main.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tree Classification with Deep Neural Network / Random Forest"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataset Preprocessing\n",
    "\n",
    "The dataset preprocessing includes loading the raw npy dataset that has been created with the convertToNpy.py script. The loaded data is then separated into the X (input array) and Y (ground truth label). The input array is then normalized using the min-max scaler which is implemented in the NormalizeMatrix function. Then the data is split into training, validation, and test dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15657, 30, 5, 5)\n",
      "(15657,)\n",
      "NVIDIA GeForce GTX 1050\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "from torch.utils.data import DataLoader\n",
    "from TreeDataset import TreeDataset\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "from sklearn.preprocessing import OneHotEncoder, LabelEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "from torch.nn import init\n",
    "import torchvision.models as models\n",
    "from torchvision.models import ResNet50_Weights\n",
    "import torchvision.transforms as transforms\n",
    "import matplotlib.pyplot as plt\n",
    "#convert image matrix to a numpy array and feed it into dataset\n",
    "\n",
    "dataset = np.load(\"assets/new_data_preprocessed_30bands.npz\", allow_pickle=True)\n",
    "\n",
    "# X_train = train_data[:,:,:-1]\n",
    "# X_train = np.c_[X_train[:][:][:].ravel().tolist()]\n",
    "# y_train = train_data[:,:,-1].reshape(-1)\n",
    "\n",
    "# X_test = test_data[:,:,:-1]\n",
    "# X_test = np.c_[X_test[:][:][:].ravel().tolist()]\n",
    "# y_test = test_data[:,:,-1].reshape(-1)\n",
    "\n",
    "# np.savez(\"sample_train_interm\",image=X_train,label=y_train)\n",
    "# np.savez(\"sample_test_interm\",image=X_test,label=y_test)\n",
    "\n",
    "\n",
    "print(dataset['image'].shape)\n",
    "print(dataset['labels'].shape)\n",
    "\n",
    "image = dataset['image']\n",
    "labels = dataset['labels']\n",
    "\n",
    "le = LabelEncoder()\n",
    "encoded_labels = le.fit_transform(labels)\n",
    "\n",
    "if torch.cuda.is_available():\n",
    "    device = torch.device('cuda')\n",
    "    print(torch.cuda.get_device_name())\n",
    "else:\n",
    "    device = torch.device('cpu')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "transform_img = transforms.Compose([\n",
    "    transforms.RandomHorizontalFlip(0.3),\n",
    "    transforms.RandomVerticalFlip(0.3),\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 2 3 4 5 6 7 8 9]\n",
      "[ 588  956  488  442  509 2303 1865  872  778  592]\n",
      "9393\n"
     ]
    }
   ],
   "source": [
    "#separate dataset into training, validation, and test dataset (stratified split)\n",
    "\n",
    "X_train_val, X_test, y_train_val, y_test = train_test_split(image, encoded_labels, test_size=0.2, random_state=42,stratify=labels)\n",
    "\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train_val, y_train_val, test_size=0.25, random_state=28,stratify=y_train_val)\n",
    "\n",
    "# np.savez(\"dataset/train_dataset.npz\", data= X_train, labels= y_train)\n",
    "# np.savez(\"dataset/val_dataset.npz\", data= X_val, labels= y_val)\n",
    "# np.savez(\"dataset/test_dataset.npz\", data= X_test, labels= y_test)\n",
    "\n",
    "\n",
    "X_train_mean = torch.from_numpy(np.mean(X_train, axis=(0,2,3)))\n",
    "X_train_std = torch.from_numpy(np.std(X_train, axis=(0,2,3)))\n",
    "\n",
    "normalized_X_train =  transforms.Normalize(X_train_mean, X_train_std)(torch.from_numpy(X_train))\n",
    "normalized_X_val = transforms.Normalize(X_train_mean, X_train_std)(torch.from_numpy(X_val))\n",
    "normalized_X_test = transforms.Normalize(X_train_mean, X_train_std)(torch.from_numpy(X_test))\n",
    "\n",
    "batch_size = 32\n",
    "\n",
    "train_dataset = TreeDataset(normalized_X_train,y_train,resize=False, transforms=transform_img)\n",
    "val_dataset = TreeDataset(normalized_X_val, y_val,resize=False)\n",
    "test_dataset = TreeDataset(normalized_X_test, y_test,resize=False)\n",
    "\n",
    "print(np.unique(train_dataset.labels))\n",
    "#weight classes\n",
    "class_counts = np.bincount(train_dataset.labels.astype(int))\n",
    "total_samples = len(train_dataset)\n",
    "class_weights = 1.0 / (class_counts/ total_samples) #(class_counts/ total_samples)\n",
    "print(class_counts)\n",
    "print(total_samples)\n",
    "# sampler = torch.utils.data.sampler.WeightedRandomSampler(class_weights, total_samples,replacement=True)\n",
    "\n",
    "# Create data loaders for the training and test sets\n",
    "\n",
    "train_loader = DataLoader(train_dataset, batch_size=batch_size,shuffle=True, drop_last=True)\n",
    "val_loader = DataLoader(val_dataset, batch_size=batch_size, shuffle=False, drop_last=True)\n",
    "test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False,drop_last=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ResNetModified(\n",
      "  (resnet): ResNet(\n",
      "    (conv1): Conv2d(30, 64, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), bias=False)\n",
      "    (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "    (relu): ReLU(inplace=True)\n",
      "    (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
      "    (layer1): Sequential(\n",
      "      (0): Bottleneck(\n",
      "        (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (downsample): Sequential(\n",
      "          (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        )\n",
      "      )\n",
      "      (1): Bottleneck(\n",
      "        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (2): Bottleneck(\n",
      "        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "    )\n",
      "    (layer2): Sequential(\n",
      "      (0): Bottleneck(\n",
      "        (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (downsample): Sequential(\n",
      "          (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        )\n",
      "      )\n",
      "      (1): Bottleneck(\n",
      "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (2): Bottleneck(\n",
      "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (3): Bottleneck(\n",
      "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "    )\n",
      "    (layer3): Sequential(\n",
      "      (0): Bottleneck(\n",
      "        (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (downsample): Sequential(\n",
      "          (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "          (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        )\n",
      "      )\n",
      "      (1): Bottleneck(\n",
      "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (2): Bottleneck(\n",
      "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (3): Bottleneck(\n",
      "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (4): Bottleneck(\n",
      "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (5): Bottleneck(\n",
      "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "    )\n",
      "    (layer4): Sequential(\n",
      "      (0): Bottleneck(\n",
      "        (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (downsample): Sequential(\n",
      "          (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "          (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        )\n",
      "      )\n",
      "      (1): Bottleneck(\n",
      "        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "      (2): Bottleneck(\n",
      "        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "      )\n",
      "    )\n",
      "    (avgpool): AdaptiveAvgPool2d(output_size=(1, 1))\n",
      "    (fc): Linear(in_features=2048, out_features=10, bias=True)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "from experimental_models import ResNetExtended, ResNetModified\n",
    "\n",
    "model = ResNetModified(30,10)\n",
    "\n",
    "print(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from experimental_models import FullyConnectedNetwork\n",
    "\n",
    "# model = FullyConnectedNetwork(5,30,0.3,10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from torch.optim.lr_scheduler import CyclicLR, ReduceLROnPlateau\n",
    "\n",
    "loss_fn = nn.CrossEntropyLoss(weight=torch.from_numpy(class_weights).float()) # weight=torch.from_numpy(class_weights).float()\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.0001)\n",
    "# scheduler = ReduceLROnPlateau(optimizer, factor=0.1,patience=5)\n",
    "scheduler = CyclicLR(optimizer, base_lr=0.001,max_lr=0.0001,cycle_momentum=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def train_loop(dataloader, model, loss_fn, optimizer):\n",
    "#     size = len(dataloader.dataset)\n",
    "#     # Set the model to training mode - important for batch normalization and dropout layers\n",
    "#     # Unnecessary in this situation but added for best practices\n",
    "#     model.train()\n",
    "#     for batch, (X, y) in enumerate(dataloader):\n",
    "#         # Compute prediction and loss\n",
    "#         pred = model(X.float())\n",
    "#         loss = loss_fn(pred, y)\n",
    "\n",
    "#         # Backpropagation\n",
    "#         loss.backward()\n",
    "#         optimizer.step()\n",
    "#         optimizer.zero_grad()\n",
    "#         # Calculate Loss\n",
    "#         if batch % 100 == 0:\n",
    "#             loss, current = loss.item(), (batch + 1) * len(X)\n",
    "#             print(f\"loss: {loss:>7f}  [{current:>5d}/{size:>5d}]\")\n",
    "        \n",
    "        \n",
    "def train_one_epoch(epoch_index, tb_writer):\n",
    "    running_loss = 0.\n",
    "    last_loss = 0.\n",
    "\n",
    "    # Here, we use enumerate(training_loader) instead of\n",
    "    # iter(training_loader) so that we can track the batch\n",
    "    # index and do some intra-epoch reporting\n",
    "    for i, data in enumerate(train_loader):\n",
    "        # Every data instance is an input + label pair\n",
    "        inputs, labels = data\n",
    "        # Zero your gradients for every batch\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        # Make predictions for this batch\n",
    "        outputs = model(inputs.float())\n",
    "        # Compute the loss and its gradients\n",
    "        loss = loss_fn(outputs, labels.long())\n",
    "        loss.backward()\n",
    "\n",
    "        # Adjust learning weights\n",
    "        optimizer.step()\n",
    "\n",
    "        # Gather data and report\n",
    "        running_loss += loss.item()\n",
    "        if i % 100 == 99:\n",
    "            last_loss = running_loss / 1000 # loss per batch\n",
    "            print('  batch {} loss: {}'.format(i + 1, last_loss))\n",
    "            tb_x = epoch_index * len(train_loader) + i + 1\n",
    "            tb_writer.add_scalar('Loss/train', last_loss, tb_x)\n",
    "            running_loss = 0.\n",
    "        \n",
    "        \n",
    "\n",
    "    return last_loss\n",
    "\n",
    "def test_loop(dataloader, model, loss_fn):\n",
    "    # Set the model to evaluation mode - important for batch normalization and dropout layers\n",
    "    # Unnecessary in this situation but added for best practices\n",
    "    model.eval()\n",
    "    size = len(dataloader.dataset)\n",
    "    num_batches = len(dataloader)\n",
    "    test_loss, correct = 0, 0\n",
    "\n",
    "    # Evaluating the model with torch.no_grad() ensures that no gradients are computed during test mode\n",
    "    # also serves to reduce unnecessary gradient computations and memory usage for tensors with requires_grad=True\n",
    "    with torch.no_grad():\n",
    "        for X, y in dataloader:\n",
    "            pred = model(X.float())\n",
    "            test_loss += loss_fn(pred, y.long()).item()\n",
    "            correct += (pred.argmax(1) == y).type(torch.float).sum().item()\n",
    "\n",
    "    test_loss /= num_batches\n",
    "    correct /= size\n",
    "    print(f\"Test Error: \\n Accuracy: {(100*correct):>0.1f}%, Avg loss: {test_loss:>8f} \\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Training\n",
    "\n",
    "The cell below implements the tensorboard utility from the torch library to save epoch results and best model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EPOCH 1:\n"
     ]
    },
    {
     "ename": "RuntimeError",
     "evalue": "Given groups=1, weight of size [64, 61, 5, 5], expected input[32, 5, 5, 61] to have 61 channels, but got 5 channels instead",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[18], line 42\u001b[0m\n\u001b[0;32m     40\u001b[0m \u001b[39m# Make sure gradient tracking is on, and do a pass over the data\u001b[39;00m\n\u001b[0;32m     41\u001b[0m model\u001b[39m.\u001b[39mtrain(\u001b[39mTrue\u001b[39;00m)\n\u001b[1;32m---> 42\u001b[0m avg_loss \u001b[39m=\u001b[39m train_one_epoch(epoch_number, writer)\n\u001b[0;32m     45\u001b[0m running_vloss \u001b[39m=\u001b[39m \u001b[39m0.0\u001b[39m\n\u001b[0;32m     46\u001b[0m \u001b[39m# Set the model to evaluation mode, disabling dropout and using population\u001b[39;00m\n\u001b[0;32m     47\u001b[0m \u001b[39m# statistics for batch normalization.\u001b[39;00m\n",
      "Cell \u001b[1;32mIn[17], line 35\u001b[0m, in \u001b[0;36mtrain_one_epoch\u001b[1;34m(epoch_index, tb_writer)\u001b[0m\n\u001b[0;32m     32\u001b[0m optimizer\u001b[39m.\u001b[39mzero_grad()\n\u001b[0;32m     34\u001b[0m \u001b[39m# Make predictions for this batch\u001b[39;00m\n\u001b[1;32m---> 35\u001b[0m outputs \u001b[39m=\u001b[39m model(inputs\u001b[39m.\u001b[39;49mfloat())\n\u001b[0;32m     36\u001b[0m \u001b[39m# Compute the loss and its gradients\u001b[39;00m\n\u001b[0;32m     37\u001b[0m loss \u001b[39m=\u001b[39m loss_fn(outputs, labels\u001b[39m.\u001b[39mlong())\n",
      "File \u001b[1;32mc:\\Users\\USER\\anaconda3\\envs\\DSEO_project\\Lib\\site-packages\\torch\\nn\\modules\\module.py:1501\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1496\u001b[0m \u001b[39m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[0;32m   1497\u001b[0m \u001b[39m# this function, and just call forward.\u001b[39;00m\n\u001b[0;32m   1498\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m (\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_pre_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_pre_hooks\n\u001b[0;32m   1499\u001b[0m         \u001b[39mor\u001b[39;00m _global_backward_pre_hooks \u001b[39mor\u001b[39;00m _global_backward_hooks\n\u001b[0;32m   1500\u001b[0m         \u001b[39mor\u001b[39;00m _global_forward_hooks \u001b[39mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[1;32m-> 1501\u001b[0m     \u001b[39mreturn\u001b[39;00m forward_call(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[0;32m   1502\u001b[0m \u001b[39m# Do not call functions when jit is used\u001b[39;00m\n\u001b[0;32m   1503\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[39m=\u001b[39m [], []\n",
      "File \u001b[1;32md:\\Programming\\Python\\machinelearning\\DSEO\\experimental_models.py:31\u001b[0m, in \u001b[0;36mResNetModified.forward\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m     30\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mforward\u001b[39m(\u001b[39mself\u001b[39m, x):\n\u001b[1;32m---> 31\u001b[0m     x \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mresnet(x)\n\u001b[0;32m     32\u001b[0m     \u001b[39mreturn\u001b[39;00m x\n",
      "File \u001b[1;32mc:\\Users\\USER\\anaconda3\\envs\\DSEO_project\\Lib\\site-packages\\torch\\nn\\modules\\module.py:1501\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1496\u001b[0m \u001b[39m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[0;32m   1497\u001b[0m \u001b[39m# this function, and just call forward.\u001b[39;00m\n\u001b[0;32m   1498\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m (\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_pre_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_pre_hooks\n\u001b[0;32m   1499\u001b[0m         \u001b[39mor\u001b[39;00m _global_backward_pre_hooks \u001b[39mor\u001b[39;00m _global_backward_hooks\n\u001b[0;32m   1500\u001b[0m         \u001b[39mor\u001b[39;00m _global_forward_hooks \u001b[39mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[1;32m-> 1501\u001b[0m     \u001b[39mreturn\u001b[39;00m forward_call(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[0;32m   1502\u001b[0m \u001b[39m# Do not call functions when jit is used\u001b[39;00m\n\u001b[0;32m   1503\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[39m=\u001b[39m [], []\n",
      "File \u001b[1;32mc:\\Users\\USER\\anaconda3\\envs\\DSEO_project\\Lib\\site-packages\\torchvision\\models\\resnet.py:285\u001b[0m, in \u001b[0;36mResNet.forward\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m    284\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mforward\u001b[39m(\u001b[39mself\u001b[39m, x: Tensor) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m Tensor:\n\u001b[1;32m--> 285\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_forward_impl(x)\n",
      "File \u001b[1;32mc:\\Users\\USER\\anaconda3\\envs\\DSEO_project\\Lib\\site-packages\\torchvision\\models\\resnet.py:268\u001b[0m, in \u001b[0;36mResNet._forward_impl\u001b[1;34m(self, x)\u001b[0m\n\u001b[0;32m    266\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m_forward_impl\u001b[39m(\u001b[39mself\u001b[39m, x: Tensor) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m Tensor:\n\u001b[0;32m    267\u001b[0m     \u001b[39m# See note [TorchScript super()]\u001b[39;00m\n\u001b[1;32m--> 268\u001b[0m     x \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mconv1(x)\n\u001b[0;32m    269\u001b[0m     x \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mbn1(x)\n\u001b[0;32m    270\u001b[0m     x \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mrelu(x)\n",
      "File \u001b[1;32mc:\\Users\\USER\\anaconda3\\envs\\DSEO_project\\Lib\\site-packages\\torch\\nn\\modules\\module.py:1501\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1496\u001b[0m \u001b[39m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[0;32m   1497\u001b[0m \u001b[39m# this function, and just call forward.\u001b[39;00m\n\u001b[0;32m   1498\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m (\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_backward_pre_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_hooks \u001b[39mor\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_forward_pre_hooks\n\u001b[0;32m   1499\u001b[0m         \u001b[39mor\u001b[39;00m _global_backward_pre_hooks \u001b[39mor\u001b[39;00m _global_backward_hooks\n\u001b[0;32m   1500\u001b[0m         \u001b[39mor\u001b[39;00m _global_forward_hooks \u001b[39mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[1;32m-> 1501\u001b[0m     \u001b[39mreturn\u001b[39;00m forward_call(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[0;32m   1502\u001b[0m \u001b[39m# Do not call functions when jit is used\u001b[39;00m\n\u001b[0;32m   1503\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[39m=\u001b[39m [], []\n",
      "File \u001b[1;32mc:\\Users\\USER\\anaconda3\\envs\\DSEO_project\\Lib\\site-packages\\torch\\nn\\modules\\conv.py:463\u001b[0m, in \u001b[0;36mConv2d.forward\u001b[1;34m(self, input)\u001b[0m\n\u001b[0;32m    462\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mforward\u001b[39m(\u001b[39mself\u001b[39m, \u001b[39minput\u001b[39m: Tensor) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m Tensor:\n\u001b[1;32m--> 463\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_conv_forward(\u001b[39minput\u001b[39;49m, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mweight, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mbias)\n",
      "File \u001b[1;32mc:\\Users\\USER\\anaconda3\\envs\\DSEO_project\\Lib\\site-packages\\torch\\nn\\modules\\conv.py:459\u001b[0m, in \u001b[0;36mConv2d._conv_forward\u001b[1;34m(self, input, weight, bias)\u001b[0m\n\u001b[0;32m    455\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mpadding_mode \u001b[39m!=\u001b[39m \u001b[39m'\u001b[39m\u001b[39mzeros\u001b[39m\u001b[39m'\u001b[39m:\n\u001b[0;32m    456\u001b[0m     \u001b[39mreturn\u001b[39;00m F\u001b[39m.\u001b[39mconv2d(F\u001b[39m.\u001b[39mpad(\u001b[39minput\u001b[39m, \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_reversed_padding_repeated_twice, mode\u001b[39m=\u001b[39m\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mpadding_mode),\n\u001b[0;32m    457\u001b[0m                     weight, bias, \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mstride,\n\u001b[0;32m    458\u001b[0m                     _pair(\u001b[39m0\u001b[39m), \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdilation, \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mgroups)\n\u001b[1;32m--> 459\u001b[0m \u001b[39mreturn\u001b[39;00m F\u001b[39m.\u001b[39;49mconv2d(\u001b[39minput\u001b[39;49m, weight, bias, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mstride,\n\u001b[0;32m    460\u001b[0m                 \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mpadding, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mdilation, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mgroups)\n",
      "\u001b[1;31mRuntimeError\u001b[0m: Given groups=1, weight of size [64, 61, 5, 5], expected input[32, 5, 5, 61] to have 61 channels, but got 5 channels instead"
     ]
    }
   ],
   "source": [
    "from torch.utils.tensorboard import SummaryWriter\n",
    "from datetime import datetime\n",
    "\n",
    "class EarlyStopper:\n",
    "    def __init__(self, patience=1, min_delta=0):\n",
    "        self.patience = patience\n",
    "        self.min_delta = min_delta\n",
    "        self.counter = 0\n",
    "        self.min_validation_loss = np.inf\n",
    "\n",
    "    def early_stop(self, validation_loss):\n",
    "        if validation_loss < self.min_validation_loss:\n",
    "            self.min_validation_loss = validation_loss\n",
    "            self.counter = 0\n",
    "        elif validation_loss > (self.min_validation_loss + self.min_delta):\n",
    "            self.counter += 1\n",
    "            if self.counter >= self.patience:\n",
    "                return True\n",
    "        return False\n",
    "\n",
    "\n",
    "\n",
    "# Initializing in a separate cell so we can easily add more epochs to the same run\n",
    "timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')\n",
    "writer = SummaryWriter('runs/tree_class_trainer_{}'.format(timestamp))\n",
    "\n",
    "epoch_number = 0\n",
    "\n",
    "EPOCHS = 100\n",
    "\n",
    "best_vloss = 1_000_000.\n",
    "\n",
    "epochs_cum = []\n",
    "tlosses_cum = []\n",
    "vlosses_cum = []\n",
    "\n",
    "for epoch in range(EPOCHS):\n",
    "    print('EPOCH {}:'.format(epoch_number + 1))\n",
    "\n",
    "    # Make sure gradient tracking is on, and do a pass over the data\n",
    "    model.train(True)\n",
    "    avg_loss = train_one_epoch(epoch_number, writer)\n",
    "\n",
    "\n",
    "    running_vloss = 0.0\n",
    "    # Set the model to evaluation mode, disabling dropout and using population\n",
    "    # statistics for batch normalization.\n",
    "    model.eval()\n",
    "\n",
    "    # Disable gradient computation and reduce memory consumption.\n",
    "    with torch.no_grad():\n",
    "        for i, vdata in enumerate(val_loader):\n",
    "            vinputs, vlabels = vdata\n",
    "            voutputs = model(vinputs.float())\n",
    "            vloss = loss_fn(voutputs, vlabels.long())\n",
    "            running_vloss += vloss\n",
    "\n",
    "    avg_vloss = running_vloss / (i + 1)\n",
    "    print('LOSS train {} valid {}'.format(avg_loss, avg_vloss))\n",
    "\n",
    "    scheduler.step()\n",
    "    \n",
    "    tlosses_cum.append(avg_loss)\n",
    "    vlosses_cum.append(avg_vloss)\n",
    "    epochs_cum.append(epoch+1)\n",
    "\n",
    "    # Log the running loss averaged per batch\n",
    "    # for both training and validation\n",
    "    writer.add_scalars('Training vs. Validation Loss',\n",
    "                    { 'Training' : avg_loss, 'Validation' : avg_vloss },\n",
    "                    epoch_number + 1)\n",
    "    writer.flush()\n",
    "\n",
    "    # Track best performance, and save the model's state\n",
    "    if avg_vloss < best_vloss:\n",
    "        best_vloss = avg_vloss\n",
    "        model_path = 'fc_model_{}_{}.pth'.format(timestamp, epoch_number)\n",
    "        torch.save(model, model_path)\n",
    "\n",
    "    early_stopper = EarlyStopper(patience=3, min_delta=1)\n",
    "    if early_stopper.early_stop(avg_loss):             \n",
    "        break\n",
    "    epoch_number += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(epochs_cum)\n",
    "plt.figure()\n",
    "plt.plot(epochs_cum, tlosses_cum, color='blue', label='Training Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Training loss')\n",
    "plt.title('Training Losses over Epochs')\n",
    "plt.figure()\n",
    "plt.plot(epochs_cum, vlosses_cum, color='red', label='Validation Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Validation Loss')\n",
    "plt.title('Validation Losses over Epochs')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Abies alba': 0, 'Betula pendula': 1, 'Carpinus betulus': 2, 'Fraxinus excelsior': 3, 'Larix decidua': 4, 'Picea abies': 5, 'Pinus sylvestris': 6, 'Quercus petraea': 7, 'Quercus robur': 8, 'Sorbus aucuparia': 9}\n"
     ]
    }
   ],
   "source": [
    "le_name_mapping = dict(zip(le.classes_, le.transform(le.classes_)))\n",
    "print(le_name_mapping)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Grid Search\n",
    "\n",
    "The code below was used to look for optimal hyperparameters (learning rate, epoch number). The cell is commented for now to test the obtained hyperparameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from skorch import NeuralNetClassifier\n",
    "# from sklearn.model_selection import GridSearchCV\n",
    "# from utils import create_search_run, save_best_hyperparam\n",
    "\n",
    "# search_folder = create_search_run()\n",
    "\n",
    "# # Learning parameters. \n",
    "# lr = 0.001\n",
    "# epochs = 20\n",
    "# print(f\"Computation device: {device}\\n\")\n",
    "\n",
    "# # Loss function. Required for defining `NeuralNetClassifier`\n",
    "# criterion = nn.CrossEntropyLoss()\n",
    "\n",
    "\n",
    "# # Instance of `NeuralNetClassifier` to be passed to `GridSearchCV` \n",
    "# net = NeuralNetClassifier(\n",
    "#     module=model, max_epochs=epochs,\n",
    "#     optimizer=optim.Adam,\n",
    "#     criterion=criterion,\n",
    "#     lr=lr, verbose=1\n",
    "# )\n",
    "\n",
    "\n",
    "# params = {\n",
    "#     'lr': [0.005, 0.01, 0.1],\n",
    "#     'max_epochs': list(range(10, 55, 20)),\n",
    "# }\n",
    "\n",
    "# \"\"\"\n",
    "# Define `GridSearchCV`.\n",
    "# 4 lrs * 7 max_epochs * 4 module__first_conv_out * 3 module__first_fc_out\n",
    "# * 2 CVs = 672 fits.\n",
    "# \"\"\"\n",
    "# gs = GridSearchCV(\n",
    "#     net, params, refit=False, scoring='accuracy', cv=2, verbose=1\n",
    "# )\n",
    "\n",
    "# counter = 0\n",
    "# # Run each fit for 2 batches. So, if we have `n` fits, then it will\n",
    "# # actually for `n*2` times. We have 672 fits, so total, \n",
    "# # 672 * 2 = 1344 runs.\n",
    "# search_batches = 2\n",
    "# \"\"\"\n",
    "# This will run `n` (`n` is calculated from `params`) number of fits \n",
    "# on each batch of data, so be careful.\n",
    "# If you want to run the `n` number of fits just once, \n",
    "# that is, on one batch of data,\n",
    "# add `break` after this line:\n",
    "#     `outputs = gs.fit(image, labels)`\n",
    "\n",
    "# Note: This will take a lot of time to run\n",
    "# \"\"\"\n",
    "# for i, data in enumerate(train_loader):\n",
    "#     counter += 1\n",
    "#     image, labels = data\n",
    "#     # image = image.to(device)\n",
    "#     # labels = labels.to(device)\n",
    "#     outputs = gs.fit(image.float(), labels.long())\n",
    "#     # GridSearch for `search_batches` number of times.\n",
    "#     if counter == search_batches:\n",
    "#         break\n",
    "\n",
    "# print('SEARCH COMPLETE')\n",
    "# print(\"best score: {:.3f}, best params: {}\".format(gs.best_score_, gs.best_params_))\n",
    "# save_best_hyperparam(gs.best_score_, f\"./outputs/{search_folder}/best_param.yml\")\n",
    "# save_best_hyperparam(gs.best_params_, f\"./outputs/{search_folder}/best_param.yml\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# epochs = 10\n",
    "# for t in range(epochs):\n",
    "#     print(f\"Epoch {t+1}\\n-------------------------------\")\n",
    "#     train_loop(train_loader, model, loss_fn, optimizer)\n",
    "#     test_loop(test_loader, model, loss_fn)\n",
    "# print(\"Done!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of the network on the test images: 11 %\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x1d80d9e66d0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "correct = 0\n",
    "total = 0\n",
    "# since we're not training, we don't need to calculate the gradients for our outputs\n",
    "label_collector = []\n",
    "prediction_collector = []\n",
    "model = torch.load(\"fc_model_20230719_130151_38.pth\")\n",
    "\n",
    "with torch.no_grad():\n",
    "    for data in test_loader:\n",
    "        images, labels = data\n",
    "        # calculate outputs by running images through the network\n",
    "        outputs = model(images.float())\n",
    "        # the class with the highest energy is what we choose as prediction\n",
    "        _, predicted = torch.max(outputs.data, 1)\n",
    "        total += labels.size(0)\n",
    "        correct += (predicted == labels).sum().item()\n",
    "        label_collector.append(labels)\n",
    "        prediction_collector.append(predicted)\n",
    "print(f'Accuracy of the network on the test images: {100 * correct // total} %')\n",
    "\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
    "predictions = le.inverse_transform(torch.cat(prediction_collector))\n",
    "labels = le.inverse_transform(torch.cat(label_collector))\n",
    "\n",
    "# for i in range(len(prediction_collector)):\n",
    "#     input_pred_CM = torch.cat(torch.unsqueeze(prediction_collector[i]))\n",
    "#     input_gt_CM = torch.cat(torch.unsqueeze(label_collector[i]))\n",
    "\n",
    "cm = confusion_matrix(labels, predictions)\n",
    "ConfusionMatrixDisplay(cm).plot()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11.872565772547357\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import f1_score\n",
    "\n",
    "f1_scores = f1_score(labels, predictions, average='weighted')\n",
    "\n",
    "print(f1_scores*100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(model.state_dict(),\"D:\\Programming\\Python\\machinelearning\\DSEO\\\\resnet50_model_20230717_183329_94resnet50_model_20230717_183329_200\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(model,\"D:\\Programming\\Python\\machinelearning\\DSEO\\\\fc_model_20230719_130151_38.pth\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "===============================================================================================\n",
       "Layer (type:depth-idx)                        Output Shape              Param #\n",
       "===============================================================================================\n",
       "ResNetExtended                                [32, 10]                  --\n",
       "├─Sequential: 1-1                             [32, 3, 20, 20]           --\n",
       "│    └─Conv2d: 2-1                            [32, 64, 5, 5]            17,344\n",
       "│    └─BatchNorm2d: 2-2                       [32, 64, 5, 5]            128\n",
       "│    └─ReLU: 2-3                              [32, 64, 5, 5]            --\n",
       "│    └─Conv2d: 2-4                            [32, 128, 5, 5]           73,856\n",
       "│    └─BatchNorm2d: 2-5                       [32, 128, 5, 5]           256\n",
       "│    └─ReLU: 2-6                              [32, 128, 5, 5]           --\n",
       "│    └─Conv2d: 2-7                            [32, 3, 5, 5]             3,459\n",
       "│    └─BatchNorm2d: 2-8                       [32, 3, 5, 5]             6\n",
       "│    └─ReLU: 2-9                              [32, 3, 5, 5]             --\n",
       "│    └─Upsample: 2-10                         [32, 3, 20, 20]           --\n",
       "├─ResNet: 1-2                                 [32, 10]                  --\n",
       "│    └─Conv2d: 2-11                           [32, 64, 10, 10]          (9,408)\n",
       "│    └─BatchNorm2d: 2-12                      [32, 64, 10, 10]          (128)\n",
       "│    └─ReLU: 2-13                             [32, 64, 10, 10]          --\n",
       "│    └─MaxPool2d: 2-14                        [32, 64, 5, 5]            --\n",
       "│    └─Sequential: 2-15                       [32, 256, 5, 5]           --\n",
       "│    │    └─Bottleneck: 3-1                   [32, 256, 5, 5]           (75,008)\n",
       "│    │    └─Bottleneck: 3-2                   [32, 256, 5, 5]           (70,400)\n",
       "│    │    └─Bottleneck: 3-3                   [32, 256, 5, 5]           (70,400)\n",
       "│    └─Sequential: 2-16                       [32, 512, 3, 3]           --\n",
       "│    │    └─Bottleneck: 3-4                   [32, 512, 3, 3]           (379,392)\n",
       "│    │    └─Bottleneck: 3-5                   [32, 512, 3, 3]           (280,064)\n",
       "│    │    └─Bottleneck: 3-6                   [32, 512, 3, 3]           (280,064)\n",
       "│    │    └─Bottleneck: 3-7                   [32, 512, 3, 3]           (280,064)\n",
       "│    └─Sequential: 2-17                       [32, 1024, 2, 2]          --\n",
       "│    │    └─Bottleneck: 3-8                   [32, 1024, 2, 2]          (1,512,448)\n",
       "│    │    └─Bottleneck: 3-9                   [32, 1024, 2, 2]          (1,117,184)\n",
       "│    │    └─Bottleneck: 3-10                  [32, 1024, 2, 2]          (1,117,184)\n",
       "│    │    └─Bottleneck: 3-11                  [32, 1024, 2, 2]          (1,117,184)\n",
       "│    │    └─Bottleneck: 3-12                  [32, 1024, 2, 2]          (1,117,184)\n",
       "│    │    └─Bottleneck: 3-13                  [32, 1024, 2, 2]          (1,117,184)\n",
       "│    └─Sequential: 2-18                       [32, 2048, 1, 1]          --\n",
       "│    │    └─Bottleneck: 3-14                  [32, 2048, 1, 1]          (6,039,552)\n",
       "│    │    └─Bottleneck: 3-15                  [32, 2048, 1, 1]          (4,462,592)\n",
       "│    │    └─Bottleneck: 3-16                  [32, 2048, 1, 1]          (4,462,592)\n",
       "│    └─AdaptiveAvgPool2d: 2-19                [32, 2048, 1, 1]          --\n",
       "│    └─Linear: 2-20                           [32, 10]                  20,490\n",
       "===============================================================================================\n",
       "Total params: 23,623,571\n",
       "Trainable params: 115,539\n",
       "Non-trainable params: 23,508,032\n",
       "Total mult-adds (Units.GIGABYTES): 2.10\n",
       "===============================================================================================\n",
       "Input size (MB): 0.10\n",
       "Forward/backward pass size (MB): 69.54\n",
       "Params size (MB): 94.49\n",
       "Estimated Total Size (MB): 164.13\n",
       "==============================================================================================="
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from torchinfo import summary\n",
    "\n",
    "# DOES A FORWARD PASS USING THE DEFAULT DEVICE (HERE CUDA) IF NOT SPECIFIED, CAUSED AN ERROR IN TRAINING LOOP BECAUSE WEIGHTS ARE IN CUDA AND INPUTS ARE IN CPU\n",
    "\n",
    "summary(model, input_size=(batch_size, 30, 5, 5))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "DSEO_project",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.3"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}