Updating stuff

Check_Performance_of_Model.ipynb

@@ -0,0 +1,176 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "83f8ecbf",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "import matplotlib.pyplot as plt\n",
+    "import os\n",
+    "import sys\n",
+    "from tqdm import tqdm\n",
+    "import torch.nn.functional as F\n",
+    "import math\n",
+    "import random\n",
+    "import h5py\n",
+    "# Import utils and models\n",
+    "import utils, models\n",
+    "\n",
+    "print(\"Is PyTorch using GPU?\", torch.cuda.is_available())\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "DTYPE = torch.float32"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8bc27adc",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(\"Loading data...\")\n",
+    "# Import data\n",
+    "hf = h5py.File(\"Data/GP_galactic_data.h5\", \"r\")\n",
+    "\n",
+    "X_training = torch.tensor( np.array(hf.get(\"X_training\")), dtype = DTYPE )\n",
+    "Y_training = torch.tensor( np.array(hf.get(\"Y_training\")), dtype = DTYPE )\n",
+    "\n",
+    "X_val = torch.tensor( np.array(hf.get(\"X_val\")), dtype = DTYPE )\n",
+    "Y_val = torch.tensor( np.array(hf.get(\"Y_val\")), dtype = DTYPE )\n",
+    "\n",
+    "X_test = torch.tensor( np.array(hf.get(\"X_test\")), dtype = DTYPE )\n",
+    "Y_test = torch.tensor( np.array(hf.get(\"Y_test\")), dtype = DTYPE )\n",
+    "\n",
+    "hf.close()\n",
+    "\n",
+    "print(\"Training shape: \", X_training.shape, Y_training.shape)\n",
+    "print(\"Val shape: \", X_val.shape, Y_val.shape)\n",
+    "print(\"Test shape: \", X_test.shape, Y_test.shape)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "78670814",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "batch_size = 5\n",
+    "n_batches = int( X_test.shape[0] / batch_size )\n",
+    "\n",
+    "labeling_order = torch.unique(Y_test, return_counts = False).type(torch.int).detach().cpu().numpy().tolist()\n",
+    "print( \"labeling_order: \", labeling_order )\n",
+    "\n",
+    "# Compute weights from frequency of classes in the training set\n",
+    "unique_training, counts_training = torch.unique( Y_test, return_counts = True )\n",
+    "\n",
+    "inverse_freqs =  1 / counts_training\n",
+    "weights_tensor = ( inverse_freqs / torch.sum(inverse_freqs) ).to(device)\n",
+    "\n",
+    "# Define model\n",
+    "input_dim = 6\n",
+    "n_classes = len(labeling_order)\n",
+    "d_model = 32\n",
+    "nhead = 4\n",
+    "num_layers = 4\n",
+    "\n",
+    "model = models.TransformerClassifier(input_dim = input_dim, n_classes = n_classes, d_model = d_model, nhead = nhead, num_layers = num_layers, weights_tensor = weights_tensor, DTYPE = DTYPE, device = device, labeling_order = labeling_order).to(device)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "3cd7e61e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Load weights\n",
+    "model.load_state_dict( torch.load(\"Different_Tests/Test_1/Models_Data/Model_Saved/model_epoch_61.pt\") )"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "bbf1f331",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "y_true_list, y_pred_list = [], []\n",
+    "\n",
+    "for n in tqdm(range(n_batches)):\n",
+    "    y_pred_batch = model(X_test[n*batch_size:(n+1)*batch_size,0,:,:].to(device))\n",
+    "    y_pred_batch = torch.tensor([labeling_order[i] for i in torch.argmax(y_pred_batch, dim = 1).detach().cpu().numpy().tolist()], dtype = DTYPE).view(-1,1)\n",
+    "    y_true = Y_test[n*batch_size:(n+1)*batch_size]\n",
+    "    y_true_list.append(y_true)\n",
+    "    y_pred_list.append(y_pred_batch)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "fa8d1624",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "y_pred_list"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "96409b9a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model(X_training[n*batch_size:(n+1)*batch_size,0,:,:].to(device))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "cf12d396",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "torch.unique(Y_test, return_counts = True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "174d13a6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "y_true"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.8.10"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}