{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true,"_kg_hide-input":false},"cell_type":"markdown","source":"# 3rd Place Solution Part: BestGuess + Optimizer\n\n\nMain architecture:\n![grafik.png](attachment:grafik.png)\n\n\n\nHi all,\n\nthanks for the really interesting competition. I quickly want to share my solution. It is split up into 2 main blocks:\n* **BestGuess**, which is an iterative neural network approach. It predicts the start state for delta=1. Repeating it delta times for delta>1\n* And an **Optimizer**, which picks a random 3x3 reagion in the input, tries every possible combination of the 3x3 grid at once and selects randomly between the ones which produce the lowest score. It is entirely implemented on the GPU.\n\n\nThis approach reached a score of 0.018 before merging. \n","attachments":{"grafik.png":{"image/png":"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"}}},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","collapsed":true,"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":false},"cell_type":"markdown","source":"## 1 delta BestGuess\nArchitecture:\n![grafik.png](attachment:grafik.png)\n\nThe network only predicts one cell at a time, because there are many possible start states which lead to the same end state. By giving the network its previous prediction, it is enough to just estimate which cells are more likely to be alive or dead. While at the beginning the prediction might be very uncertain, by adding more and more predicted cells, the possible start state combinations narrow down and the network gets more certain.\n\nTraining is done supervised by just marking random cells in an end state as unpredicted and setting the start state as target.\n\nHere is a visualization of the solving process:","attachments":{"grafik.png":{"image/png":"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"}}},{"metadata":{"_kg_hide-input":true,"trusted":true},"cell_type":"code","source":"import gc\n\nimport pandas as pd\nimport math\n\nimport cv2\nimport numpy as np\nimport matplotlib\nimport random\n\nfrom tensorboard.backend.event_processing import event_accumulator\n# matplotlib.use('agg')\nimport matplotlib.pyplot as plt\nimport time\nimport torch\nimport torch.nn as nn\nimport os\n\nimport torch.nn.functional as F\n\nfrom torch import optim\nfrom torch.utils.data import Dataset, DataLoader\nfrom torch.utils.tensorboard import SummaryWriter\nfrom matplotlib import animation, rc\nfrom IPython.display import HTML","execution_count":null,"outputs":[]},{"metadata":{"_kg_hide-input":true,"_kg_hide-output":true,"trusted":true},"cell_type":"code","source":"\n\n\ndef state_step(state: np.array):\n    neighbour_sum = \\\n        np.roll(state, -1, axis=0) + \\\n        np.roll(state, 1, axis=0) + \\\n        np.roll(state, -1, axis=1) + \\\n        np.roll(state, 1, axis=1) + \\\n        np.roll(np.roll(state, -1, axis=0), -1, axis=1) + \\\n        np.roll(np.roll(state, -1, axis=0), 1, axis=1) + \\\n        np.roll(np.roll(state, 1, axis=0), -1, axis=1) + \\\n        np.roll(np.roll(state, 1, axis=0), 1, axis=1)\n    out = np.zeros(state.shape, dtype=np.int)\n    out[neighbour_sum == 3] = 1\n    out[np.logical_and(neighbour_sum == 2, state == 1)] = 1\n    return out\n\n\nclass TiledConv2d(nn.Module):\n    def __init__(self, in_features, out_features):\n        super().__init__()\n        self.conv = nn.Conv2d(in_features, out_features, kernel_size=3, bias=False)\n\n    def forward(self, x):\n        return self.conv(F.pad(x, [1, 1, 1, 1], mode='circular'))\n\n\nclass TiledResBlock(nn.Module):\n    def __init__(self, features):\n        super().__init__()\n        self.main = nn.Sequential(\n            nn.BatchNorm2d(features),\n            nn.ReLU(True),\n            TiledConv2d(features, features),\n            nn.BatchNorm2d(features),\n            nn.ReLU(True),\n            TiledConv2d(features, features)\n        )\n\n    def forward(self, x):\n        return self.main(x) + x\n\n\ndef best_guess_block(input_channels, channels):\n    return nn.Sequential(\n        nn.Conv2d(input_channels, channels, 1, bias=False),\n        TiledResBlock(channels),\n        TiledResBlock(channels),\n        nn.BatchNorm2d(channels),\n        nn.ReLU(True),\n    )\n\n\ndef create_net_input_array(state: np.array, predicted_mask: np.array, predictions: np.array):\n    input_dead = (1 - state).astype(np.float)\n    input_alive = state.astype(np.float)\n    input_unpredicted = (1 - predicted_mask).astype(np.float)\n    input_predicted_dead = ((1 - predictions) * predicted_mask).astype(np.float)\n    input_predicted_alive = (predictions * predicted_mask).astype(np.float)\n    sample_input = np.stack([\n        input_dead,\n        input_alive,\n        input_unpredicted,\n        input_predicted_dead,\n        input_predicted_alive],\n        axis=2\n    )\n\n    return sample_input.transpose((2, 0, 1)).astype(np.float)\n\n\nclass BestGuessModule(nn.Module):\n    def __init__(self, channels=16 * 8):\n        super().__init__()\n\n        self.main = nn.Sequential(\n            best_guess_block(5, channels),\n            best_guess_block(channels, channels * 2),\n            best_guess_block(channels * 2, channels * 4),\n            nn.Conv2d(channels * 4, 2, 1)\n        )\n\n    def forward(self, x):\n        return self.main(x)\n\n    def get_probabilities(self, x):\n        return torch.softmax(self.main(x), 1)\n\n    def get_best_guess(self, x: torch.tensor, mask: np.array):\n        probabilities = self.get_probabilities(x)\n        masked_probabilities = np.array(probabilities.tolist()) * (1 - mask)\n        guess = np.unravel_index(masked_probabilities.argmax(), masked_probabilities.shape)\n        return {\n            \"coord_yx\": np.array([guess[2], guess[3]]),\n            \"alive\": guess[1]\n        }\n\n    def get_best_guesses(self, x: torch.tensor, mask: np.array, num_guesses=2):\n        probabilities = self.get_probabilities(x)\n        masked_probabilities = np.array(probabilities.tolist()) * (1 - mask)\n        guess = np.unravel_index(masked_probabilities.argmax(), masked_probabilities.shape)\n        return {\n            \"coord_yx\": np.array([guess[2], guess[3]]),\n            \"alive\": guess[1]\n        }\n\n    def get_best_by_threshold(self, x: torch.tensor, mask: np.array, threshold: float):\n        probabilities = self.get_probabilities(x)\n        masked_probabilities = np.array(probabilities.tolist()) * (1 - mask)\n        results = np.where(masked_probabilities >= threshold)\n        return {\n            \"coord_yx\": np.array([results[2], results[3]]),\n            \"alive\": results[1]\n        }\n\n    def get_tendencies_img(self, x):\n        return np.array(self.get_probabilities(x).tolist()).transpose((0, 2, 3, 1))[0, :, :, 1]\n\n\n    def solve(self,\n              state: np.array,\n              device: torch.device,\n              ground_truth=None):\n\n        predicted_mask = np.zeros(state.shape)\n        predictions = np.zeros(state.shape)\n        total_runs = state.shape[0] * state.shape[1]\n        \n        fig = plt.figure(figsize=(8, 6))\n\n        def animate(i):\n            sample_input = create_net_input_array(state, predicted_mask, predictions)\n            batch_input = torch.from_numpy(np.expand_dims(sample_input, 0)).float().to(device)\n            guess = self.get_best_guess(batch_input, predicted_mask)\n            predicted_mask[guess['coord_yx'][0], guess['coord_yx'][1]] = 1\n            predictions[guess['coord_yx'][0], guess['coord_yx'][1]] = guess['alive']\n\n\n            # input\n            plt.clf()\n            sub = fig.add_subplot(2, 3, 1)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"start (ground truth)\")\n            if ground_truth is not None:\n                sub.imshow(ground_truth.astype(np.float))\n            sub = fig.add_subplot(2, 3, 4)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"end state\")\n            sub.imshow(state.astype(np.float))\n\n            # net\n            sub = fig.add_subplot(2, 3, 3)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"net probabilities\")\n            prob = self.get_tendencies_img(batch_input)\n            overlay = np.ones((state.shape[0], state.shape[1], 4), dtype=np.float)\n            overlay[:, :, 3] = predicted_mask\n            # prob[prob < 0.5] *= -1\n            # prob[prob < 0.5] += 1.0\n            # prob *= (1 - prev_predicted_mask)\n            sub.imshow(prob, vmin=0.0, vmax=1.0)\n            sub.imshow(overlay, vmin=0.0, vmax=1.0)\n\n            # outcome\n            sub = fig.add_subplot(2, 3, 2)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"net prediction\")\n            overlay = np.ones((state.shape[0], state.shape[1], 4), dtype=np.float)\n            overlay[:, :, 3] = (1.0 - predicted_mask) * 0.66\n            sub.imshow(predictions.astype(np.float))\n            sub.imshow(overlay, vmin=0.0, vmax=1.0)\n            sub = fig.add_subplot(2, 3, 5)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"forwarded prediction\".format(1))\n            outc = predictions\n            for d in range(1):\n                outc = state_step(outc)\n            overlay = np.zeros((state.shape[0], state.shape[1], 4), dtype=np.float)\n            overlay[outc - state != 0] = [1, 0.2, 0.2, 1]\n            sub.imshow(outc.astype(np.float))\n            sub.imshow(overlay, vmin=0.0, vmax=1.0)\n\n        anim = animation.FuncAnimation(fig, animate, frames=total_runs, interval=50)\n        plt.clf()\n        return predictions, anim\n\n    \ndevice = torch.device('cuda')\nbestGuess = BestGuessModule()\nbestGuess.load_state_dict(torch.load('../input/gol-pretrained-model/epoch_040.pt'))\nbestGuess.to(device)\nbestGuess.eval()\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# feel free to enter your own end state here\nend_state = np.array([\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0],\n       [1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0],\n       [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0],\n       [0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0],\n       [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1]])\n\n# this may take a while to calculate with the animation...\npred_start_state, anim = bestGuess.solve(end_state, device)\nHTML(anim.to_jshtml())","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"It can be seen how it starts with simple cells and solves the more difficult ones at the end.\nSome things I noticed:\n* It reconstrcuts the glider (middle left) correctly\n* Because it is trained to learn the distribution of given GoL states, it generates more \"game of life looking\" patterns.\n* It starts solving (is most certain) around simple patterns. Before filling empty space. This could be, because in the empty areas the probability that some structure died in the start state is higher than arround a well known pattern. \n\nI also tried training the network to predict delta>1 directy but it seemed to require a huge ammount of time and way more parameters. This model has alread 12 mio. parameters, which was enough for my gtx 980 ti.\n\nSadly, the error increases a lot when repeating the guessing algorithm for delta>1. While at delta=1 its average loss is about 0.002. With delta=2 it increased ten times to 0.017. For delta=3 already 0.11."},{"metadata":{},"cell_type":"markdown","source":"## Optimizer\n\n\nThe optimizer improves the input state by randomly selecting 3 by 3 regions and testinal all possible inputs for this region. It then takes only the solutions which generate the lowest score. If there are multiple solutions with the same score, it will select randomly one of them. This process gets repeated multiple times.\n\nBelow is an example with delta=3 where first BestGuess is applied 3 times and then the optimizer at the end."},{"metadata":{"_kg_hide-input":true,"_kg_hide-output":true,"trusted":true},"cell_type":"code","source":"def binary_clamp(x: torch.tensor):\n    return torch.clamp(x, 0, 1)\n\n\ndef conway_layer(x: torch.tensor):\n    surround_sum = torch.roll(x, 1, 2) + torch.roll(x, -1, 2) + torch.roll(x, 1, 3) + torch.roll(x, -1, 3) +\\\n        torch.roll(x, (-1, -1), (2, 3)) + torch.roll(x, (1, -1), (2, 3)) + torch.roll(x, (-1, 1), (2, 3)) + torch.roll(x, (1, 1), (2, 3))\n    return binary_clamp(surround_sum + x - 2) - binary_clamp(surround_sum - 3)\n\n\nclass TilePad2d(nn.Module):\n    def __init__(self, left, right, top, bottom):\n        super().__init__()\n        self.left = left\n        self.right = right\n        self.top = top\n        self.bottom = bottom\n\n    def forward(self, x):\n        return F.pad(x, [self.left, self.right, self.top, self.bottom], mode='circular')\n\n\nclass BestChangeLayer(nn.Module):\n    def __init__(self, delta=1, window=(3, 3), device=torch.device('cpu')):\n        super().__init__()\n        self.device = device\n        self.window = window\n        self.influence_window = (window[0] + 4 * delta, window[1] + 4 * delta)\n        self.delta = delta\n        self.num_bins = window[0] * window[1]\n        self.num_possible_window_inputs = 2 ** self.num_bins\n        self.possible_inputs = np.zeros((self.num_possible_window_inputs, self.num_bins))\n\n        # compute all possible\n        for i in range(self.num_possible_window_inputs):\n            self.possible_inputs[i] = np.array(list(np.binary_repr(i, self.num_bins)), dtype=np.float)\n        self.possible_inputs = self.possible_inputs.reshape((1, self.num_possible_window_inputs, window[0], window[1]))\n\n        self.possible_inputs_mask = np.zeros((1, self.num_possible_window_inputs, self.influence_window[0], self.influence_window[1]), dtype=np.float)\n        self.possible_inputs_mask[:, :, 2*delta:-2*delta, 2*delta:-2*delta] = 1\n        self.possible_inputs_window = np.zeros((1, self.num_possible_window_inputs, self.influence_window[0], self.influence_window[1]), dtype=np.float)\n        self.possible_inputs_window[:, :, 2 * delta:-2 * delta, 2 * delta:-2 * delta] = self.possible_inputs\n\n        self.pi = torch.from_numpy(self.possible_inputs).float().to(device)\n        self.pi_window = torch.from_numpy(self.possible_inputs_window).float().to(device)\n        self.pi_window_mask = torch.from_numpy(self.possible_inputs_mask).float().to(device)\n        self.pi_window_inv_mask = -self.pi_window_mask + 1\n\n        self.replication_input_layer = TilePad2d(\n            delta * 2,\n            delta * 2 + window[1] - 1,\n            delta * 2,\n            delta * 2 + window[0] - 1)\n\n        self.replication_target_layer = TilePad2d(\n            delta,\n            delta + window[1] - 1,\n            delta,\n            delta + window[0] - 1)\n\n        self.unpool = nn.MaxUnpool1d(self.num_possible_window_inputs)\n\n    def forward(self, x: torch.Tensor, target: torch.Tensor):\n\n        random_x = np.random.randint(0, x.size()[3])\n        random_y = np.random.randint(0, x.size()[2])\n\n        influence_window = self.replication_input_layer(x)[:, :, random_y:(random_y + self.influence_window[0]), random_x:(random_x + self.influence_window[1])]\n        target_window = self.replication_target_layer(target)[:, :, random_y:(random_y + self.window[0] + 2 * self.delta), random_x:(random_x + self.window[1] + 2 * self.delta)]\n        process_window = influence_window.repeat(1, self.num_possible_window_inputs, 1, 1)\n        process_window = process_window * self.pi_window_inv_mask + self.pi_window\n        end = process_window\n        for d in range(self.delta):\n            end = conway_layer(end)[:, :, 1:-1, 1:-1]\n\n        errors = torch.sum(torch.abs(end - target_window.repeat(1, self.num_possible_window_inputs, 1, 1)), (2, 3))\n        seeded_errors = errors + torch.rand(errors.size(), device=self.device) * 0.5\n\n        indices = torch.argmin(seeded_errors, 1)\n        best_mask = self.unpool(torch.ones((seeded_errors.size()[0], 1, 1), device=self.device), indices.reshape(-1, 1, 1)).reshape(-1, self.num_possible_window_inputs, 1, 1)\n        best_inputs = torch.sum(self.pi * best_mask, 1).reshape((-1, 1, self.window[0], self.window[1]))\n\n        out = torch.roll(x, shifts=(-random_y, -random_x), dims=(2, 3))\n        out[:, :, :self.window[0], :self.window[1]] = best_inputs\n        out = torch.roll(out, shifts=(random_y, random_x), dims=(2, 3))\n\n        return out\n\n    def solve(self, initial_state: np.array, target: np.array, device: torch.device, num_steps=1000):\n        self.inputs_batch = torch.from_numpy(initial_state).reshape(1, 1, initial_state.shape[0], initial_state.shape[1]).float().to(device)\n        self.targets_batch = torch.from_numpy(target).reshape(1, 1, target.shape[0], target.shape[1]).float().to(device)\n        \n        fig = plt.figure(figsize=(8, 6))\n        \n        def animate(i):\n            self.inputs_batch = self(self.inputs_batch, self.targets_batch)\n            state = np.clip(np.rint(np.array(self.inputs_batch.tolist())).astype(np.int), 0, 1).reshape(initial_state.shape)\n            \n            # input\n            plt.clf()\n            sub = fig.add_subplot(2, 2, 3)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"target state\")\n            sub.imshow(target.astype(np.float))\n\n            # outcome\n            sub = fig.add_subplot(2, 2, 2)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"optimized pred.\")\n            sub.imshow(state.astype(np.float))\n            sub = fig.add_subplot(2, 2, 4)\n            plt.xticks([])\n            plt.yticks([])\n            sub.set_title(\"forwarded prediction\".format(1))\n            outc = state\n            for d in range(self.delta):\n                outc = state_step(outc)\n            overlay = np.zeros((state.shape[0], state.shape[1], 4), dtype=np.float)\n            overlay[outc - target != 0] = [1, 0.2, 0.2, 1]\n            sub.imshow(outc.astype(np.float))\n            sub.imshow(overlay, vmin=0.0, vmax=1.0)\n            \n        anim = animation.FuncAnimation(fig, animate, frames=num_steps, interval=10)\n        plt.clf()\n        state = np.clip(np.rint(np.array(self.inputs_batch.tolist())).astype(np.int), 0, 1).reshape(initial_state.shape)\n        return state, anim\n    \noptimizer = BestChangeLayer(delta=3, device=device)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# feel freee to enter your own end state here\nend_state = np.array([\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0],\n       [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0],\n       [1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0],\n       [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0],\n       [0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0],\n       [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1]])\ncurrent_prediction = end_state.copy()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Predict delta=1:"},{"metadata":{"trusted":true},"cell_type":"code","source":"current_prediction, anim = bestGuess.solve(current_prediction, device)\nHTML(anim.to_jshtml())","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Predict from delta=1 to delta=2:"},{"metadata":{"trusted":true},"cell_type":"code","source":"current_prediction, anim = bestGuess.solve(current_prediction, device)\nHTML(anim.to_jshtml())","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Predict from delta=2 to delta=3:"},{"metadata":{"trusted":true},"cell_type":"code","source":"current_prediction, anim = bestGuess.solve(current_prediction, device)\nHTML(anim.to_jshtml())","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Optimize the prediction with 2000 iterations:"},{"metadata":{"trusted":true},"cell_type":"code","source":"optimized_start_state, anim = optimizer.solve(current_prediction, end_state, device, 2000)\nHTML(anim.to_jshtml())","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"Thanks for reading and have a nice day!"}],"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}