pred = cnn_feed_forward(start, iters=delta)\nplot_comparison(start, stop, model_pred)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# 2) The \"Reverse\" problem."},{"metadata":{"trusted":true},"cell_type":"code","source":"class CglCNN_Reverse(nn.Module):\n    def __init__(self):\n        super().__init__() \n\n        self.conv1   = nn.Conv2d(1, 128, kernel_size=(3,3), padding=1, padding_mode='circular') #padding_mode='circular'\n        self.conv2   = nn.Conv2d(128, 1,  kernel_size=(1,1))\n        self.bnorm1  = nn.BatchNorm2d(512)\n        self.apply(self.weights_init)\n\n\n    def forward(self, x):\n        x = self.cast_inputs(x)\n\n        x = self.conv1(x)\n        x = F.leaky_relu(x)\n        x = self.bnorm1(x)\n        x = self.conv2(x)\n        x = torch.sigmoid(x)\n        \n        return x\n    \n    @torch.no_grad()\n    def evaluate(self, loader, loss_fn):\n        loss, accur = [], []\n        for delta, _, stop in loader:\n            start = self(stop)\n            out = cnn_model(start)\n            target = stop\n            loss.append(loss_fn(out, target))\n            accur.append(self.accuracy(out, target))\n        mean_loss  = torch.stack(loss).mean()\n        mean_accur = torch.stack(accur).mean()\n        return mean_loss, mean_accur\n    \n    def accuracy(self, outputs, expected):\n        res = torch.sum(self.cast_int(outputs) == self.cast_int(expected)).cpu().numpy() / np.prod(outputs.shape)\n        return torch.tensor(res)\n    \n    def cast_int(self, num):\n        return self.cast_bool(num).to(torch.int8)\n    \n    def cast_bool(self, x):\n        return (x > 0.5)\n    \n    @staticmethod\n    def weights_init(layer):\n        if isinstance(layer, (nn.Conv2d, nn.ConvTranspose2d)):\n            nn.init.kaiming_normal_(layer.weight)\n            nn.init.constant_(layer.bias, 0.1)\n            \n    def cast_to_tensor(self, x):\n        if torch.is_tensor(x):\n            return x.to(torch.float32).to(device)\n        if isinstance(x, list):\n            x = np.array(x)\n        if isinstance(x, np.ndarray):\n            x = torch.from_numpy(x).to(torch.float32)\n            x = x.to(device)\n            return x  # x.shape = (42,3)\n        raise TypeError(f'{self.__class__.__name__}.cast_to_tensor() invalid type(x) = {type(x)}')\n\n\n    # Credit: \n    def cast_inputs(self, x):\n        x = self.cast_to_tensor(x)\n        if x.dim() == 1:             # single row from dataframe\n            x = x.view(1, 1, torch.sqrt(x.shape[0]), torch.sqrt(x.shape[0]))\n        elif x.dim() == 2:\n            if x.shape[0] == x.shape[1]:  # single 2d board\n                x = x.view(1, 1, x.shape[0], x.shape[1])\n            else: # rows of flattened boards\n                x = x.view(-1, 1, torch.sqrt(x.shape[1]), torch.sqrt(x.shape[1]))\n        elif x.dim() == 3:                                 # numpy  == (batch_size, height, width)\n            x = x.view(x.shape[0], 1, x.shape[1], x.shape[2])   # x.shape = (batch_size, channels, height, width)\n        elif x.dim() == 4:\n            pass  # already in (batch_size, channels, height, width) format, so do nothing\n        return x","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Convolutional Neural Net (Reverse Case)\ncnn_model_reverse = CglCNN()\nto_device(cnn_model_reverse, device)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def cnn_fit_reverse(epochs, lr, model, train_loader, val_loader, opt_func=torch.optim.SGD, loss_fn=nn.MSELoss()):\n    \n    optimizer = opt_func(model.parameters(), lr)\n    history = []\n    \n    for e in range(epochs):\n        \n        # Training Phase \n        for delta, _, stop in train_loader:\n            start = model(stop)\n            out = cnn_model(start)\n            loss = loss_fn(out, stop)\n            loss.backward()\n            optimizer.step()\n            optimizer.zero_grad()\n        \n        # Validation phase\n        val_loss, val_accur = model.evaluate(val_loader, loss_fn)\n        print('Epoch [{} / {}], Validation Loss: {}, Validation Accuracy: {}'.format(e+1, epochs, val_loss, val_accur))\n        history.append((val_loss, val_accur))\n\n    return history","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}