{
"cells": [
{
"cell_type": "markdown",
"id": "afa53af3",
"metadata": {},
"source": [
"# 06 — Dynamic-Graph iTransformer (v3)\n",
"\n",
"## v3 Improvements\n",
"- **Cross-battery split** (no data leakage)\n",
"- **18 features** per timestep (6 new physics-informed features)\n",
"\n",
"## Architecture\n",
"- **iTransformer backbone** with inverted (feature-wise) attention\n",
"- **Dynamic Graph Convolution** layer that learns inter-feature adjacency dynamically\n",
"- Fuses local (graph) + global (attention) representations\n",
"- Built with TensorFlow/Keras"
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "20d26377",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:16.341799Z",
"iopub.status.busy": "2026-02-24T18:15:16.341799Z",
"iopub.status.idle": "2026-02-24T18:15:18.735982Z",
"shell.execute_reply": "2026-02-24T18:15:18.735982Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"TensorFlow: 2.20.0\n",
"GPUs: []\n"
]
},
{
"data": {
"text/plain": [
"{'root': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3'),\n",
" 'models_classical': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3/models/classical'),\n",
" 'models_deep': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3/models/deep'),\n",
" 'models_ensemble': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3/models/ensemble'),\n",
" 'scalers': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3/scalers'),\n",
" 'figures': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3/figures'),\n",
" 'results': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3/results'),\n",
" 'logs': WindowsPath('E:/VIT/aiBatteryLifecycle/artifacts/v3/logs')}"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import sys, os\n",
"sys.path.insert(0, os.path.abspath(\"..\"))\n",
"\n",
"import numpy as np\n",
"import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"import seaborn as sns\n",
"import warnings\n",
"warnings.filterwarnings(\"ignore\")\n",
"\n",
"import tensorflow as tf\n",
"from src.models.deep.itransformer import build_dynamic_graph_itransformer\n",
"from src.evaluation.metrics import regression_metrics, tolerance_accuracy\n",
"from src.utils.plotting import save_fig\n",
"from src.utils.config import (\n",
" ARTIFACTS_DIR, FIGURES_DIR, MODELS_DIR,\n",
" get_version_paths, ensure_version_dirs,\n",
" WINDOW_SIZE, BATCH_SIZE, MAX_EPOCHS, EARLY_STOP_PATIENCE,\n",
" TRANSFORMER_D_MODEL, TRANSFORMER_NHEAD, TRANSFORMER_LAYERS, DROPOUT,\n",
")\n",
"\n",
"print(f\"TensorFlow: {tf.__version__}\")\n",
"print(f\"GPUs: {tf.config.list_physical_devices('GPU')}\")\n",
"plt.style.use(\"seaborn-v0_8-whitegrid\")\n",
"\n",
"# v3 paths\n",
"v3 = get_version_paths('v3')\n",
"ensure_version_dirs('v3')"
]
},
{
"cell_type": "markdown",
"id": "2b806d3a",
"metadata": {},
"source": [
"## 1. Load & Prepare Data"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "bdd1403f",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:18.738372Z",
"iopub.status.busy": "2026-02-24T18:15:18.738372Z",
"iopub.status.idle": "2026-02-24T18:15:18.758405Z",
"shell.execute_reply": "2026-02-24T18:15:18.758405Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Train: (1444, 32, 18) | Test: (290, 32, 18)\n",
"Overlap: NONE ✓\n"
]
}
],
"source": [
"V3_FEATURES = v3[\"root\"] / \"features\"\n",
"data = np.load(str(V3_FEATURES / \"battery_sequences.npz\"), allow_pickle=True)\n",
"X_multi = data[\"X_multi\"]\n",
"y_multi = data[\"y_multi\"]\n",
"bids = data[\"bids_multi\"]\n",
"\n",
"# ── v3 FIX: Cross-battery grouped split ──\n",
"unique_bids = np.unique(bids)\n",
"rng = np.random.RandomState(42)\n",
"shuffled = unique_bids.copy()\n",
"rng.shuffle(shuffled)\n",
"n_train = max(1, int(len(shuffled) * 0.8))\n",
"train_bats = set(shuffled[:n_train])\n",
"test_bats = set(shuffled[n_train:])\n",
"\n",
"train_mask = np.isin(bids, list(train_bats))\n",
"test_mask = np.isin(bids, list(test_bats))\n",
"\n",
"X_train, y_train = X_multi[train_mask], y_multi[train_mask]\n",
"X_test, y_test = X_multi[test_mask], y_multi[test_mask]\n",
"\n",
"from sklearn.preprocessing import StandardScaler\n",
"n_samples, seq_len, n_feat = X_train.shape\n",
"scaler = StandardScaler().fit(X_train.reshape(-1, n_feat))\n",
"X_train = scaler.transform(X_train.reshape(-1, n_feat)).reshape(n_samples, seq_len, n_feat)\n",
"X_test = scaler.transform(X_test.reshape(-1, n_feat)).reshape(X_test.shape[0], seq_len, n_feat)\n",
"\n",
"print(f\"Train: {X_train.shape} | Test: {X_test.shape}\")\n",
"print(f\"Overlap: {train_bats & test_bats if train_bats & test_bats else 'NONE ✓'}\")"
]
},
{
"cell_type": "markdown",
"id": "71722721",
"metadata": {},
"source": [
"## 2. Build Dynamic-Graph iTransformer"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "06243bd4",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:18.760606Z",
"iopub.status.busy": "2026-02-24T18:15:18.759601Z",
"iopub.status.idle": "2026-02-24T18:15:19.494309Z",
"shell.execute_reply": "2026-02-24T18:15:19.494309Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"WARNING:tensorflow:From e:\\VIT\\aiBatteryLifecycle\\venv\\Lib\\site-packages\\keras\\src\\backend\\tensorflow\\core.py:232: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
"\n"
]
},
{
"data": {
"text/html": [
"Model: \"DynamicGraph_iTransformer\"\n",
"\n"
],
"text/plain": [
"\u001b[1mModel: \"DynamicGraph_iTransformer\"\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ input_layer (InputLayer) │ (None, 32, 18) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dyn_graph (DynamicGraphConv) │ (None, 32, 18) │ 378 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ proj (Dense) │ (None, 32, 64) │ 1,216 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_feat_0 (FeatureWiseMHA) │ (None, 32, 64) │ 8,480 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_token_0 (TokenWiseMHA) │ (None, 32, 64) │ 16,768 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_ff_0 (Conv1DFeedForward) │ (None, 32, 64) │ 33,216 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_feat_1 (FeatureWiseMHA) │ (None, 32, 64) │ 8,480 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_token_1 (TokenWiseMHA) │ (None, 32, 64) │ 16,768 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_ff_1 (Conv1DFeedForward) │ (None, 32, 64) │ 33,216 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ global_average_pooling1d │ (None, 64) │ 0 │\n",
"│ (GlobalAveragePooling1D) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (Dense) │ (None, 128) │ 8,320 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_10 (Dropout) │ (None, 128) │ 0 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (Dense) │ (None, 1) │ 129 │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"\n"
],
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m18\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dyn_graph (\u001b[38;5;33mDynamicGraphConv\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m18\u001b[0m) │ \u001b[38;5;34m378\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ proj (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m1,216\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_feat_0 (\u001b[38;5;33mFeatureWiseMHA\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m8,480\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_token_0 (\u001b[38;5;33mTokenWiseMHA\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m16,768\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_ff_0 (\u001b[38;5;33mConv1DFeedForward\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m33,216\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_feat_1 (\u001b[38;5;33mFeatureWiseMHA\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m8,480\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_token_1 (\u001b[38;5;33mTokenWiseMHA\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m16,768\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dg_ff_1 (\u001b[38;5;33mConv1DFeedForward\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m33,216\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ global_average_pooling1d │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"│ (\u001b[38;5;33mGlobalAveragePooling1D\u001b[0m) │ │ │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m8,320\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dropout_10 (\u001b[38;5;33mDropout\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m129\u001b[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Total params: 126,971 (495.98 KB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m126,971\u001b[0m (495.98 KB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Trainable params: 126,971 (495.98 KB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m126,971\u001b[0m (495.98 KB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Non-trainable params: 0 (0.00 B)\n",
"\n"
],
"text/plain": [
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Total parameters: 126,971\n"
]
}
],
"source": [
"dg_model = build_dynamic_graph_itransformer(\n",
" seq_len=seq_len, n_features=n_feat,\n",
" d_model=TRANSFORMER_D_MODEL, n_heads=TRANSFORMER_NHEAD,\n",
" n_blocks=TRANSFORMER_LAYERS, dropout=DROPOUT,\n",
")\n",
"dg_model.compile(optimizer=tf.keras.optimizers.Adam(1e-3), loss=\"mse\", metrics=[\"mae\"])\n",
"dg_model.summary()\n",
"print(f\"\\nTotal parameters: {dg_model.count_params():,}\")"
]
},
{
"cell_type": "markdown",
"id": "f6d25fe4",
"metadata": {},
"source": [
"## 3. Train"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "a5f69f03",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:19.496320Z",
"iopub.status.busy": "2026-02-24T18:15:19.495314Z",
"iopub.status.idle": "2026-02-24T18:15:53.431445Z",
"shell.execute_reply": "2026-02-24T18:15:53.431445Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 33ms/step - loss: 3107.6074 - mae: 51.8714 - val_loss: 1967.3826 - val_mae: 42.0192 - learning_rate: 0.0010\n",
"Epoch 2/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 25ms/step - loss: 856.7082 - mae: 24.9853 - val_loss: 379.6254 - val_mae: 16.3422 - learning_rate: 0.0010\n",
"Epoch 3/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 541.1316 - mae: 16.8246 - val_loss: 404.9633 - val_mae: 16.7222 - learning_rate: 0.0010\n",
"Epoch 4/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 325.5964 - mae: 13.5706 - val_loss: 277.5700 - val_mae: 11.9630 - learning_rate: 0.0010\n",
"Epoch 5/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 25ms/step - loss: 172.9639 - mae: 9.5096 - val_loss: 276.4447 - val_mae: 12.7127 - learning_rate: 0.0010\n",
"Epoch 6/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 137.8238 - mae: 8.5541 - val_loss: 198.3174 - val_mae: 11.9075 - learning_rate: 0.0010\n",
"Epoch 7/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 108.1718 - mae: 7.3416 - val_loss: 192.5336 - val_mae: 11.3964 - learning_rate: 0.0010\n",
"Epoch 8/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 106.9558 - mae: 7.3110 - val_loss: 165.8024 - val_mae: 10.5384 - learning_rate: 0.0010\n",
"Epoch 9/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 25ms/step - loss: 92.9293 - mae: 6.6659 - val_loss: 157.7313 - val_mae: 10.4017 - learning_rate: 0.0010\n",
"Epoch 10/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 76.1936 - mae: 5.9393 - val_loss: 124.5216 - val_mae: 9.2149 - learning_rate: 0.0010\n",
"Epoch 11/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 72.6647 - mae: 5.9936 - val_loss: 147.7324 - val_mae: 10.5025 - learning_rate: 0.0010\n",
"Epoch 12/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 61.0835 - mae: 5.4357 - val_loss: 107.3129 - val_mae: 8.7674 - learning_rate: 0.0010\n",
"Epoch 13/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 53.9811 - mae: 5.1831 - val_loss: 161.3696 - val_mae: 10.1498 - learning_rate: 0.0010\n",
"Epoch 14/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 55.9300 - mae: 5.3362 - val_loss: 120.3780 - val_mae: 8.9371 - learning_rate: 0.0010\n",
"Epoch 15/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 54.5672 - mae: 5.2052 - val_loss: 175.7188 - val_mae: 11.8765 - learning_rate: 0.0010\n",
"Epoch 16/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 50.6375 - mae: 4.9938 - val_loss: 120.6923 - val_mae: 9.1595 - learning_rate: 0.0010\n",
"Epoch 17/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 49.6132 - mae: 5.0517 - val_loss: 164.1767 - val_mae: 10.8718 - learning_rate: 0.0010\n",
"Epoch 18/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 28ms/step - loss: 47.0034 - mae: 5.0077 - val_loss: 129.4520 - val_mae: 9.8728 - learning_rate: 0.0010\n",
"Epoch 19/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 45.4484 - mae: 4.9093 - val_loss: 171.7626 - val_mae: 11.5587 - learning_rate: 0.0010\n",
"Epoch 20/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 38.2535 - mae: 4.4474 - val_loss: 157.3609 - val_mae: 11.2681 - learning_rate: 0.0010\n",
"Epoch 21/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 35.8599 - mae: 4.3157 - val_loss: 158.3537 - val_mae: 11.0643 - learning_rate: 0.0010\n",
"Epoch 22/150\n",
"\u001b[1m44/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m━\u001b[0m \u001b[1m0s\u001b[0m 23ms/step - loss: 34.2164 - mae: 4.3049\n",
"Epoch 22: ReduceLROnPlateau reducing learning rate to 0.0005000000237487257.\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 36.1667 - mae: 4.4353 - val_loss: 128.8475 - val_mae: 9.5550 - learning_rate: 0.0010\n",
"Epoch 23/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 33.1296 - mae: 4.2523 - val_loss: 117.9625 - val_mae: 9.1469 - learning_rate: 5.0000e-04\n",
"Epoch 24/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 31.2907 - mae: 4.3110 - val_loss: 131.2014 - val_mae: 9.6215 - learning_rate: 5.0000e-04\n",
"Epoch 25/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 33.1731 - mae: 4.2189 - val_loss: 120.6982 - val_mae: 9.2732 - learning_rate: 5.0000e-04\n",
"Epoch 26/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 29.7279 - mae: 4.0669 - val_loss: 133.3877 - val_mae: 9.9621 - learning_rate: 5.0000e-04\n",
"Epoch 27/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 30.7823 - mae: 4.1318 - val_loss: 134.9048 - val_mae: 9.9791 - learning_rate: 5.0000e-04\n",
"Epoch 28/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 31.4873 - mae: 4.2751 - val_loss: 126.3903 - val_mae: 9.8283 - learning_rate: 5.0000e-04\n",
"Epoch 29/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 31.1747 - mae: 4.2124 - val_loss: 155.8230 - val_mae: 10.9685 - learning_rate: 5.0000e-04\n",
"Epoch 30/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 27ms/step - loss: 26.1791 - mae: 3.8438 - val_loss: 146.5038 - val_mae: 10.5429 - learning_rate: 5.0000e-04\n",
"Epoch 31/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 28.4097 - mae: 4.0506 - val_loss: 139.7865 - val_mae: 10.3443 - learning_rate: 5.0000e-04\n",
"Epoch 32/150\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 23ms/step - loss: 27.9504 - mae: 4.1128\n",
"Epoch 32: ReduceLROnPlateau reducing learning rate to 0.0002500000118743628.\n",
"\u001b[1m46/46\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 26ms/step - loss: 28.7068 - mae: 4.1198 - val_loss: 136.8158 - val_mae: 10.0341 - learning_rate: 5.0000e-04\n",
"Epoch 32: early stopping\n",
"Restoring model weights from the end of the best epoch: 12.\n",
"\n",
"Best epoch: 12\n",
"Best val loss: 107.312881\n"
]
}
],
"source": [
"callbacks = [\n",
" tf.keras.callbacks.EarlyStopping(\n",
" monitor=\"val_loss\", patience=EARLY_STOP_PATIENCE,\n",
" restore_best_weights=True, verbose=1,\n",
" ),\n",
" tf.keras.callbacks.ReduceLROnPlateau(\n",
" monitor=\"val_loss\", factor=0.5, patience=10, verbose=1,\n",
" ),\n",
"]\n",
"\n",
"history = dg_model.fit(\n",
" X_train, y_train,\n",
" validation_data=(X_test, y_test),\n",
" epochs=MAX_EPOCHS, batch_size=BATCH_SIZE,\n",
" callbacks=callbacks, verbose=1,\n",
")\n",
"\n",
"best_epoch = np.argmin(history.history[\"val_loss\"]) + 1\n",
"print(f\"\\nBest epoch: {best_epoch}\")\n",
"print(f\"Best val loss: {min(history.history['val_loss']):.6f}\")"
]
},
{
"cell_type": "markdown",
"id": "b88c6b3a",
"metadata": {},
"source": [
"## 4. Evaluate"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "12650600",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:53.433453Z",
"iopub.status.busy": "2026-02-24T18:15:53.433453Z",
"iopub.status.idle": "2026-02-24T18:15:54.126753Z",
"shell.execute_reply": "2026-02-24T18:15:54.126753Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Dynamic-Graph iTransformer Results (v3):\n",
" MAE = 8.7674\n",
" RMSE = 10.3592\n",
" R² = 0.7090\n",
" MAPE = 23.28%\n",
" Tol.Acc (±2%) = 8.97%\n",
" Tol.Acc (±5%) = 24.14%\n",
"Model saved.\n"
]
}
],
"source": [
"y_pred = dg_model.predict(X_test, verbose=0).flatten()\n",
"metrics = regression_metrics(y_test, y_pred)\n",
"metrics[\"tol_2pct\"] = tolerance_accuracy(y_test, y_pred, 2.0)\n",
"metrics[\"tol_5pct\"] = tolerance_accuracy(y_test, y_pred, 5.0)\n",
"\n",
"print(\"Dynamic-Graph iTransformer Results (v3):\")\n",
"print(f\" MAE = {metrics['MAE']:.4f}\")\n",
"print(f\" RMSE = {metrics['RMSE']:.4f}\")\n",
"print(f\" R² = {metrics['R2']:.4f}\")\n",
"print(f\" MAPE = {metrics['MAPE']:.2f}%\")\n",
"print(f\" Tol.Acc (±2%) = {metrics['tol_2pct']:.2%}\")\n",
"print(f\" Tol.Acc (±5%) = {metrics['tol_5pct']:.2%}\")\n",
"\n",
"dg_model.save(str(v3[\"models_deep\"] / \"dynamic_graph_itransformer.keras\"))\n",
"print(\"Model saved.\")"
]
},
{
"cell_type": "markdown",
"id": "5c7a78fc",
"metadata": {},
"source": [
"## 5. Training Curves"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "77842c13",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:54.128764Z",
"iopub.status.busy": "2026-02-24T18:15:54.128764Z",
"iopub.status.idle": "2026-02-24T18:15:54.402919Z",
"shell.execute_reply": "2026-02-24T18:15:54.402919Z"
}
},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"fig, axes = plt.subplots(1, 2, figsize=(14, 5))\n",
"\n",
"# Loss\n",
"axes[0].plot(history.history[\"loss\"], \"b-\", linewidth=1.5, label=\"Train\")\n",
"axes[0].plot(history.history[\"val_loss\"], \"r-\", linewidth=1.5, label=\"Validation\")\n",
"axes[0].axvline(x=best_epoch-1, color=\"green\", linestyle=\"--\", alpha=0.7, label=f\"Best: {best_epoch}\")\n",
"axes[0].set_xlabel(\"Epoch\"); axes[0].set_ylabel(\"MSE Loss\")\n",
"axes[0].set_title(\"Loss Curve\", fontweight=\"bold\")\n",
"axes[0].legend(); axes[0].grid(True, alpha=0.3)\n",
"\n",
"# MAE\n",
"if \"mae\" in history.history:\n",
" axes[1].plot(history.history[\"mae\"], \"b-\", linewidth=1.5, label=\"Train\")\n",
" axes[1].plot(history.history[\"val_mae\"], \"r-\", linewidth=1.5, label=\"Validation\")\n",
" axes[1].set_ylabel(\"MAE\")\n",
"else:\n",
" axes[1].plot(history.history[\"loss\"], \"b-\", linewidth=1.5, label=\"Train Loss\")\n",
" axes[1].set_ylabel(\"Loss\")\n",
"axes[1].set_xlabel(\"Epoch\")\n",
"axes[1].set_title(\"MAE Metric\", fontweight=\"bold\")\n",
"axes[1].legend(); axes[1].grid(True, alpha=0.3)\n",
"\n",
"plt.suptitle(\"Dynamic-Graph iTransformer — Training (v3)\", fontsize=14, fontweight=\"bold\")\n",
"plt.tight_layout(); save_fig(fig, \"v3_dg_itransformer_training\", directory=v3[\"figures\"]); plt.show()"
]
},
{
"cell_type": "markdown",
"id": "3460b623",
"metadata": {},
"source": [
"## 6. Actual vs Predicted"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "7578a82d",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:54.404452Z",
"iopub.status.busy": "2026-02-24T18:15:54.404452Z",
"iopub.status.idle": "2026-02-24T18:15:54.670860Z",
"shell.execute_reply": "2026-02-24T18:15:54.670860Z"
}
},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"fig, axes = plt.subplots(1, 2, figsize=(14, 5))\n",
"\n",
"# Scatter\n",
"axes[0].scatter(y_test, y_pred, s=8, alpha=0.4, c=\"steelblue\")\n",
"lims = [min(y_test.min(), y_pred.min()), max(y_test.max(), y_pred.max())]\n",
"axes[0].plot(lims, lims, \"r--\", linewidth=1.5)\n",
"axes[0].annotate(f\"R² = {metrics['R2']:.4f}\", xy=(0.05, 0.9), xycoords=\"axes fraction\",\n",
" fontsize=12, fontweight=\"bold\", bbox=dict(boxstyle=\"round\", facecolor=\"lightyellow\"))\n",
"axes[0].set_xlabel(\"Actual SOH (%)\"); axes[0].set_ylabel(\"Predicted SOH (%)\")\n",
"axes[0].set_title(\"Actual vs Predicted\", fontweight=\"bold\")\n",
"axes[0].set_aspect(\"equal\"); axes[0].grid(True, alpha=0.3)\n",
"\n",
"# Residuals\n",
"residuals = y_test - y_pred\n",
"axes[1].scatter(y_pred, residuals, s=8, alpha=0.4, c=\"coral\")\n",
"axes[1].axhline(y=0, color=\"black\", linewidth=1)\n",
"axes[1].set_xlabel(\"Predicted SOH (%)\"); axes[1].set_ylabel(\"Residual (Actual - Predicted)\")\n",
"axes[1].set_title(\"Residual Plot\", fontweight=\"bold\"); axes[1].grid(True, alpha=0.3)\n",
"\n",
"plt.suptitle(\"Dynamic-Graph iTransformer — Predictions (v3)\", fontsize=14, fontweight=\"bold\")\n",
"plt.tight_layout(); save_fig(fig, \"v3_dg_itransformer_predictions\", directory=v3[\"figures\"]); plt.show()"
]
},
{
"cell_type": "markdown",
"id": "cd602cc6",
"metadata": {},
"source": [
"## 7. Learned Adjacency Matrix Visualization\n",
"Visualizing the learned inter-feature adjacency (if exposed by the model)."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "ff82778a",
"metadata": {
"execution": {
"iopub.execute_input": "2026-02-24T18:15:54.672868Z",
"iopub.status.busy": "2026-02-24T18:15:54.672868Z",
"iopub.status.idle": "2026-02-24T18:15:54.677055Z",
"shell.execute_reply": "2026-02-24T18:15:54.677055Z"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Adjacency matrix not directly accessible; model uses softmax(QK^T) for dynamic graphs.\n",
"Architecture relies on implicit feature correlations via attention.\n",
"Results + predictions saved to v3\n"
]
}
],
"source": [
"# Attempt to extract adjacency matrix from DynamicGraphConv layers\n",
"adj_found = False\n",
"for layer in dg_model.layers:\n",
" if hasattr(layer, \"adj\"):\n",
" adj = layer.adj.numpy()\n",
" adj_found = True\n",
" fig, ax = plt.subplots(figsize=(8, 7))\n",
" im = ax.imshow(adj, cmap=\"viridis\", aspect=\"auto\")\n",
" ax.set_xlabel(\"Feature j\"); ax.set_ylabel(\"Feature i\")\n",
" ax.set_title(\"Learned Dynamic Adjacency Matrix (v3)\", fontsize=13, fontweight=\"bold\")\n",
" plt.colorbar(im, ax=ax)\n",
" save_fig(fig, \"v3_dg_adjacency_matrix\", directory=v3[\"figures\"])\n",
" plt.show()\n",
" break\n",
"\n",
"if not adj_found:\n",
" print(\"Adjacency matrix not directly accessible; model uses softmax(QK^T) for dynamic graphs.\")\n",
" print(\"Architecture relies on implicit feature correlations via attention.\")\n",
"\n",
"# Save metrics\n",
"import json\n",
"with open(v3[\"results\"] / \"v3_dg_itransformer_results.json\", \"w\") as f:\n",
" json.dump({k: float(v) for k, v in metrics.items()}, f, indent=2)\n",
"\n",
"# v3: save raw predictions\n",
"np.savez_compressed(\n",
" str(v3[\"results\"] / \"v3_dg_predictions.npz\"),\n",
" y_test=y_test, y_pred=y_pred,\n",
")\n",
"print(\"Results + predictions saved to v3\")"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "venv",
"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": "3.12.10"
}
},
"nbformat": 4,
"nbformat_minor": 5
}