{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":36363,"databundleVersionId":4050810,"sourceType":"competition"},{"sourceId":6392848,"sourceType":"datasetVersion","datasetId":3607309},{"sourceId":6403207,"sourceType":"datasetVersion","datasetId":3692048},{"sourceId":6665364,"sourceType":"datasetVersion","datasetId":3846147},{"sourceId":6874921,"sourceType":"datasetVersion","datasetId":3950522},{"sourceId":7219769,"sourceType":"datasetVersion","datasetId":4178645},{"sourceId":7220928,"sourceType":"datasetVersion","datasetId":4179450},{"sourceId":7222318,"sourceType":"datasetVersion","datasetId":4180446}],"dockerImageVersionId":30559,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"!pip install --upgrade pip --quiet\n!pip install nibabel --quiet\n!pip install torch torchvision torchaudio --quiet","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:49:05.156635Z","iopub.execute_input":"2023-12-14T10:49:05.157332Z","iopub.status.idle":"2023-12-14T10:49:57.555915Z","shell.execute_reply.started":"2023-12-14T10:49:05.157302Z","shell.execute_reply":"2023-12-14T10:49:57.554546Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os\nimport pandas as pd\nimport numpy as np\nimport nibabel as nib\nimport torch\nimport torch.nn as nn\nimport torch.optim as optim\nfrom torch.utils.data import Dataset, DataLoader\nfrom torchvision import transforms\nfrom sklearn.metrics import (accuracy_score, precision_recall_fscore_support, roc_auc_score)\nfrom sklearn.model_selection import train_test_split\nfrom torchvision import models\nfrom scipy.ndimage import zoom\nimport torch.nn.functional as F\nfrom PIL import Image\n\n# Constants and configuration settings\nsegmentation_dir = '/kaggle/input/rsna-2022-cervical-spine-fracture-detection/segmentations'\ncsv_file = '/kaggle/input/file-mask-path/train_file_mask_path.csv'\nbatch_size = 4\nnum_workers = 4\nnum_classes = 7\ndesired_shape = (128, 128, 128)\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")","metadata":{"_kg_hide-output":true,"execution":{"iopub.status.busy":"2023-12-14T10:50:53.360554Z","iopub.execute_input":"2023-12-14T10:50:53.360968Z","iopub.status.idle":"2023-12-14T10:50:53.369023Z","shell.execute_reply.started":"2023-12-14T10:50:53.360939Z","shell.execute_reply":"2023-12-14T10:50:53.367817Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Specify the path to your CSV file containing data\ncsv_file_path = '/kaggle/input/file-mask-path/train_file_mask_path.csv'\n\n# Load the CSV data using pandas\ndata_frame = pd.read_csv(csv_file_path)\n\n# Extract relevant information from the data\nimage_paths = data_frame['file_path'].values\nmask_paths = data_frame['mask_path'].values\nlabels = data_frame[['C1', 'C2', 'C3', 'C4', 'C5', 'C6', 'C7', 'patient_overall']].values\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:53.678654Z","iopub.execute_input":"2023-12-14T10:50:53.679004Z","iopub.status.idle":"2023-12-14T10:50:53.696094Z","shell.execute_reply.started":"2023-12-14T10:50:53.678975Z","shell.execute_reply":"2023-12-14T10:50:53.69505Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import nibabel as nib\nimport pandas as pd\nimport torch\nfrom torch.utils.data import Dataset\n\nclass CustomDataset(Dataset):\n    def __init__(self, image_paths, mask_paths, labels, transform=None):\n        self.image_paths = image_paths\n        self.mask_paths = mask_paths\n        self.labels = labels\n        self.transform = transform\n\n    def __len__(self):\n        return len(self.image_paths)\n\n    def __getitem__(self, idx):\n        image_path = self.image_paths[idx]\n        mask_path = self.mask_paths[idx]\n\n        image = nib.load(image_path).get_fdata()\n        segmentation_mask = None\n\n        if pd.notna(mask_path):\n            segmentation_mask = nib.load(mask_path)\n            segmentation_mask_data = segmentation_mask.get_fdata()\n            resized_data = resize_nifti(segmentation_mask_data, desired_shape)\n            segmentation_mask_data_affine = segmentation_mask.affine\n            resized_affine = segmentation_mask_data_affine\n            segmentation_mask = nib.Nifti1Image(resized_data, affine=resized_affine).get_fdata()\n\n        if self.transform:\n            image = self.transform(image)\n\n        if segmentation_mask is not None and self.transform:\n            \n            segmentation_mask = self.transform(segmentation_mask)\n        else:\n            segmentation_mask = torch.zeros_like(image)\n\n        label = torch.tensor(self.labels[idx], dtype=torch.float32)\n        \n        return image, segmentation_mask, label\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:54.046416Z","iopub.execute_input":"2023-12-14T10:50:54.046821Z","iopub.status.idle":"2023-12-14T10:50:54.057889Z","shell.execute_reply.started":"2023-12-14T10:50:54.046774Z","shell.execute_reply":"2023-12-14T10:50:54.056693Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Create an instance of the custom dataset\ndataset = CustomDataset(image_paths, mask_paths, labels, transform=None)\n\n# Define the index of the sample you want to access\nsample_index = 0  # You can change this to any index you're interested in\n\n# Access the image and mask paths for the specific sample\nimage_path = dataset.image_paths[sample_index]\nmask_path = dataset.mask_paths[sample_index]","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:54.568378Z","iopub.execute_input":"2023-12-14T10:50:54.56876Z","iopub.status.idle":"2023-12-14T10:50:54.574708Z","shell.execute_reply.started":"2023-12-14T10:50:54.568732Z","shell.execute_reply":"2023-12-14T10:50:54.573704Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import torch\nimport torch.nn as nn\nimport torchvision.models as models\nfrom scipy.ndimage import zoom\n# Function to resize NIfTI data\ndef resize_nifti(nifti_data, target_shape):\n    factors = (target_shape[0] / nifti_data.shape[0],\n               target_shape[1] / nifti_data.shape[1],\n               target_shape[2] / nifti_data.shape[2])\n    resized_data = zoom(nifti_data, factors, order=3)  # Cubic interpolation (higher quality)\n    return resized_data\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:55.081931Z","iopub.execute_input":"2023-12-14T10:50:55.082299Z","iopub.status.idle":"2023-12-14T10:50:55.088746Z","shell.execute_reply.started":"2023-12-14T10:50:55.082243Z","shell.execute_reply":"2023-12-14T10:50:55.087746Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torchvision import models\nfrom torchvision.models.video import mvit_v2_s, MViT_V2_S_Weights\n\nclass MultiLabel3DAttentionModel(nn.Module):\n    def __init__(self, num_classes, num_classes_segmentation):\n        super(MultiLabel3DAttentionModel, self).__init__()\n\n        # Load a pre-trained ResNet3D backbone\n        self.backbone = models.video.mvit_v2_s(pretrained=True)\n        \n        # Modify the stem to accept the correct input channels (128)\n#         self.backbone.stem[0] = nn.Sequential(\n#             nn.Conv3d(1, 64, kernel_size=(3, 7, 7), stride=(1, 2, 2), padding=(1, 3, 3)),\n#             nn.BatchNorm3d(64),\n#             nn.ReLU(inplace=True))\n#         self.backbone.conv_proj = nn.Sequential(\n#             nn.Conv3d(1, 96, kernel_size=(3, 7, 7), stride=(2, 4, 4), padding=(1, 3, 3)),\n    #    )\n\n\n        # Attention block\n        self.attention = nn.Sequential(\n            nn.Conv3d(1, 128, kernel_size=1),\n            nn.ReLU(inplace=True),\n            nn.Conv3d(128, 1, kernel_size=1),\n            nn.Sigmoid()\n        )\n        \n        # Classification head\n        self.classification_head = nn.Sequential(\n            nn.AdaptiveAvgPool3d(1),\n            nn.Flatten(),\n            nn.Linear(1, 64),\n            nn.ReLU(inplace=True),\n            nn.Linear(64, num_classes + 1),\n            nn.Sigmoid()\n        )\n        \n        # Segmentation head\n        self.segmentation_head = nn.Sequential(\n            nn.Conv3d(1, 128, kernel_size=1),\n            nn.ReLU(inplace=True),\n            nn.Conv3d(128, num_classes_segmentation, kernel_size=1),\n            nn.Sigmoid()\n        )\n        \n    def forward(self, x, segmentation_mask):\n        # Feature extraction with the backbone\n#         print(x.shape, \"........................\")\n        transforms = MViT_V2_S_Weights.KINETICS400_V1.transforms()\n        #x = transforms(torch.rand(4, 16, 3, 128, 128))\n        x = transforms(x)\n#         print(x.shape,\"...........after transformation...........\")\n        features = self.backbone(x)\n        \n#         print(\"features shape = \",features.shape)\n\n        # Apply attention to features\n        features = features.view(features.size(0), 1, 1, 1, features.size(1))\n        attention_weights = self.attention(features)\n        attended_features = features * attention_weights\n\n        # Classification branch\n        classification_output = self.classification_head(attended_features)\n        \n        # Initialize segmentation_output as None\n        segmentation_output = None\n\n        # Check if segmentation_mask is None\n        if segmentation_mask is not None:\n            \n            # Segmentation branch\n            segmentation_output = self.segmentation_head(attended_features)\n            segmentation_output = F.interpolate(segmentation_output, size=segmentation_mask.shape[2:], mode='trilinear')\n            segmentation_output = segmentation_output * segmentation_mask\n            \n        else:\n            # No segmentation branch in the test phase\n            segmentation_mask = torch.zeros_like(x)\n\n        return classification_output, segmentation_output\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:55.635974Z","iopub.execute_input":"2023-12-14T10:50:55.636419Z","iopub.status.idle":"2023-12-14T10:50:55.651342Z","shell.execute_reply.started":"2023-12-14T10:50:55.636383Z","shell.execute_reply":"2023-12-14T10:50:55.650056Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"transform2 = transforms.Compose([\n    transforms.ToTensor(),  # Convert to tensor\n    # Add more transformations if necessary\n])","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:56.16811Z","iopub.execute_input":"2023-12-14T10:50:56.169184Z","iopub.status.idle":"2023-12-14T10:50:56.173824Z","shell.execute_reply.started":"2023-12-14T10:50:56.169146Z","shell.execute_reply":"2023-12-14T10:50:56.172778Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import pandas as pd\n\n# Load the CSV file\ndata = pd.read_csv(csv_file)\ndata_length = len(data)\nprint(\"Length of DataFrame:\", data_length)\n\n# Remove leading and trailing whitespaces from column names\ndata.columns = data.columns.str.strip()\n\n# Split the data into training, validation, and test sets\n# train_data, temp_data = train_test_split(data, test_size=0.3, random_state=42)\n# val_data, test_data = train_test_split(temp_data, test_size=0.5, random_state=42)\n# train_data.to_csv('exp_train.csv', index=False)\n# test_data.to_csv('exp_test.csv', index=False)\n# val_data.to_csv('exp_val.csv', index=False)\ntrain_data = pd.read_csv('/kaggle/input/data-divided/exp_train.csv')\ntest_data = pd.read_csv('/kaggle/input/data-divided/exp_test.csv')\nval_data = pd.read_csv('/kaggle/input/data-divided/exp_val.csv')\n\n# Limit the number of samples for testing purposes\n# train_data = train_data[:12]\n# val_data = val_data[:12]\n# test_data = test_data[:100]\n\n# Extract file paths and labels from the data\ntrain_paths = train_data['file_path'].values\ntrain_mask_paths = train_data['mask_path'].values\ntrain_labels = train_data[['C1', 'C2', 'C3', 'C4', 'C5', 'C6', 'C7', 'patient_overall']].values\n\nval_paths = val_data['file_path'].values\nval_mask_paths = val_data['mask_path'].values\nval_labels = val_data[['C1', 'C2', 'C3', 'C4', 'C5', 'C6', 'C7', 'patient_overall']].values\n\ntest_paths = test_data['file_path'].values\ntest_mask_paths = test_data['mask_path'].values\ntest_labels = test_data[['C1', 'C2', 'C3', 'C4', 'C5', 'C6', 'C7', 'patient_overall']].values\n\ntrain_length = len(train_data)\nprint(\"Length of train_data:\", train_length)\nval_length = len(val_data)\nprint(\"Length of val_data:\", val_length)\ntest_length = len(test_data)\nprint(\"Length of test_data:\", test_length)\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:56.645889Z","iopub.execute_input":"2023-12-14T10:50:56.646234Z","iopub.status.idle":"2023-12-14T10:50:56.681274Z","shell.execute_reply.started":"2023-12-14T10:50:56.646207Z","shell.execute_reply":"2023-12-14T10:50:56.680335Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Instantiate the datasets\ntrain_dataset = CustomDataset(train_paths, train_mask_paths, train_labels, transform=transform2)\nval_dataset = CustomDataset(val_paths, val_mask_paths, val_labels, transform=transform2)\ntest_dataset = CustomDataset(test_paths, test_mask_paths, test_labels, transform=transform2)","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:57.839856Z","iopub.execute_input":"2023-12-14T10:50:57.840221Z","iopub.status.idle":"2023-12-14T10:50:57.846124Z","shell.execute_reply.started":"2023-12-14T10:50:57.840192Z","shell.execute_reply":"2023-12-14T10:50:57.844823Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Instantiate the data loaders\ntrain_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True, num_workers=num_workers, drop_last=True)\nval_loader = DataLoader(val_dataset, batch_size=batch_size, shuffle=False, num_workers=num_workers)\ntest_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False, num_workers=num_workers)","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:58.542668Z","iopub.execute_input":"2023-12-14T10:50:58.543635Z","iopub.status.idle":"2023-12-14T10:50:58.549544Z","shell.execute_reply.started":"2023-12-14T10:50:58.543597Z","shell.execute_reply":"2023-12-14T10:50:58.548422Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"\n# Instantiate the model with the appropriate number of classes for both classification and segmentation\nnum_classes_classification = 7  # Number of classes for classification\nnum_classes_segmentation = 1    # Number of classes for segmentation (change this according to your task)\nmodel = MultiLabel3DAttentionModel(num_classes_classification, num_classes_segmentation)\n\n# print(model)\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:50:59.174703Z","iopub.execute_input":"2023-12-14T10:50:59.17542Z","iopub.status.idle":"2023-12-14T10:51:00.663338Z","shell.execute_reply.started":"2023-12-14T10:50:59.175386Z","shell.execute_reply":"2023-12-14T10:51:00.662305Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from tqdm import tqdm\nimport pickle \nimport random","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:51:00.666604Z","iopub.execute_input":"2023-12-14T10:51:00.666974Z","iopub.status.idle":"2023-12-14T10:51:00.671407Z","shell.execute_reply.started":"2023-12-14T10:51:00.666941Z","shell.execute_reply":"2023-12-14T10:51:00.670554Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import torch\n\ndef weighted_cross_entropy(predicted, label, weights):\n    num_samples = predicted.shape[0]\n    \n    # Calculate element-wise losses\n    losses = weights * (-label * torch.log(predicted) - (1 - label) * torch.log(1 - predicted))\n    \n    # Sum the total_loss for all inputs and divide by the sum of weights\n    total_loss = torch.sum(losses, dim=1) / torch.sum(weights, dim=1)\n    \n    # Sum the total_loss for all inputs and divide by the number of samples\n    final_loss = torch.sum(total_loss) / num_samples\n    \n    return final_loss","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:51:02.827495Z","iopub.execute_input":"2023-12-14T10:51:02.827858Z","iopub.status.idle":"2023-12-14T10:51:02.834575Z","shell.execute_reply.started":"2023-12-14T10:51:02.82783Z","shell.execute_reply":"2023-12-14T10:51:02.833534Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import torch\n\ndef focal_loss(predicted, label, weights, gamma=2, epsilon=1e-7):\n    # Calculate the probability of the positive class (pt) for each sample and class\n    pt = torch.where(label == 1, predicted, 1 - predicted)\n    \n    # Calculate the loss components for each class\n    loss = -weights * ((1 - pt) ** gamma) * torch.log(pt + epsilon)\n\n    # Sum the loss components for each class\n    final_loss = torch.sum(loss, dim=1) / torch.sum(weights, dim=1)\n    \n    return final_loss.mean()\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:51:03.757168Z","iopub.execute_input":"2023-12-14T10:51:03.757565Z","iopub.status.idle":"2023-12-14T10:51:03.764207Z","shell.execute_reply.started":"2023-12-14T10:51:03.757535Z","shell.execute_reply":"2023-12-14T10:51:03.763088Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import torch\nimport torch.nn as nn\n\ndef binary_cross_entropy_loss(segmentation_outputs, newbatch_segmentation_masks):\n    # Ensure that both batch_segmentation_masks and segmentation_outputs are tensors of type torch.float32\n    newbatch_segmentation_masks = newbatch_segmentation_masks.to(torch.float32)\n    segmentation_outputs = segmentation_outputs.to(torch.float32)\n\n    # Check shapes\n    assert newbatch_segmentation_masks.shape == segmentation_outputs.shape, \"Input shapes must match\"\n\n    # Check for NaN or Infinite values\n    assert not torch.isnan(newbatch_segmentation_masks).any() and not torch.isinf(newbatch_segmentation_masks).any(), \"Input contains NaN or Infinite values\"\n    assert not torch.isnan(segmentation_outputs).any() and not torch.isinf(segmentation_outputs).any(), \"Input contains NaN or Infinite values\"\n\n    # Clip values in newbatch_segmentation_masks to [0, 1] range\n    newbatch_segmentation_masks = torch.clamp(newbatch_segmentation_masks, 0, 1)\n\n    # Create an instance of BCELoss\n    criterion = nn.BCELoss()\n\n    # Compute the BCE loss\n    loss = criterion(segmentation_outputs, newbatch_segmentation_masks)\n\n    return loss","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:51:04.408216Z","iopub.execute_input":"2023-12-14T10:51:04.409218Z","iopub.status.idle":"2023-12-14T10:51:04.416902Z","shell.execute_reply.started":"2023-12-14T10:51:04.409182Z","shell.execute_reply":"2023-12-14T10:51:04.415801Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import torch\nimport torch.nn as nn\nfrom torch.optim import AdamW  # Import AdamW\nfrom sklearn.metrics import accuracy_score\n\n# Function to calculate weights based on class labels\ndef weight_calculate(class_labels):\n    # Calculate the weight for each class label\n    weights = []\n    total_weight = 0\n    c = 1\n    for label in class_labels:\n        if c % 8 == 0:\n            if label == 0:\n                weights.append(7)\n                total_weight += 7\n            else:\n                weights.append(14)\n                total_weight += 14\n        else:\n            if label == 0:\n                weights.append(1)\n                total_weight += 1\n            else:\n                weights.append(2)\n                total_weight += 2\n        c += 1\n    # Calculate the overall weight based on the presence of any class label being 1\n    return torch.tensor(weights, dtype=torch.float32)\n\ndef rsnamodel(num_epochs=10, initial_epoch=0, ranOnce=False, model_path='sample.pth', history_path='sample_history.pkl'):\n    criterion = nn.BCEWithLogitsLoss()  # Binary Cross-Entropy loss\n\n    if ranOnce:\n        # Load the existing model if ranOnce is True\n        model = MultiLabel3DAttentionModel(num_classes_classification, num_classes_segmentation)\n        model.load_state_dict(torch.load(model_path))\n        model = model.to(device)\n\n        # Load the existing metrics history for plotting\n        with open(history_path, 'rb') as f:\n            metrics_history = pickle.load(f)\n\n        val_loss_history = metrics_history['val_loss_history']\n        val_acc_history = metrics_history['val_acc_history']\n        train_loss_history = metrics_history['train_loss_history']\n        train_acc_history = metrics_history['train_acc_history']\n    else:\n        # Create a new model if ranOnce is False\n        # Instantiate the model with the appropriate number of classes for both classification and segmentation\n        model = MultiLabel3DAttentionModel(num_classes_classification, num_classes_segmentation)\n        model.to(device)\n\n        # Initialize empty lists for metrics history\n        val_loss_history = []\n        val_acc_history = []\n        train_loss_history = []\n        train_acc_history = []\n\n    optimizer = optim.Adam(model.parameters(), lr=0.01)\n\n    for epoch in range(num_epochs):\n        model.train()\n        running_loss = 0.0\n        correct_train = 0\n        total_train = 0\n\n        for batch_images, batch_segmentation_masks, batch_labels in tqdm(train_loader, desc=f\"Epoch {epoch+1+initial_epoch}/{num_epochs+initial_epoch} Training:\"):\n            optimizer.zero_grad()\n\n            # Move data to the GPU if available\n            batch_images = batch_images.to(torch.float32).to(device)\n            batch_segmentation_masks = batch_segmentation_masks.to(torch.float32).to(device)\n            batch_labels = batch_labels.to(torch.float32).to(device)\n\n            # Assuming batch_images has shape (batch_size, num_frames, num_channels, height, width)\n            batch_images = batch_images.unsqueeze(1)  # Add a singleton dimension for channels\n            batch_segmentation_masks = batch_segmentation_masks.unsqueeze(1)\n#             print(batch_images.shape)\n#             # Assuming batch_images has shape (4, 1, 128, 128, 128)\n#             batch_images = batch_images.squeeze(1)  # Remove the singleton dimension for channels\n            # Number of frames to add\n            num_frames_to_add = 16\n            \n            tensor_list = [batch_images] * num_frames_to_add\n            # Concatenate along the frames dimension (dimension 1)\n            nbatch_images = torch.cat(tensor_list, dim=1)\n            nbatch_segmentation_masks = torch.cat(tensor_list, dim=1)\n            \n            newbatch_images = nbatch_images[:, :, :3, ...].contiguous()\n            newbatch_segmentation_masks = nbatch_segmentation_masks[:, :, :3, ...].contiguous()\n            \n            # Check the current shape\n#             print(\"Current shape:\", newbatch_images.shape)\n#             print(\"Current shape:\", newbatch_segmentation_masks.shape)\n#             # Reshape to the desired shape (2, 16, 3, 224, 224)\n#             batch_images = batch_images.view(2, 16, 3, 128, 128)  # Assuming the original spatial dimensions are 128x128\n\n#             # Check the new shape\n#             print(\"New shape:\", batch_images.shape)            # Forward pass\n            classification_outputs, segmentation_outputs = model(newbatch_images, newbatch_segmentation_masks)\n\n            # Apply sigmoid activation to the classification outputs\n            classification_outputs = torch.sigmoid(classification_outputs)\n\n            # Calculate weights for each sample based on class labels\n            weights = torch.stack([weight_calculate(labels) for labels in batch_labels]).to(device)\n\n            # Calculate binary cross-entropy loss with weights\n            weighted_loss = focal_loss(classification_outputs, batch_labels, weights)\n\n            # Check if segmentation mask is available\n            if newbatch_segmentation_masks is not None and (newbatch_segmentation_masks != 0).any():\n                # Ensure that both input and target tensors are of type torch.float32\n                newbatch_segmentation_masks = newbatch_segmentation_masks.to(torch.float32)\n\n                # Apply sigmoid activation to segmentation_outputs\n                segmentation_outputs = torch.sigmoid(segmentation_outputs)\n                segmentation_outputs = segmentation_outputs.to(torch.float32)\n\n                # Calculate segmentation loss\n                segmentation_loss = binary_cross_entropy_loss(segmentation_outputs, newbatch_segmentation_masks)\n\n                # Add segmentation loss to the weighted loss\n                weighted_loss += segmentation_loss\n\n            running_loss += weighted_loss.item()\n\n            # Calculate accuracy for each class separately\n            accuracies = []\n            for class_index in range(num_classes_classification+1):\n                class_labels = batch_labels[:, class_index]\n                class_outputs = classification_outputs[:, class_index]\n\n                # Calculate binary predictions based on a threshold (e.g., 0.5)\n                predicted = (class_outputs > 0.5).float()\n\n                class_accuracy = accuracy_score(class_labels.cpu(), predicted.cpu())\n                accuracies.append(class_accuracy)\n\n            # Calculate overall accuracy\n            batch_accuracy = sum(accuracies) / (num_classes_classification+1)\n            correct_train += batch_accuracy\n            total_train += 1\n\n            # Backpropagation and optimization\n            weighted_loss.backward()\n            optimizer.step()\n\n        # Calculate and print average training accuracy and loss\n        avg_train_accuracy = correct_train / total_train\n        avg_train_loss = running_loss / (len(train_loader))\n        train_acc_history.append(avg_train_accuracy)\n        train_loss_history.append(avg_train_loss)\n\n        print(f\"Epoch [{epoch+initial_epoch+1}/{num_epochs+initial_epoch}]\")\n        print(f\"Train Accuracy: {avg_train_accuracy:.4f} | Train Loss: {avg_train_loss:.4f}\")\n\n        # Validation loop\n        model.eval()\n        total_val_loss = 0.0\n        correct_val = 0\n        total_val = 0\n\n        with torch.no_grad():\n            for batch_images, batch_segmentation_masks, batch_labels in tqdm(val_loader, desc=f\"Epoch {epoch+initial_epoch+1}/{num_epochs+initial_epoch} Validation:\"):\n                optimizer.zero_grad()\n\n                # Move data to the GPU if available\n                batch_images = batch_images.to(torch.float32).to(device)\n                batch_segmentation_masks = batch_segmentation_masks.to(torch.float32).to(device)\n                batch_labels = batch_labels.to(torch.float32).to(device)\n\n                # Assuming batch_images has shape (batch_size, num_frames, num_channels, height, width)\n                batch_images = batch_images.unsqueeze(1)  # Add a singleton dimension for channels\n                batch_segmentation_masks = batch_segmentation_masks.unsqueeze(1)\n                \n                num_frames_to_add = 16\n            \n                tensor_list = [batch_images] * num_frames_to_add\n                # Concatenate along the frames dimension (dimension 1)\n                nbatch_images = torch.cat(tensor_list, dim=1)\n                nbatch_segmentation_masks = torch.cat(tensor_list, dim=1)\n\n                newbatch_images = nbatch_images[:, :, :3, ...].contiguous()\n                newbatch_segmentation_masks = nbatch_segmentation_masks[:, :, :3, ...].contiguous()\n                # Check the current shape\n#                 print(\"Current shape:\", newbatch_images.shape)\n#                 print(\"Current shape:\", newbatch_segmentation_masks.shape)\n\n\n                # Forward pass\n                classification_outputs, segmentation_outputs = model(newbatch_images, newbatch_segmentation_masks)\n\n                # Apply sigmoid activation to the classification outputs\n                classification_outputs = torch.sigmoid(classification_outputs)\n\n                # Calculate weights for each sample based on class labels\n                weights = torch.stack([weight_calculate(labels) for labels in batch_labels]).to(device)\n\n                # Calculate binary cross-entropy loss with weights\n                weighted_loss = focal_loss(classification_outputs, batch_labels, weights)\n\n                # Check if segmentation mask is available\n                if newbatch_segmentation_masks is not None and (newbatch_segmentation_masks != 0).any():\n                    # Ensure that both input and target tensors are of type torch.float32\n                    newbatch_segmentation_masks = newbatch_segmentation_masks.to(torch.float32)\n\n                    # Apply sigmoid activation to segmentation_outputs\n                    segmentation_outputs = torch.sigmoid(segmentation_outputs)\n                    segmentation_outputs = segmentation_outputs.to(torch.float32)\n\n                    # Calculate segmentation loss\n                    segmentation_loss = binary_cross_entropy_loss(segmentation_outputs, newbatch_segmentation_masks)\n\n                    # Add segmentation loss to the weighted loss\n                    weighted_loss += segmentation_loss\n\n                total_val_loss += weighted_loss.item()\n\n                # Calculate accuracy for each class separately\n                accuracies = []\n                for class_index in range(num_classes_classification+1):\n                    class_labels = batch_labels[:, class_index]\n                    class_outputs = classification_outputs[:, class_index]\n\n                    # Calculate binary predictions based on a threshold (e.g., 0.5)\n                    predicted = (class_outputs > 0.5).float()\n\n                    class_accuracy = accuracy_score(class_labels.cpu(), predicted.cpu())\n                    accuracies.append(class_accuracy)\n\n                batch_accuracy = sum(accuracies) / (num_classes_classification+1)\n                correct_val += batch_accuracy\n                total_val += 1\n\n        val_accuracy = correct_val / total_val\n        avg_val_loss = total_val_loss / len(val_loader)\n        val_loss_history.append(avg_val_loss)\n        val_acc_history.append(val_accuracy)\n\n        print(f\"Epoch [{epoch+1+initial_epoch}/{num_epochs+initial_epoch}]\")\n        print(f\"Validation Accuracy: {val_accuracy:.4f} | Validation Loss: {avg_val_loss:.4f}\")\n\n        if (epoch+1+initial_epoch) % 10== 0:\n            torch.save(model.state_dict(), f\"focalMvitsample_{epoch+1+initial_epoch}.pth\")\n\n            # Save the lists of metrics to a file for later plotting\n        metrics_history = {\n            'val_loss_history': val_loss_history,\n            'val_acc_history': val_acc_history,\n            'train_loss_history': train_loss_history,\n            'train_acc_history': train_acc_history,\n        }\n\n        with open(f\"focalMvitsample_metrics_history.pkl\", 'wb') as f:\n            # This will be a single file, containing all the history\n            pickle.dump(metrics_history, f)\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:51:05.355822Z","iopub.execute_input":"2023-12-14T10:51:05.356189Z","iopub.status.idle":"2023-12-14T10:51:05.39376Z","shell.execute_reply.started":"2023-12-14T10:51:05.356158Z","shell.execute_reply":"2023-12-14T10:51:05.392681Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"rsnamodel(num_epochs=30, initial_epoch=70, ranOnce=True, model_path='/kaggle/input/mvitfocaladam-40-70epoches/focalMvitsample_70.pth',history_path='/kaggle/input/mvitfocaladam-40-70epoches/focalMvitsample_metrics_history.pkl')","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import torch\n\n# # Your original tensor\n# original_tensor = torch.randn(4, 1, 128, 128, 128)\n\n# # Number of frames to add\n# num_frames_to_add = 3\n\n# # Create a list of tensors by repeating the original tensor\n# tensor_list = [original_tensor] * num_frames_to_add\n\n# # Concatenate along the frames dimension (dimension 1)\n# new_tensor = torch.cat(tensor_list, dim=1)\n\n# # Check the shape of the new tensor\n# print(new_tensor.shape)\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:51:07.46609Z","iopub.execute_input":"2023-12-14T10:51:07.466848Z","iopub.status.idle":"2023-12-14T10:51:07.471555Z","shell.execute_reply.started":"2023-12-14T10:51:07.466816Z","shell.execute_reply":"2023-12-14T10:51:07.470315Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import torch\n\n# # Assuming your input tensor is named 'input_tensor'\n# input_tensor = torch.randn(4, 16, 128, 128, 128)\n\n# # Reshape the tensor\n# resized_tensor = input_tensor[:, :, :3, ...].contiguous()\n\n# # Print the new shape\n# print(resized_tensor.shape)\n","metadata":{"execution":{"iopub.status.busy":"2023-12-14T10:51:08.249222Z","iopub.execute_input":"2023-12-14T10:51:08.250101Z","iopub.status.idle":"2023-12-14T10:51:08.25432Z","shell.execute_reply.started":"2023-12-14T10:51:08.250071Z","shell.execute_reply":"2023-12-14T10:51:08.25332Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# rsnamodel(num_epochs=10, initial_epoch=0, ranOnce=False)","metadata":{"execution":{"iopub.status.busy":"2023-12-14T07:07:15.3906Z","iopub.execute_input":"2023-12-14T07:07:15.390958Z","iopub.status.idle":"2023-12-14T07:15:59.501183Z","shell.execute_reply.started":"2023-12-14T07:07:15.39093Z","shell.execute_reply":"2023-12-14T07:15:59.499979Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import torch\n\n# # Assuming your input tensor is named 'input_tensor'\n# input_tensor = torch.randn(4, 16, 3, 128, 128)\n# weight = torch.randn(96, 16, 3, 7, 7)  # Adjust the second dimension to match input channels\n\n# # Perform convolution\n# output = torch.nn.functional.conv3d(input_tensor, weight, groups=1)\n\n# # Print the output shape\n# print(output.shape)\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import numpy as np\n\n# # Generate a random tensor\n# random_tensor = np.random.randn(4, 1, 16, 224, 224)\n\n# # Print the shape of the generated tensor\n# print(random_tensor.shape)\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import torch\n# from torchvision.models.video import mvit_v2_s, MViT_V2_S_Weights\n\n# model = mvit_v2_s(weights=\"DEFAULT\")\n# model.eval()\n# transforms = MViT_V2_S_Weights.KINETICS400_V1.transforms()\n# input = transforms(torch.rand(4, 16, 3, 128, 128))\n# print(input.shape)\n# output = model(input)\n# print(output.shape)\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import torch\n# from torchvision.models.video import mvit_v2_s, MViT_V2_S_Weights\n\n# model = mvit_v2_s(weights=\"DEFAULT\")\n# model.eval()\n# transforms = MViT_V2_S_Weights.KINETICS400_V1.transforms()\n# input = transforms(torch.rand(2, 16, 3, 224, 224))\n# output = model(input)\n# print(output)\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import torch\n# from torchvision.models.video import swin3d_b, Swin3D_B_Weights\n\n# model = swin3d_b(weights=\"DEFAULT\")\n# model.eval()\n# transforms = Swin3D_B_Weights.KINETICS400_V1.transforms()\n# input = transforms(torch.rand(2, 16, 3, 224, 224))\n# output = model(input)\n# print(output)\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import pickle\n# import matplotlib.pyplot as plt\n\n# # history_path = '/kaggle/input/train-run-1/metrics_history.pkl'\n# history_path = '/kaggle/input/adam-focal-loss-after-100-epoch/adam_focal_loss_metrics_history.pkl'\n\n# # Load the metrics history from the saved file\n# with open(history_path, 'rb') as f:\n#     metrics_history = pickle.load(f)\n\n# # Extract the lists of metrics\n# train_accuracies = metrics_history['train_acc_history']\n# val_accuracies = metrics_history['val_acc_history']\n# train_losses = metrics_history['train_loss_history']\n# val_losses = metrics_history['val_loss_history']\n\n# # Create a list of epoch numbers for the x-axis\n# epochs = list(range(1, len(train_accuracies) + 1))\n\n# # Plot training and validation accuracies\n# plt.figure(figsize=(12, 5))\n# plt.subplot(1, 2, 1)\n# plt.plot(epochs, train_accuracies, label='Train Accuracy', marker='o')\n# plt.plot(epochs, val_accuracies, label='Validation Accuracy', marker='o')\n# plt.xlabel('Epoch')\n# plt.ylabel('Accuracy')\n# plt.title('Training and Validation Accuracies')\n# plt.legend()\n\n# # Plot training and validation losses\n# plt.subplot(1, 2, 2)\n# plt.plot(epochs, train_losses, label='Train Loss', marker='o')\n# plt.plot(epochs, val_losses, label='Validation Loss', marker='o')\n# plt.xlabel('Epoch')\n# plt.ylabel('Loss')\n# plt.title('Training and Validation Losses')\n# plt.legend()\n\n# plt.tight_layout()\n# plt.show()\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# import numpy as np\n\n# # Check if there is data in the lists before concatenating\n# if all_labels and all_predictions:\n#     all_labels = np.concatenate(all_labels)\n#     all_predictions = np.concatenate(all_predictions)\n# else:\n#     # Handle the case when there is no data\n#     print(\"No data points to calculate metrics.\")\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# from sklearn.metrics import precision_score, recall_score, f1_score, accuracy_score\n# # Define your test data loader here\n# test_data = pd.read_csv('/kaggle/input/data-divided/exp_test.csv')\n# test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False, num_workers=num_workers)\n\n# # Instantiate your model\n# model = MultiLabel3DAttentionModel(num_classes_classification, num_classes_segmentation)\n# device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n# model.to(device)\n\n# # Load the saved model checkpoint\n# model.load_state_dict(torch.load('/kaggle/input/adam-focal-loss-after-100-epoch/adam_focal_loss_metrics_history.pkl'))\n\n\n# # Assuming your test_loader is already defined\n# # test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False, num_workers=num_workers)\n\n# # Switch to evaluation mode\n# model.eval()\n\n\n# # Initialize variables for calculating test loss and accuracy\n# test_loss = 0.0\n# correct_test = 0\n# total_test = 0\n# all_labels = []\n# all_predictions = []\n\n# with torch.no_grad():\n#     for batch_images, batch_segmentation_masks, batch_labels in test_loader:\n#         # Move data to the GPU if available\n#         batch_images = batch_images.to(device, dtype=torch.float32)\n#         batch_labels = batch_labels.to(device, dtype=torch.float32)\n\n#         # Assuming batch_images has shape (batch_size, num_frames, num_channels, height, width)\n#         batch_images = batch_images.unsqueeze(1)  # Add a singleton dimension for channels\n\n#         # Forward pass\n#         classification_outputs, _ = model(batch_images, None)  # No need for segmentation in the test phase\n\n#         # Apply sigmoid activation to the classification outputs\n#         classification_outputs = torch.sigmoid(classification_outputs)\n#         weights = torch.stack([weight_calculate(labels) for labels in batch_labels]).to(device)\n\n#         # Calculate test loss using your defined criterion (focal_loss)\n#         weighted_loss = weighted_cross_entropy(classification_outputs, batch_labels, weights)\n#         test_loss += weighted_loss.item()\n\n#         # Calculate accuracy for each class separately (similar to training)\n#         accuracies = []\n#         for class_index in range(num_classes_classification):\n#             class_labels = batch_labels[:, class_index]\n#             class_outputs = classification_outputs[:, class_index]\n\n#             # Calculate binary predictions based on a threshold (e.g., 0.5)\n#             predicted = (class_outputs > 0.5).float()\n\n#             class_accuracy = accuracy_score(class_labels.cpu(), predicted.cpu())\n#             accuracies.append(class_accuracy)\n\n#         # Calculate overall accuracy for this batch\n#         batch_accuracy = sum(accuracies) / num_classes_classification\n#         correct_test += batch_accuracy\n#         total_test += 1\n\n# # Calculate and print average test accuracy and loss\n# avg_test_accuracy = correct_test / total_test\n# avg_test_loss = test_loss / len(test_loader)\n# print(\"Test Accuracy: {:.4f} | Test Loss: {:.4f}\".format(avg_test_accuracy, avg_test_loss))\n\n\n# # Flatten the lists of labels and predictions\n# # Check if there is data in the lists before concatenating\n# if all_labels and all_predictions:\n#     all_labels = np.concatenate(all_labels)\n#     all_predictions = np.concatenate(all_predictions)\n# else:\n#     # Handle the case when there is no data\n#     print(\"No data points to calculate metrics.\")\n\n# # Calculate F1 score, precision, recall, and accuracy\n# f1 = f1_score(all_labels, all_predictions, average='weighted')\n# precision = precision_score(all_labels, all_predictions, average='weighted')\n# recall = recall_score(all_labels, all_predictions, average='weighted')\n# accuracy = accuracy_score(all_labels, all_predictions)\n\n# print(\"Test F1 Score: {:.4f}\".format(f1))\n# print(\"Test Precision: {:.4f}\".format(precision))\n# print(\"Test Recall: {:.4f}\".format(recall))\n# print(\"Test Accuracy: {:.4f}\".format(accuracy))\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}