{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "5125745b",
   "metadata": {},
   "source": [
    "# BaseAttentive: Attention Stack Configuration\n",
    "\n",
    "This notebook demonstrates how to configure different attention mechanisms in the decoder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "fd748c45",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:18.786046Z",
     "iopub.status.busy": "2026-04-21T01:46:18.786046Z",
     "iopub.status.idle": "2026-04-21T01:46:22.312639Z",
     "shell.execute_reply": "2026-04-21T01:46:22.312639Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Backend: tensorflow\n"
     ]
    }
   ],
   "source": [
    "# ── v2.2.0 Backend Setup ─────────────────────────────────────────────────────\n",
    "# BASE_ATTENTIVE_BACKEND must be set *before* importing base_attentive.\n",
    "# Choose your installed backend: \"tensorflow\" | \"torch\" | \"jax\" | \"auto\"\n",
    "import os\n",
    "os.environ.setdefault(\"BASE_ATTENTIVE_BACKEND\", \"tensorflow\")\n",
    "os.environ.setdefault(\"KERAS_BACKEND\", os.environ[\"BASE_ATTENTIVE_BACKEND\"])\n",
    "import keras  # initialise Keras 3 backend before base_attentive\n",
    "BACKEND = os.environ[\"BASE_ATTENTIVE_BACKEND\"]\n",
    "print(f\"Backend: {BACKEND}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "b1c9d69a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:22.315975Z",
     "iopub.status.busy": "2026-04-21T01:46:22.314714Z",
     "iopub.status.idle": "2026-04-21T01:46:22.359838Z",
     "shell.execute_reply": "2026-04-21T01:46:22.359084Z"
    }
   },
   "outputs": [],
   "source": [
    "from base_attentive import BaseAttentive"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6cd8aace",
   "metadata": {},
   "source": [
    "## Available Attention Mechanisms\n",
    "\n",
    "1. **Cross-Attention**: Allows decoder to focus on encoder outputs\n",
    "2. **Hierarchical Attention**: Self-attention within decoder for structure finding\n",
    "3. **Memory-Augmented Attention**: Long-term memory retrieval mechanism"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f3f416b6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:22.362030Z",
     "iopub.status.busy": "2026-04-21T01:46:22.362030Z",
     "iopub.status.idle": "2026-04-21T01:46:22.375391Z",
     "shell.execute_reply": "2026-04-21T01:46:22.374221Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "✅ Base configuration prepared\n"
     ]
    }
   ],
   "source": [
    "# Base configuration\n",
    "base_config = {\n",
    "    \"static_input_dim\": 4,\n",
    "    \"dynamic_input_dim\": 8,\n",
    "    \"future_input_dim\": 6,\n",
    "    \"output_dim\": 1,\n",
    "    \"forecast_horizon\": 24,\n",
    "}\n",
    "\n",
    "print(\"✅ Base configuration prepared\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "943a583b",
   "metadata": {},
   "source": [
    "## Configuration 1: Cross Attention Only\n",
    "\n",
    "Simple encoder-decoder attention without self-attention."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "cb08105c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:22.377511Z",
     "iopub.status.busy": "2026-04-21T01:46:22.377511Z",
     "iopub.status.idle": "2026-04-21T01:46:22.685651Z",
     "shell.execute_reply": "2026-04-21T01:46:22.684445Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\projects\\base-attentive\\src\\base_attentive\\core\\base_attentive.py:148: DeprecatedParameterWarning: BaseAttentive: 'static_input_dim' is deprecated since 2.1.0 and will be removed in 3.0.0. Use 'static_dim' instead.\n",
      "  resolved = resolve_deprecated_kwargs(\n",
      "D:\\projects\\base-attentive\\src\\base_attentive\\core\\base_attentive.py:148: DeprecatedParameterWarning: BaseAttentive: 'dynamic_input_dim' is deprecated since 2.1.0 and will be removed in 3.0.0. Use 'dynamic_dim' instead.\n",
      "  resolved = resolve_deprecated_kwargs(\n",
      "D:\\projects\\base-attentive\\src\\base_attentive\\core\\base_attentive.py:148: DeprecatedParameterWarning: BaseAttentive: 'future_input_dim' is deprecated since 2.1.0 and will be removed in 3.0.0. Use 'future_dim' instead.\n",
      "  resolved = resolve_deprecated_kwargs(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "🔍 Model 1: Cross Attention Only\n",
      "  Attention Stack: ['cross']\n",
      "  Complexity: Low\n",
      "  Best for: Simple encoder-decoder interactions\n"
     ]
    }
   ],
   "source": [
    "cross_only = {\n",
    "    **base_config,\n",
    "    \"architecture_config\": {\n",
    "        \"decoder_attention_stack\": [\"cross\"],\n",
    "    },\n",
    "    \"name\": \"CrossAttentionOnly\",\n",
    "}\n",
    "\n",
    "model1 = BaseAttentive(**cross_only)\n",
    "\n",
    "print(\"🔍 Model 1: Cross Attention Only\")\n",
    "print(\n",
    "    f\"  Attention Stack: {model1.architecture_config['decoder_attention_stack']}\"\n",
    ")\n",
    "print(\"  Complexity: Low\")\n",
    "print(\"  Best for: Simple encoder-decoder interactions\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1f7f08f",
   "metadata": {},
   "source": [
    "## Configuration 2: Cross + Hierarchical Attention\n",
    "\n",
    "Encoder-decoder attention followed by self-attention within decoder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "63511b9d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:22.687950Z",
     "iopub.status.busy": "2026-04-21T01:46:22.687950Z",
     "iopub.status.idle": "2026-04-21T01:46:22.874205Z",
     "shell.execute_reply": "2026-04-21T01:46:22.872973Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "🔍 Model 2: Cross + Hierarchical Attention\n",
      "  Attention Stack: ['cross', 'hierarchical']\n",
      "  Complexity: Medium\n",
      "  Best for: Multi-horizon forecasts with structural patterns\n"
     ]
    }
   ],
   "source": [
    "cross_hierarchical = {\n",
    "    **base_config,\n",
    "    \"architecture_config\": {\n",
    "        \"decoder_attention_stack\": [\"cross\", \"hierarchical\"],\n",
    "    },\n",
    "    \"name\": \"CrossHierarchical\",\n",
    "}\n",
    "\n",
    "model2 = BaseAttentive(**cross_hierarchical)\n",
    "\n",
    "print(\"🔍 Model 2: Cross + Hierarchical Attention\")\n",
    "print(\n",
    "    f\"  Attention Stack: {model2.architecture_config['decoder_attention_stack']}\"\n",
    ")\n",
    "print(\"  Complexity: Medium\")\n",
    "print(\n",
    "    \"  Best for: Multi-horizon forecasts with structural patterns\"\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9bf3dc50",
   "metadata": {},
   "source": [
    "## Configuration 3: Full Stack (All Attention Types)\n",
    "\n",
    "Cross, Hierarchical, and Memory-Augmented attention combined."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "37bcad1e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:22.876287Z",
     "iopub.status.busy": "2026-04-21T01:46:22.876287Z",
     "iopub.status.idle": "2026-04-21T01:46:23.062878Z",
     "shell.execute_reply": "2026-04-21T01:46:23.061398Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "🔍 Model 3: Full Attention Stack\n",
      "  Attention Stack: ['cross', 'hierarchical', 'memory']\n",
      "  Memory Size: 128\n",
      "  Complexity: High\n",
      "  Best for: Complex long-term dependencies, recall mechanisms\n"
     ]
    }
   ],
   "source": [
    "full_stack = {\n",
    "    **base_config,\n",
    "    \"architecture_config\": {\n",
    "        \"decoder_attention_stack\": [\n",
    "            \"cross\",\n",
    "            \"hierarchical\",\n",
    "            \"memory\",\n",
    "        ],\n",
    "    },\n",
    "    \"memory_size\": 128,  # Memory buffer size\n",
    "    \"name\": \"FullAttentionStack\",\n",
    "}\n",
    "\n",
    "model3 = BaseAttentive(**full_stack)\n",
    "\n",
    "print(\"🔍 Model 3: Full Attention Stack\")\n",
    "print(\n",
    "    f\"  Attention Stack: {model3.architecture_config['decoder_attention_stack']}\"\n",
    ")\n",
    "print(f\"  Memory Size: {model3.memory_size}\")\n",
    "print(\"  Complexity: High\")\n",
    "print(\n",
    "    \"  Best for: Complex long-term dependencies, recall mechanisms\"\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ac83c4d",
   "metadata": {},
   "source": [
    "## Configuration 4: Custom Ordering\n",
    "\n",
    "You can specify the order of attention mechanisms."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "ef15f3c2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:23.064752Z",
     "iopub.status.busy": "2026-04-21T01:46:23.064752Z",
     "iopub.status.idle": "2026-04-21T01:46:23.250810Z",
     "shell.execute_reply": "2026-04-21T01:46:23.249736Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "🎨 Model 4: Custom Attention Order\n",
      "  Attention Stack: ['memory', 'cross', 'hierarchical']\n",
      "  Note: Order matters! First layer uses input, others chain the output\n"
     ]
    }
   ],
   "source": [
    "# Memory first, then cross attention (unusual but possible)\n",
    "custom_order = {\n",
    "    **base_config,\n",
    "    \"architecture_config\": {\n",
    "        \"decoder_attention_stack\": [\n",
    "            \"memory\",\n",
    "            \"cross\",\n",
    "            \"hierarchical\",\n",
    "        ],\n",
    "    },\n",
    "    \"memory_size\": 100,\n",
    "    \"name\": \"CustomOrderAttention\",\n",
    "}\n",
    "\n",
    "model4 = BaseAttentive(**custom_order)\n",
    "\n",
    "print(\"🎨 Model 4: Custom Attention Order\")\n",
    "print(\n",
    "    f\"  Attention Stack: {model4.architecture_config['decoder_attention_stack']}\"\n",
    ")\n",
    "print(\n",
    "    \"  Note: Order matters! First layer uses input, others chain the output\"\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd302d0a",
   "metadata": {},
   "source": [
    "## Comparison: Attention Configurations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "06a93acd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:23.253476Z",
     "iopub.status.busy": "2026-04-21T01:46:23.253476Z",
     "iopub.status.idle": "2026-04-21T01:46:23.266406Z",
     "shell.execute_reply": "2026-04-21T01:46:23.265252Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       Configuration                               Stack Parameters Training Time       Best Use Case\n",
      "          Cross Only                           ['cross']        Low          Fast  Simple forecasting\n",
      "Cross + Hierarchical           ['cross', 'hierarchical']     Medium        Medium Structured horizons\n",
      "          Full Stack ['cross', 'hierarchical', 'memory']       High          Slow    Complex patterns\n",
      "        Custom Order ['memory', 'cross', 'hierarchical']       High          Slow   Specialized tasks\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "attention_comparison = {\n",
    "    \"Configuration\": [\n",
    "        \"Cross Only\",\n",
    "        \"Cross + Hierarchical\",\n",
    "        \"Full Stack\",\n",
    "        \"Custom Order\",\n",
    "    ],\n",
    "    \"Stack\": [\n",
    "        \"['cross']\",\n",
    "        \"['cross', 'hierarchical']\",\n",
    "        \"['cross', 'hierarchical', 'memory']\",\n",
    "        \"['memory', 'cross', 'hierarchical']\",\n",
    "    ],\n",
    "    \"Parameters\": [\"Low\", \"Medium\", \"High\", \"High\"],\n",
    "    \"Training Time\": [\"Fast\", \"Medium\", \"Slow\", \"Slow\"],\n",
    "    \"Best Use Case\": [\n",
    "        \"Simple forecasting\",\n",
    "        \"Structured horizons\",\n",
    "        \"Complex patterns\",\n",
    "        \"Specialized tasks\",\n",
    "    ],\n",
    "}\n",
    "\n",
    "df = pd.DataFrame(attention_comparison)\n",
    "print(df.to_string(index=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c71831b",
   "metadata": {},
   "source": [
    "## Attention Mechanisms Explained"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "921f7570",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:23.269557Z",
     "iopub.status.busy": "2026-04-21T01:46:23.269557Z",
     "iopub.status.idle": "2026-04-21T01:46:23.282187Z",
     "shell.execute_reply": "2026-04-21T01:46:23.280760Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Cross-Attention:\n",
      "──────────────────────────────────────────────────\n",
      "  Role................ Queries (decoder) attend to Keys/Values (encoder)\n",
      "  Purpose............. Bridge encoder-decoder, extract relevant context\n",
      "  When to use......... Always (fundamental for seq2seq)\n",
      "  Complexity.......... O(n*m) where n=decoder len, m=encoder len\n",
      "\n",
      "Hierarchical Attention:\n",
      "──────────────────────────────────────────────────\n",
      "  Role................ Self-attention within decoder outputs\n",
      "  Purpose............. Find patterns and structure in predictions\n",
      "  When to use......... When temporal structure matters (seasonality, trends)\n",
      "  Complexity.......... O(n²) where n=decoder length\n",
      "\n",
      "Memory-Augmented Attention:\n",
      "──────────────────────────────────────────────────\n",
      "  Role................ Retrieve from external memory buffer\n",
      "  Purpose............. Store and recall long-term patterns\n",
      "  When to use......... When data has recurring patterns or long-term deps\n",
      "  Complexity.......... O(n*k) where k=memory size\n"
     ]
    }
   ],
   "source": [
    "mechanisms = {\n",
    "    \"Cross-Attention\": {\n",
    "        \"Role\": \"Queries (decoder) attend to Keys/Values (encoder)\",\n",
    "        \"Purpose\": \"Bridge encoder-decoder, extract relevant context\",\n",
    "        \"When to use\": \"Always (fundamental for seq2seq)\",\n",
    "        \"Complexity\": \"O(n*m) where n=decoder len, m=encoder len\",\n",
    "    },\n",
    "    \"Hierarchical Attention\": {\n",
    "        \"Role\": \"Self-attention within decoder outputs\",\n",
    "        \"Purpose\": \"Find patterns and structure in predictions\",\n",
    "        \"When to use\": \"When temporal structure matters (seasonality, trends)\",\n",
    "        \"Complexity\": \"O(n²) where n=decoder length\",\n",
    "    },\n",
    "    \"Memory-Augmented Attention\": {\n",
    "        \"Role\": \"Retrieve from external memory buffer\",\n",
    "        \"Purpose\": \"Store and recall long-term patterns\",\n",
    "        \"When to use\": \"When data has recurring patterns or long-term deps\",\n",
    "        \"Complexity\": \"O(n*k) where k=memory size\",\n",
    "    },\n",
    "}\n",
    "\n",
    "for name, details in mechanisms.items():\n",
    "    print(f\"\\n{name}:\")\n",
    "    print(\"─\" * 50)\n",
    "    for key, value in details.items():\n",
    "        print(f\"  {key:.<20} {value}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "92e509b3",
   "metadata": {},
   "source": [
    "## Training Attention Configurations\n",
    "\n",
    "We train two representative stacks on the same data to empirically compare their fit quality:\n",
    "\n",
    "- **Cross Only** — lightweight baseline\n",
    "- **Full Stack** (cross + hierarchical + memory) — maximum expressiveness"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d3e8be1b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:23.284491Z",
     "iopub.status.busy": "2026-04-21T01:46:23.284491Z",
     "iopub.status.idle": "2026-04-21T01:46:23.297410Z",
     "shell.execute_reply": "2026-04-21T01:46:23.296302Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "static:(64, 4) dynamic:(64, 16, 8) future:(64, 24, 6) target:(64, 24, 1)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import keras\n",
    "\n",
    "np.random.seed(0)\n",
    "N, T, H = 64, 16, 24\n",
    "S_DIM, D_DIM, F_DIM = 4, 8, 6\n",
    "\n",
    "# ── Synthetic data ────────────────────────────────────────────────────\n",
    "t_past   = np.linspace(0, 4*np.pi, T)\n",
    "t_future = np.linspace(4*np.pi, 6*np.pi, H)\n",
    "static_d  = np.random.randn(N, S_DIM).astype('float32')\n",
    "dynamic_d = (np.tile(np.sin(t_past),   (N,1))[:,:,None]\n",
    "             + 0.1*np.random.randn(N,T,D_DIM)).astype('float32')\n",
    "future_d  = (np.tile(np.cos(t_future), (N,1))[:,:,None]\n",
    "             + 0.1*np.random.randn(N,H,F_DIM)).astype('float32')\n",
    "target_d  = (np.tile(np.sin(t_future), (N,1))[:,:,None]\n",
    "             + 0.1*np.random.randn(N,H,1)).astype('float32')\n",
    "\n",
    "print(f'static:{static_d.shape} dynamic:{dynamic_d.shape} future:{future_d.shape} target:{target_d.shape}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4f7cf714",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:46:23.299622Z",
     "iopub.status.busy": "2026-04-21T01:46:23.299622Z",
     "iopub.status.idle": "2026-04-21T01:47:28.681719Z",
     "shell.execute_reply": "2026-04-21T01:47:28.681719Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\projects\\base-attentive\\src\\base_attentive\\core\\base_attentive.py:148: DeprecatedParameterWarning: BaseAttentive: 'static_input_dim' is deprecated since 2.1.0 and will be removed in 3.0.0. Use 'static_dim' instead.\n",
      "  resolved = resolve_deprecated_kwargs(\n",
      "D:\\projects\\base-attentive\\src\\base_attentive\\core\\base_attentive.py:148: DeprecatedParameterWarning: BaseAttentive: 'dynamic_input_dim' is deprecated since 2.1.0 and will be removed in 3.0.0. Use 'dynamic_dim' instead.\n",
      "  resolved = resolve_deprecated_kwargs(\n",
      "D:\\projects\\base-attentive\\src\\base_attentive\\core\\base_attentive.py:148: DeprecatedParameterWarning: BaseAttentive: 'future_input_dim' is deprecated since 2.1.0 and will be removed in 3.0.0. Use 'future_dim' instead.\n",
      "  resolved = resolve_deprecated_kwargs(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross Only                                     val_MSE=0.0195\n",
      "Full Stack (cross+hierarchical+memory)         val_MSE=0.0202\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "from base_attentive import BaseAttentive\n",
    "\n",
    "def build_and_train(stack, label, memory_size=None, epochs=12):\n",
    "    cfg = {'decoder_attention_stack': stack}\n",
    "    kw  = dict(memory_size=memory_size) if memory_size else {}\n",
    "    m   = BaseAttentive(\n",
    "        static_input_dim=S_DIM, dynamic_input_dim=D_DIM,\n",
    "        future_input_dim=F_DIM, output_dim=1, forecast_horizon=H,\n",
    "        architecture_config=cfg, **kw,\n",
    "    )\n",
    "    _ = m([static_d, dynamic_d, future_d])  # build weights\n",
    "    m.compile(optimizer=keras.optimizers.Adam(1e-3), loss='mse', metrics=['mae'])\n",
    "    h = m.fit([static_d, dynamic_d, future_d], target_d,\n",
    "              epochs=epochs, batch_size=16, validation_split=0.2, verbose=0)\n",
    "    print(f'{label:45s}  val_MSE={h.history[\"val_loss\"][-1]:.4f}')\n",
    "    return m, h\n",
    "\n",
    "print('Training...')\n",
    "m_cross, h_cross = build_and_train(['cross'],                          'Cross Only')\n",
    "m_full,  h_full  = build_and_train(['cross','hierarchical','memory'],  'Full Stack (cross+hierarchical+memory)', memory_size=64)\n",
    "print('Done.')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10ec7592",
   "metadata": {},
   "source": [
    "## Plot 1 — Convergence Comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "9cc99016",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:47:28.684951Z",
     "iopub.status.busy": "2026-04-21T01:47:28.683928Z",
     "iopub.status.idle": "2026-04-21T01:47:28.836693Z",
     "shell.execute_reply": "2026-04-21T01:47:28.835518Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(9, 4))\n",
    "for h, label, color, ls in [\n",
    "    (h_cross, 'Cross Only — train',     'steelblue',   '-'),\n",
    "    (h_cross, 'Cross Only — val',        'steelblue',   '--'),\n",
    "    (h_full,  'Full Stack — train',      'darkorange',  '-'),\n",
    "    (h_full,  'Full Stack — val',         'darkorange',  '--'),\n",
    "]:\n",
    "    key = 'loss' if 'train' in label else 'val_loss'\n",
    "    ax.plot(h.history[key], label=label, color=color, linestyle=ls, linewidth=2)\n",
    "\n",
    "ax.set_title('Convergence: Cross Only vs Full Attention Stack', fontsize=13)\n",
    "ax.set_xlabel('Epoch')\n",
    "ax.set_ylabel('MSE')\n",
    "ax.legend(fontsize=9)\n",
    "ax.grid(True, alpha=0.3)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "628c4ad4",
   "metadata": {},
   "source": [
    "## Plot 2 — Forecast Quality per Configuration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ccbe2a63",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:47:28.838702Z",
     "iopub.status.busy": "2026-04-21T01:47:28.838702Z",
     "iopub.status.idle": "2026-04-21T01:47:31.275658Z",
     "shell.execute_reply": "2026-04-21T01:47:31.273570Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross Only  MAE: 0.1053\n",
      "Full Stack  MAE: 0.1090\n"
     ]
    }
   ],
   "source": [
    "test_sl = slice(-10, None)\n",
    "pred_cross = m_cross.predict([static_d[test_sl], dynamic_d[test_sl], future_d[test_sl]], verbose=0)\n",
    "pred_full  = m_full.predict( [static_d[test_sl], dynamic_d[test_sl], future_d[test_sl]], verbose=0)\n",
    "\n",
    "sample = 0\n",
    "steps  = np.arange(1, H + 1)\n",
    "actual = target_d[test_sl][sample, :, 0]\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 4))\n",
    "ax.plot(steps, actual,                   label='Actual',       color='black',      linewidth=2.5)\n",
    "ax.plot(steps, pred_cross[sample, :, 0], label='Cross Only',   color='steelblue',  linewidth=2, linestyle='--')\n",
    "ax.plot(steps, pred_full [sample, :, 0], label='Full Stack',   color='darkorange', linewidth=2, linestyle=':')\n",
    "\n",
    "ax.set_title('Forecast vs Actual — Attention Stack Configurations', fontsize=13)\n",
    "ax.set_xlabel('Forecast step')\n",
    "ax.set_ylabel('Value')\n",
    "ax.legend()\n",
    "ax.grid(True, alpha=0.3)\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "mae_cross = float(np.mean(np.abs(pred_cross - target_d[test_sl])))\n",
    "mae_full  = float(np.mean(np.abs(pred_full  - target_d[test_sl])))\n",
    "print(f'Cross Only  MAE: {mae_cross:.4f}')\n",
    "print(f'Full Stack  MAE: {mae_full:.4f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "346ba5a0",
   "metadata": {},
   "source": [
    "## Plot 3 — MAE Bar Chart"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "938d0ab1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-04-21T01:47:31.278272Z",
     "iopub.status.busy": "2026-04-21T01:47:31.278272Z",
     "iopub.status.idle": "2026-04-21T01:47:31.384868Z",
     "shell.execute_reply": "2026-04-21T01:47:31.383766Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeoAAAGGCAYAAAC0W8IbAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjgsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvwVt1zgAAAAlwSFlzAAAPYQAAD2EBqD+naQAARlpJREFUeJzt3QmcTfX/P/D3bMY+1iF7luyG7CqjIkTSZikRUvZBFJGhhBKy5qtfaBMpJKQkpJC9IkQUWcaWnTFmzv/xevc/tzt37p25d9yZe+6d1/PxOMw995xzzzn3nPs+nz3IMAxDiIiIyJKCfb0DRERE5BoDNRERkYUxUBMREVkYAzUREZGFMVATERFZGAM1ERGRhTFQExERWRgDNRERkYUxUBMREVkYAzVlmFGjRklQUJD8+eefPMtegHP5zDPP8Fxmkett165dcv/990v+/Pn1uHB8ODbz70CC6xrHRc4xUPuJdevW6YWMqW/fvk6XOXXqlGTLlk2XadKkicttvfTSS7pMhQoVXC4zb9482+c5m8qXLy9W9tVXX+l+hoSEyJEjR5wus3TpUpc/eKm9l1HOnz+vn4nv2qoSExPlww8/lLvvvluKFi0q2bNnlxIlSsi9994rI0eOlPj4eNuyOA4cD47Ln1y9elXefvttueeee6RAgQISFhYmRYoUkQcffFDvi5s3b2b4PuAzHnvsMTlw4IC89tpres4fffRR8Wc4dzivlA7o65usb+3ateiT3ciePbuRP39+4/r16ymWeeutt4zQ0FCdoqOjnW4nISHBKFq0qFGuXDnd3rp165wuN3fuXH2/f//+xocffphiWrp0aZr7HBsbq9s4fPiwkdkee+wxo2TJkka2bNmMUaNGOV2mS5cuun+evpdRcJ7wmThvzly7ds24ceOG4Uvt2rXTfbzrrruMN99803j33XeNkSNHGq1btzbCw8ON06dP++z798bnHThwwLjjjjt0O02bNtVjnDNnjjF+/Hh9jflDhgwxMtr+/fv1syZOnJhsflJSkl4HuI/9DX6TSpcu7fQ9XNc4LnIuND3BnXznkUcekU8++US++OILadeuXbL35s6dq0/9a9ascbn+ihUr5OTJk7pMx44dZc6cORIdHe1yeaQqHn/8cfEnp0+flmXLlskrr7wiO3fu1Cd5pPb8PWsNqVdf2r59u3z66ad6DS5evDjF+2fPnpW8efOKv7p27Zq0bt1aDh06JJ9//nmKFCxyorZu3apTRsM9CkjR28M17OvrALkqyDnJmTOn17aJXAtM5IKLAE4WTVFPmDDBqFGjhtGiRYtk7//000/6/pIlS4xcuXK5TFE/9NBDRtmyZfXJfODAgUbOnDmNCxcuuExRL1q06JZTOLt37zb69etnFClSRHME6tWrZ3z77be25eLj441ChQoZjRo1crodpGqwnfXr17v1uUiFBAUFacrqiy++0HVXr16dbBmcH8x3nHDcqb1nOn78uNGzZ09NtYeFhRm33Xab0aNHDyMuLs7pOdi3b58xbNgwo3jx4prKx3e4YsWKFN+v42SfAsFrpPQdIVVbq1YtPbd58+Y1mjVrZmzYsCHFcub6GzduNBo3bqzffYECBYzu3bsbly5dSvO8fvLJJ7qNadOmpbmsmSPhOJm5BceOHTMGDRpkREVFGfny5dPUeOXKlTXlevPmzRTbwzXyxhtv6PI5cuTQ46xdu3ayfXGWosa2nn/+eb0esH5qpk6dquu/9NJLhidwz+HaxfnEvYe/neU44bvEtbV3717jwQcfNHLnzq3HgdyfEydO2JZzdf3huFzluly5ckXvZ+SW4TqoX7++3mPOcobM/XBkXoP217n5O4D759VXX9XfDuTYmct8/fXXmsty++236+dGRETo9eeYU4fPdHZM+MzUcrB+/vlno23btnqdmtcIvkfHa8Rc//z583pfFi5cWJfHd7F582bD3zFF7Ye6desmgwYNkmPHjknx4sV1HlLGkZGRmiJI7SkdZbcjRoywVUyaPHmyLFiwQJ577jmn61y6dEnOnDmTYn6OHDkkV65cbu1v586dtawYKRJs73//+5+0aNFC96Vp06Zart6lSxeZOHGi7N+/XypWrJhsfRzbHXfcIY0bN3br88xcgjJlymj5Kc4L5uGzTMOHD5ekpCTZsGGDlv+ZGjVqpOfU1XuAMu+GDRvKjRs3pHv37lKuXDk5ePCgvPPOO7J27VrZtm2bREREJNsnHB9SDIMHD9b1UFbXtm1b+f3333U/K1eurN/FwIEDNcVqpuZy586d6rHinL755ptSr149GTt2rJ7f2bNna5kxcl2Qw+JYQQnXSNeuXeXJJ5/UcuT33ntPgoODdb3U4Dhh0aJF8tRTT2klJ1eef/55uXjxoixZskSPq1ChQjq/Ro0a+v8vv/yiqXIcK7abkJAgq1atkqFDh2qKFteICeerefPmuq8PPPCAdOrUSVOVv/76q27DVZ0NpJCRa7Ry5Ur54IMPdL3UfPbZZ/q/q3vBmZkzZ0qfPn2kUqVKmmsDyMHBd4tjcNwW7lnUH8FxT5gwQX7++WddDufqm2++sV2bd911l36fWB+5WlC4cGHNLXLmiSee0OPE5+I6P3z4sH7G7bffLt6A6xbfUY8ePTTXxLxHcaznzp3Texz3Go7v//7v/7QSHO4Fc99xvQ8bNkx/S3A9mHDdu4L7CPdxWFiYnmPUifjyyy/1msd5+/jjj1Osg+sE5wnfBXJ4Jk2aJK1atdLzkSdPHvFbvn5SIM9T1GfOnNFU2euvv67vXb16VZ9kX3jhBX3tKkWN1ApSFocOHbLNq1mzpqZwHZlP0q6mPn36pLnPZgoH20eKyHT06FHdx0qVKqUok3Ms//vhhx90flqpIROenh1TBQMGDNCn/XPnznmljLpNmzb6xI7jsLd161YjJCQkWWrHPAetWrXSXAzTli1bdP7QoUPdLqN2TFEjlY7vE+XF9ucXqVVcD0jF2Kc8sD6Wd0xhIHWHVJI7qWrkyGA7SD2izHb48OHGsmXLNEXnSZkxrln782Hq1KmTERwcrDkWJnz32A5yJBwlJiY6/byzZ88aDRs21FQrUn3uQKoNKVx34XrCdYz6Hva5UvgbKU989j///JMiVblw4cJk2+ndu7ct1yW11K2rawQ5M5j37LPPJlvWnO+NFDXK7Z19x5cvX04x7+TJk0bBggWNli1bul1G7ex+Q2oY99PPP/9sm4dr5oknntBl7XPlzPV79eqVbBuffvqpzp81a5bhz1jr2w8VLFhQ2rRpo0+zgFTFhQsXNKWdGqQq8YRr/5SNVPWWLVtkz549TtfBk+nq1atTTP369XN7f5FKRKrZhCdvpMj27dsne/fu1XlIMePpGSkf+1q1SO2FhoZqitQdWB4pfftydRzj9evXZf78+XKrcJ6XL1+u5x+pOqQQzAkpY9SGN1NG9mJiYpKVkdetW1dTy6jVm15IMSP+vvjii8nOb7FixTTF/Ndff2kZvT3kBNSvXz/ZvPvuu0/PuTvNmlB2O23aNKlWrZqmcF9//XU9F0jtIEfEXciRMc8HUsxIleEcIkWE3AykpkxIOSH1bqZY7SEnwBGOGylSpMzXr1+vqXB3IFXrSaoL98GVK1ekf//+ycrm8TfmXb58Wb799ttk6+C7caxbgvMP6b0WkMoE5LLZQ25KailWT/Tq1ctpmbR9rhqOF6lY5J7hGvvpp5/S/XlowbJx40a9tmr8/1wYwDWDHAdAbo2z3xpvnlurYKD2U/ghxsX3ww8/aABG1meVKlVcLo9sXGSzIlsM2bTmhBsKP3YIcM5Ur15d13GcHLOnU+Psx8LcV/yYmpDNFxcXp4EQkI2LykvIqkXzmLTgRxPZ+MhaRDa/eYz4MUEAdXWMnkDWPAIJtoUsNscJ7+MYHJUtW9bpAxd+2NIL2XlQtWrVFO+Z8+zPb2r7Ae7sC7IhkdWMH2EENlxXyNLEAwOyR1HR0R14MBgzZow+oOGBB/uA8/f000/r+//8849tWVznyFp2txLVQw89pMEa98add94p7kKAxTVn5fPvaj9wDztrMunJfZoafE/O/PHHH9KhQwd9kMJDDoo48D0iG97+O/RUaue2cuXKeryO59bZ+b3Vc2sVLKP2U0h5oCx19OjRWhaE8tHUmEEKqRJnKZOPPvpI3njjDZ/WvES7UaREsK8oa1u4cKEG32effdat9VF2ih9a1GzH5AzKaGvWrJnuffw3B1m0vNNVKh+pRUdIZaS2vcziaj/Ssy84TrSnxoQycaRc8d2hXDgtSP0hZd6+fXtNIaEeAa69HTt2aBkkHobSC2XvKPfFgwAeYp2lup1BLsH333+vAcBZQLXa+XfkbqsGV8ul1j7cWWoaKWjUG8E9OmDAAH2oR7DG+R43bpx89913ktlCLHKfeRsDtZ/CBYkKHLgh8IOZ2o8jghcqyjRr1sxpRRlU7EGnCmjShGDpbcjejoqKSjbvt99+0//tfxDDw8P1mKZOnSrHjx/XH308jKDimTvwo4ysxSlTpqR4D9mr2Da2iQCR1g+bq/eQasF72J595TRv8LT5mHnuUGxhVvRK7fxmpAYNGuj/qEzkzvGgkh5+5JEDYg85IM5ScygmQZMgXCNpQYU0nA8UCSD4vP/++6kGSBOufQRqVIZCRS5Pzj8qT/nq/KPIBQ82yHlwzL1CDo8jNPlCUYMjZynU1KCJJ+5T3HfI4bOHCqu3cn2bxXPOiuT27dunx5tZ17YVMOvbj/Xs2VNiY2Nl1qxZqbZfxY8hnnqxPMpuHSf8sOGJGTdcRkAtTwQ2099//63lxciWc/xhQa1StNNEqmrz5s1avuzOjyyy9ZENix9bZ8eIVBbK5/G5Zu9ZZo1qZz9art5DVhrK/lAvAPvn7MndVc3ctKS2P86g/A4/fqg9jBq5phMnTmib+tKlS0utWrXEWxAInAVSsyc3sC9+Se148J06pnJwjdrXCDahPgOyUZFCdjelNGTIEN0Wyrfx3bvTmxhybnBNvvXWW1r+76otOWp6Ax58UayCBz/7LHP8jXk4fiyT0ZDVD47nDtnPZh0QZw8+9g9VuCdmzJjh0eea96Xjd4A6Gs7Kp3E+8D26k7pFDgtaWaD8fffu3bb5WBeJE0Ct9qyCKWo/VqpUKbe6uUQqEoHYVcoU77Vs2VJ/bO2bfAGCHypiOYMfUHeekvEjiSCJVD9+xPBggaYzSDk7QuBGViqy4rHttCrImcyHjNRyBPAeKkChEgrK1ZAKnD59uvTu3VubcCDrFWX2eJpP7T0UM2AfkSJEKh3BEE/4SJHgBx7z0tP9KB4CkGLHgxVShCiXRyAwf4gdIaggIKF5FvYF2chm8yxkSyJIufOQ4y40icFnoNIf6gGgUiCCK36UUZcA2Z72xSpmKhsPXbhWUMaM7GVMeHhC9jS2h5wJlOvjOzTLFB0r4uEHG4EanY0gix3bQmoLKUbHClsmZMeikh3K1HEN4rymVrSD+wD1I/B9o+gFn4NAi33CwxeKmL7++mtNqUO+fPn03KPpEK4Nsx92VPLEAw2Oz7GZXkbAgyOKwt59912tkGc2z8J1gIpYyDGzh/OBc4Hl8PCOh2jkcHjagYnZjewLL7ygFRFxPaBoCdtCNjiaz9nD9YDzi89HEMa1icpeCMrOIGcM19o999xja56F9fEd4OHLMRcjoPm62jl53jwrLfbNs9DZCNZ79NFHU11n/vz5upzZ5Cut5lmY0urG0L7Dk759+2qHJ+iEoG7dusY333zjcr0PPvhA17vvvvsMd6AJEjocQZMp++Y6jv7++29tnoQOGQDLokkbOiFBkyD7pimpvQfoKnPw4MFGhQoV9JjQHKpatWra5eqePXvcaqLkrJkMOq4xO89wt8OT2bNnazM77EeePHm02dT333+fYjlX65vftdn5hCvozAWdyaCzHewXmrzhM8uXL28899xz2v2mIzStQmcYaP5l36wITX1w/kqVKmXbxrhx47TJjbNmSehecsyYMUaVKlVs57tOnTrGjBkz0jzXOD/43tGszr4ZmyvYt0mTJmmzN3TGgn2PjIzUZmy4Nh0721i8eLE2BcN3hgl/oxMUR540i/KkeZbZTComJkb30+xUaM2aNdqZCjqIcTRv3jxtcoXOesqUKaPfE5Z31TzL1bWBplPNmzfX84TmaDg+XHvOmlvhvHbr1k330byn0urwZNeuXcbDDz+s3SajSSqadKbW4Ykzrq57fxKEf3z9sEBkD6kzpLSQTe1OxSQicg4pWxSLIKub/BfLqMlyUFaGZh7+PloQUWZBUZIjtHxA+W5mlJNTxmIZNVkCOjhALVKUiaPmLSqMuFPDl4hEXn31Ve3cBs3kUC6OsmKzzB91BMi/MeubLAGVvPAjgwo6qOiFimYcTYfIPajhPX78eG0Wht7z0AQLFbXQ7NLqY8dT2hioiYiILIxl1ERERBbGQE1ERGRhrEzmBDqvQNd46MDB024diYiI0oKW0eigCN0ep9UfPQO1EwjSJUuWTPNEExER3YqjR49qr26pYaB2whyTFicwtT60iYiI0gPDxCJB6M4Y6AzUTpjZ3QjSDNRERJRR3CleZWUyIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC/N5oJ4xY4aUKVNGsmfPLvXr15ctW7a4XHbPnj3y2GOP6fJBQUHy9ttvp1hm3LhxUrduXcmTJ49ERkZK27ZtZf/+/Rl8FERERAEYqBcuXCiDBg2S2NhY2bFjh0RFRUnz5s3l1KlTTpe/evWqlC1bVsaPHy9FixZ1usz69eulT58+snnzZlm9erUkJCTIAw88IFeuXMngoyEiIvK+IMMwDPERpKCR+p0+fbq+TkpKkpIlS0q/fv1k6NChqa6LVPWAAQN0Ss3p06c1ZY0A3rhxY7f26+LFixIRESEXLlyQvHnzenBERERE3o0zPktR37hxQ7Zv3y5Nmzb9b2eCg/X1pk2bvPY5OAlQoEABr22TiIgos4SKj5w5c0YSExOlSJEiyebj9b59+7zyGUihI8V91113SbVq1VwuFx8fr5P9k465PiYiIiJv8iS2+CxQZwaUVe/evVt++OGHVJdDBbTRo0c7zTa/fv16Bu4hERFlRZcuXbJ+oC5UqJCEhIRIXFxcsvl47aqimCf69u0ry5cvl++//15KlCiR6rLDhg3TSm32KWqUlRcuXJhl1ERE5HVo6WT5QJ0tWzapXbu2rFmzRptQmVkBeI0gm16oG4fKaEuWLJF169bJ7bffnuY64eHhOjlCmTkmIiIib/Iktvg06xup2C5dukidOnWkXr162i4azai6du2q73fu3FmKFy+uWdNmBbTffvvN9vexY8dk165dkjt3bilfvrwtu3v+/PnyxRdfaFvqkydP6nzUrsuRI4fPjpWIiMjvmmcBmmZNmDBBA2rNmjVl6tSp2mwLmjRpos2w5s2bp6///PNPpynk6OhoTT0DOkJxZu7cufLMM8+4tU9snkVERBnJkzjj80BtRQzUREQkWb0dNREREaWNgZqIiMjCGKiJiIgsjIGaiCiAoDUMKs6il0e01a1SpYpMnjw5zZ6w0JJm8ODB0qhRI10PFXMxrVq1yunyGPgIAx6hfDVXrlzaA+TKlStTLHf8+HHp3r273HbbbdoMFi10Ro4c6bQzqfTue8BDZTJK7sKFC6hgp/8TEfmLuLg4o1SpUvr75Tj17Nkz1XX/+ecfp+t99dVXKZZdv369kS1bthTLBgUFGfPnz7ctd/LkSaNkyZJOt9u8eXMjKSnJK/se6HGGKWoiogAxatQoOXLkiP793nvv6ZDBrVu31tezZs2SLVu2uFw3LCxM+6H4+OOPpWfPnql+Dt5HChy1lrdu3SoHDx6UUqVK2Tqcunbtmi6HIYmPHj2qf8+cOVO7zcQ+wtdffy0LFizwyr4HvEx5dPAzTFETkb9JTEw0IiIiNJVWsWJF2/yNGzfaUqb9+/d3a1uxsbEuU9Tbt2+3vff888/b5o8dO9Y2f/HixTovKipKX+fMmdOWej5//rxtuQcffNDr++4vmKImIspiDh06ZBvWt1KlSrb59n/v2LHjlj/HfhtpfU5agxrt3LkzU/fdXzHrm4goAGC0P5N9Bxr2fyM7OTM/JyoqSv+/evWqZl+ji2h0FW06e/Zspu67v2KgJiIKYPadT7rqYjmjPgcjE5qjRPXu3VvHZTDLqM1ycU+3mRUxUBMRBQAMy2sys5Edxz22XyYzPgfjN6xdu1buvfdeHRQJ85966ilbljaGE87MffdXPh09i4iIvKNs2bKSL18+OX/+vOzfv982f9++fba/77zzzlv+HPttuPM5DRo0kO+++872Gtnc5uBKGHgpM/fdXzFFncVYrTMEczvOJty0phUrVkizZs2kWLFi2mlCzpw5pWrVqjJ8+HC5fPmyF84Mkf+Pb9yxY0f9G8EOIwYiKI4dO9a2DFKzgFEJcY+ZgdJ05swZnVCmbD94BOaZKV0EzMqVK+vfaF61bds2bZ6FMmgoWLCgtGjRQv8+d+6czJkzR393ULEMlcceffRRLatGtnffvn093vcsyddV1K0oUJtnWa0zBHC2TXPCZ5qGDx/ucrkWLVp48SwRBf49Xrp0aZ0XHR3t9v1ov+y6devcuscPHz7scnvTpk1L175nxTjDQH2LJ9Cf9OrVy3bhv/fee8apU6eM1q1b2+b99NNPLte9fPmy0adPH+Pjjz/Wmya1QF25cmV9D+0it27dahw8eNB2AxYsWNC4evWqbVlzO2vXrk1137/44gtjyZIlxvHjx3V9vA4PD7etf/bs2Vs8O0SB4e+//zY6d+5sFC5cWIMp7sdJkyZpW2VvBWrYtGmT0axZMyNPnjzaTrpRo0bGihUrki1z7tw545FHHjFKlCih+5IvXz7jgQceMFavXp3ufQ8UDNSZeAL9hRU7Q/AkUDtz55132ta/ePGix+sTEfkKOzwhv+gMwd4TTzyhZVao2Ym/9+7d6/Iz0D3h0qVLZc+ePfq6U6dOkidPnlvedyIiK2Kt7yzCip0h2ENlFfP/zz77TL755hvt27dixYq2ZU6ePKkj8Nhr166dVlYhIgpUrPWdxfmyMwRAre3t27drze0///xTHn74YVtN03HjxqW53U8//VS6du2aQXtNROR7DNRZhBU7Q4AxY8Zocw804SpdurS88847tvcwKo+9okWLasBH0w50olCiRAmdj9F+EOyJiAIRA3UWYXYoAFbpDMFZ22371LarFD7aUKP95+OPP26bd+DAgVvedyIiK2IZdRZhdiiAFKvZoQDGenXVGcJff/0l0dHRsm7duhTlyM46Q0BFMIxNa3aGgMpg6Azh2Wef1QcEZ50hYF9+/vlnXaZGjRoSFxcnAwYMsG0bnaSYsAz2r1q1alpxDCnozz//PNmDCJHPzIwUSbjCLyCQheUS6e2bgUGCUE3cJ59sYQg+CDrIurWvBOXvUImrbt26tsHZHQeCN7OdXQXq1Mqw7Zddv3699kqG3szsYX1kU5s9EGEUnYEDBzrdHrK5UZnM7As4tc9u06aNfPHFF2kcPVEGmpJL5OZ/D7AUgEJzisRc8UmcYdZ3FhIZGSkbN26Uzp07azlxtmzZNPU7adIkmTFjhtc+B0EbwRpdfiL1i6xqdD26fPlyW5CGhx56SIYMGaId9xcoUEBT5aVKlZLnn39eU8xmkIaYmBhNrWO5kJAQTaVjm1OnTtVa4kREgYop6iyUoiaiDMIUdeALZYqaiIiInGDWNxERkYUxUBMREVkYAzUREZGFMVATERFZGDs8ySTtJq6W6wmJmfVx5APZw0Lk0xea8dwTkVcxUGcSBOl4BmoiIvIQs76JiIgsjIGaiIjIwhioiYiILIyBmoiIyMIYqImIiCzM54EaozZhWMXs2bNL/fr1dWhDV/bs2SOPPfaYLo9hDzFM4q1uk4iIyMp8GqgXLlwogwYNktjYWNmxY4dERUVJ8+bNddxkZ65evSply5aV8ePH63jF3tgmERGRlfk0UGMc5B49ekjXrl2lSpUqMmvWLB27eM6cOU6Xr1u3rkyYMEE6dOgg4eHhXtkmERGRlfksUN+4cUO2b98uTZs2/W9ngoP19aZNmyyzTSIioizZM9mZM2ckMTFRihQpkmw+Xu/bty9TtxkfH6+T6eLFi/p/UlKSTt4QJIZOFLjw/XrreiF/TPP4vMoPZahgBAWvbc2T3wp2ISoi48aNk9GjR6c4OadPn5br16975UspkTtJEm4yUAeysNAk1oXIqnLUFEn872GfAlBIuIgX6zpdunTJ+oG6UKFCEhISInFxccnm47WrimIZtc1hw4ZpBTT7FHXJkiWlcOHCkjdvXvGGvy8HS3wCA3UgCw8LlsjISF/vBvnCtV0iN6/y3Aey0JwiXry/0SrJ7Y8WH8mWLZvUrl1b1qxZI23btrVlBeB13759M3WbqJjmrHIayrcxecO/Gd9BXtkWWRO+X29dL+RvkI3JYo/AloSg4LWtefJb4dOsb6Riu3TpInXq1JF69eppu+grV65ojW3o3LmzFC9eXLOmzcpiv/32m+3vY8eOya5duyR37txSvnx5t7ZJRETkT3waqNu3b6/lwCNHjpSTJ09KzZo1ZdWqVbbKYEeOHEn21HH8+HGpVauW7fVbb72lU3R0tKxbt86tbRIREfmTIMMw3C44RY3qH3/8UWrUqCH58uWTQIUy6oiICLlw4YLXyqjbjF/F8agDXHhYiCwb2sLXu0G+MCUXy6izQhl1zBWfxBmPMtxRUeuBBx6Qf/7551b3kYiIiNzgccl4tWrV5NChQ56uRkRERJkRqMeMGSODBw+W5cuXy4kTJzT5bj8RERGRDyuTPfjgg/p/mzZtdAQrE4q68Rrl2EREROSjQL127VovfTQRERF5PVCjKRQRERFZuB31+fPn5b333pO9e/fq66pVq0q3bt20qjkRERH5sDLZtm3bpFy5cjJ58mQ5d+6cThgDGvN27NjhxV0jIiIij1PUAwcO1Ipk7777roSG/rv6zZs35dlnn5UBAwbI999/z7NKRETkq0CNFLV9kNaNhIbKiy++qP1rExERkQ+zvtHVGfrgdnT06FHJkyePt/aLiIiI0hOoMehF9+7dZeHChRqcMS1YsECzvjt27MiTSkRE5Musb4xWhY5NMAQlyqYhLCxMevXqJePHj/fmvhEREWV5HgVq9Dq2efNmGTVqlI4R/ccff+h81PjOmTNnlj+ZREREPg3U5uhZaD99++23S/Xq1b2+Q0RERPQfjp5FRERkYRw9i4iIyMI4ehYREZGFcfQsIiKiQAnUCQkJ8uqrr8qsWbOkQoUKGbdXRERE5HkZNdpL//LLL56sQkRERJlZmaxTp046xCURERFZsIwavZHNmTNHvv32W6ldu7bkypUr2fsY8pKIiIh8FKh3794td955p/79+++/J3sPXYsSERGR97DWNxERUSCVUafm1KlT3twcERFRlud2oMagG6dPn7a9btWqlZw4ccL2Oi4uTm677bYsf0KJiIh8EqivX78uhmHYXn///fdy7dq1ZMvYv09EREQWy/pmZTIiIiILB2oiIiLyUaBGatk+xez4moiIiHzYPAvlz3fccYctOF++fFlq1aolwcH/xnqWTxMREfkwUM+dOzcDPp6IiIi8Eqi7dOni7qJERETkJaxMRkREZGEM1ERERBbGQE1ERGRhPg/UM2bMkDJlykj27Nmlfv36smXLllSXX7RokVSqVEmXr169uqxcuTLZ+6iN3rdvXylRooTkyJFDqlSpIrNmzcrgoyAiIrJYoL5x44bs379fx6dOr4ULF8qgQYMkNjZWduzYIVFRUdK8eXOXg3ts3LhROnbsKN27d5edO3dK27ZtdcLQmyZsb9WqVfLRRx/J3r17ZcCAARq4ly1blu79JCIi8ptAffXqVQ2UGKSjatWqcuTIEZ3fr18/GT9+vEfbmjRpkvTo0UO6du1qS/liu3PmzHG6/JQpU6RFixYyZMgQqVy5srz22ms6Nvb06dOTBXPUUG/SpImm1J977jl9AEgrpU5ERBQQ41EPGzZMfv75Z1m3bp0GTVPTpk1l1KhRMnToULdT5Nu3b9ftmdB5CrazadMmp+tgPlLM9pACX7p0qe11o0aNNPXcrVs3KVasmO7n77//LpMnT3a5L/Hx8TqZLl68qP8nJSXp5A1BYuhEgQvfr7euF/LHNI/PSxIpQwUjKHhta578VngcqBEUkWXdoEGDZF2IInX9xx9/uL2dM2fOSGJiohQpUiTZfLzet2+f03VOnjzpdHnMN02bNk1T0SijDg0N1eD/7rvvSuPGjV3uy7hx42T06NEp5mNYT4wa5g0lcidJwk0G6kAWFprEMdmzqhw1RRL/e9inABQSLuKiWDY9Ll26lHGBGsErMjIyxfwrV65You9vBOrNmzdrqrp06dI6HGefPn00dY3UujNI1dun1JGiLlmypBQuXFjy5s3rlf36+3KwxCcwUAey8LBgp/cGZQHXdoncvOrrvaCMFJpTxIv3NypEu/3Rnm68Tp06smLFCi2TBjM4/9///Z80bNjQ7e0UKlRIQkJCJC4uLtl8vC5atKjTdTA/teUxPvbLL78sS5YskVatWum8GjVqyK5du+Stt95yGajDw8N1coTUuNmX+a36N+Pb9w8ylHHw/XrreiF/g2xMFnsEtiQEBa9tzZPfCo8/dezYsRoMe/XqpTW+UcHrgQce0L7AX3/9dbe3ky1bNqldu7asWbMmWZ49XrsK+JhvvzysXr3atnxCQoJOjicADwQsOyQiIn/kcaC+++67NYWKII12zN98841m96GiFwKvJ5DdjPLj999/X5tSIfgjCx21wKFz587JKpvFxMRo06uJEydqOTYqr23btk2bXwGyqaOjo7VWOCqRHT58WObNmycffPCBPPLII54eKhERkc95nPUN5cqV0wB7q9q3b69l3iNHjtQKYTVr1tRAbFYYQ9Mv+9QxanTPnz9fRowYoan6ChUqaOW2atWq2ZZZsGCBBvennnpKzp07p+XUSOn37NnzlveXiIgoswUZHg4kjWzkEydOpKg0c/bsWZ2Hmtz+DpXJIiIi5MKFC16rTNZm/CqJT/D/c0OuhYeFyLKh/zVZpCxkSi5WJssKlclirvgkznic9e0qrqMdMsqdiYiIyAdZ31OnTrXV8kYN79y5c9veQyoazaDQBzcRERH5IFCbPXshRY2uPpEFbkJKGt11cvALIiIiHwVq1KCGe++9VxYvXiz58+f38q4QERHRLdf6Xrt2raerEBERUWYFagx2kRpXI18RERFRJgTqf/75J9lr9ASG8aDPnz8v9913Xzp2gYiIiLwWqNGPtiN0z4lexdARChEREXmPV3oYR+9h6A40tTGfiYiIyHNeGwoEY1Gj/28iIiLyYda3/bjNZrtqdCmKoS+7dOnixV0jIiIijwP1zp07U2R7Fy5cWEe0SqtGOBEREXmG7aiJiIiyQhk1ERER+ShFXatWLR2Mwx07duy41X0iIiIiTwJ127Zt3VmMiIiIfBGoY2Njvf25RERElBGVyUzbt2+XvXv36t9Vq1bV7HEiIiLycaA+deqUdOjQQdatWyf58uXTeejnG8NfLliwQJtqERERkY9qfffr108uXboke/bskXPnzumEQTkuXrwo/fv399JuERERUbpS1KtWrZJvv/1WKleubJtXpUoVmTFjhjzwwAM8q0RERL5MUWOkrLCwsBTzMQ/vERERkQ8DNcacjomJkePHj9vmHTt2TAYOHCj333+/F3eNiIiIPA7U06dP1/LoMmXK6PjTmG6//XadN23aNJ5RIiIiX5ZRlyxZUnsfQzn1vn37dB7Kq5s2berN/SIiIqL0tqNGd6LNmjXTyWyeRURERBbI+n7jjTdk4cKFttft2rWTggULSvHixeXnn3/29v4RERFlaR4H6lmzZmn2N6xevVqnr776Slq2bClDhgzJiH0kIiLKsjzO+j558qQtUC9fvlxT1Gg/jcpl9evXz4h9JCIiyrI8TlHnz59fjh49auv8xKxEZhiGJCYmen8PiYiIsjCPU9SPPvqoPPnkk1KhQgU5e/asZnnDzp07pXz58hmxj0RERFmWx4F68uTJms2NVPWbb74puXPn1vknTpyQ3r17Z8Q+EhERZVkeB2p0FTp48OAU89EzGREREVmgHfX+/fu1FzJzPGp0eIJRtSpWrOjl3SMiIsraPK5M9vnnn0u1atVk+/btEhUVpRN6KsM8vEdEREQ+TFG/+OKLMmzYMHn11VeTzY+NjdX3HnvsMS/uHhERUdbmcYoalcY6d+6cYn6nTp30PU9hHGtUTsuePbu2w96yZUuqyy9atEgqVaqky1evXl1WrlyZYhlkybdp00YiIiIkV65cUrduXTly5IjH+0ZEROR3gbpJkyayYcOGFPN/+OEHueeeezzaFroiHTRokKbGkX2ObPTmzZvLqVOnnC6/ceNG6dixo3Tv3l2bg7Vt21an3bt325b5448/5O6779Zgvm7dOvnll1/klVde0cBORETkb4IM9FSShmXLltn+xjjUI0eO1B7JGjRooPM2b96sKd3Ro0dLz5493f5wpKCR2sXQmZCUlKS9nqFi2tChQ1Ms3759e7ly5Yr2iGbCPtSsWVO7NoUOHTpozfQPP/xQ0gtDdiI1fuHCBcmbN694Q5vxqyQ+gR3CBLLwsBBZNrSFr3eDfGFKLpGbV3nuA1loTpGYK17bnCdxxq0yaqRaHc2cOVMne3369HE7UN+4cUMrpKG82xQcHKw9nW3atMnpOpiPFLg9pMCXLl1qC/QrVqzQsnLMR6obY2XjM5wdAxERkdW5FagRAL3tzJkz2uVokSJFks3Ha3Oca2f9jDtbHvMBWeaXL1+W8ePHy5gxY3SkL3Rzit7U1q5dK9HR0U63Gx8fr5P9k4553N469iAxdKLAhe83I+4V8pdSRI9LEsmvBCMoeG1rnvxWpKsdtTMYk/qjjz6Svn37iq+YB/7www/bOmBBtjjKtpE17ipQjxs3TrPtHZ0+fVquX7/ulX0rkTtJEm4yUAeysNAkl/UrKMDlqCmS+N/DPgWgkHCkBr22uUuXLmVeoF6zZo289957smTJEsmZM6fbgbpQoUISEhIicXFxyebjddGiRZ2ug/mpLY9thoaGSpUqVZItgw5ZUNnNFWSN22epI0WNsvLChQt7rYz678vBEp/AQB3IwsOCJTIy0te7Qb5wbRfLqLNCGXWk9+5vTyo4pytQo5/vuXPn6oRmT6jAhUB9//33u72NbNmySe3atTXQm+XHSBHjtatg37BhQ31/wIABtnkYDxvzzW2ichp6TrP3+++/S+nSpV3uS3h4uE6OUGaOyRv+zfgO8sq2yJrw/XrreiF/g9w8FnsEtiQEBa9tzZPfCreXTEhI0JrdqKSFrkJ37dolEyZM0A8bPny4tGjRQmtbewKp2HfffVfef/99bfvcq1cvrdXdtWtXfR/tte0rm8XExGiZ88SJE7Uce9SoUbJt27ZkgX3IkCHa7AvbPXjwoNYo//LLLzlgCBER+SW3U9TFixfXtsno2GTBggU6LjWgXXN6obkVyoHR3AsVwlCejEBsVhhDat3+qaNRo0Yyf/58GTFihLz88ss61CZqfKP7UtMjjzyi5dEod+7fv78+VKBrU7StJiIiCthAffPmTQkKCtIJZcvegtSwq6xudFji6IknntApNd26ddOJiIjI37md9Y2OTp577jn55JNPtPIW+vRGuTQCNxEREfk4UKOG2lNPPSXfffed/Prrr1qTGlnLSGm//vrrWqkL7aKJiIjIe9JVha1cuXLaochff/2lPYGhs5DWrVun6IyEiIiIbs0ttaNGRa+WLVvqhEpht9K/NhEREaXktUZh6BzEsR9uIiIiujXsnYGIiMjCGKiJiIgsjIGaiIjIwhioiYiIAqnWN9pKz5s3TwfHwJB+jmNqop01ERER+ShQY2AMBOpWrVppH9vsmYyIiMhCgRoDcnz66afy4IMPZsweERERUfrLqDHmc/ny5T1djYiIiDIjUL/wwgsyZcoUMQwjPZ9HREREGZn1/cMPP8jatWvlq6++kqpVq0pYWFiy9xcvXuzpJomIiMhbgTpfvnzyyCOPeLoaERERZUagnjt3bno+h4iIiNKBHZ4QEREF2jCXn332mTbROnLkiNy4cSPZezt27PDWvhEREWV5Hqeop06dKl27dpUiRYrIzp07pV69elKwYEE5dOiQjktNREREPgzUM2fOlNmzZ8u0adO0TfWLL74oq1evlv79+8uFCxe8uGtERETkcaBGdnejRo307xw5csilS5f076efflo++eQTnlEiIiJfBuqiRYvKuXPn9O9SpUrJ5s2b9e/Dhw+zExQiIiJfB+r77rtPli1bpn+jrHrgwIHSrFkzad++PdtXExER+brWN8qnzaEt+/TpoxXJNm7cKG3atJHnn3/e2/tHRESUpXkcqIODg3UydejQQSciIiKySIcnGzZskE6dOknDhg3l2LFjOu/DDz/UfsCJiIjIh4H6888/l+bNm2uNb7Sjjo+P1/lomjV27Fgv7hoRERF5HKjHjBkjs2bNknfffTfZyFl33XUXeyUjIiLydaDev3+/NG7cOMX8iIgIOX/+vLf2i4iIiNLbjvrgwYMp5qN8umzZsjypREREvgzUPXr0kJiYGPnpp58kKChIjh8/Lh9//LEMHjxYevXq5c19IyIiyvI8bp41dOhQbUd9//33y9WrVzUbPDw8XAN1v379svwJJSIi8mmgRip6+PDhMmTIEM0Cv3z5slSpUkVy587t1R0jIiKidI5HDRg5CwGaiIiILBCou3Xr5tZyc+bMuZX9ISIiovQE6nnz5knp0qWlVq1aHCWLiIjIarW+UaMbvY9hOMt7771X3nvvPVmyZEmKKT1mzJghZcqUkezZs0v9+vVly5YtqS6/aNEiqVSpki5fvXp1Wblypctle/bsqeXqb7/9drr2jYiIyC8CNYLpiRMn5MUXX5Qvv/xSSpYsKe3atZOvv/76llLYCxculEGDBklsbKz2bBYVFaVdlJ46dcrp8hipq2PHjtK9e3ftwrRt27Y67d69O8WyeHDAeNnFihVL9/4RERH5TTtqNMNCkFy9erX89ttvUrVqVendu7emhlH7Oz0mTZqkbbMxtjUqp6F70pw5c7os654yZYq0aNFCa51XrlxZXnvtNbnzzjtl+vTpyZbDYCFoLoY23vZdnRIREWWJWt8Y6hJZykhNJyYmpmsbN27ckO3bt8uwYcOSbbdp06ayadMmp+tgPlLg9pACX7p0qe012nk//fTTGszxMJEWDCxiDi4CFy9etG3HHHv7VgWJoRMFLny/3rpeyB/TPOkajJD8RjCCgte25slvhUeBGsFs8eLFmtpFl6GtW7fWlCxSuPZjVLvrzJkzGuSLFCmSbD5e79u3z+k6J0+edLo85pveeOMNCQ0Nlf79+7u1H+PGjZPRo0enmH/69Gm5fv26eEOJ3EmScJOBOpCFhSa5LLKhAJejpkjifw/7FIBCwkW8eH9funTJ+4EaWdwLFizQsmk01frkk0+kUKFCYjVIoSN7HOXdSPG7Ayl6+1Q6UtQ4zsKFC0vevHm9sl9/Xw6W+AQG6kAWHhYskZGRvt4N8oVru0RuXuW5D2ShOUW8eH+jMrTbH+3ugig7LlWqlA68sX79ep2cQYrbXQj0ISEhEhcXl2w+XmPwD2cwP7XlN2zYoKka7KsJqfYXXnhBa37/+eefTsveMTlCLkF6cgqc+Tfj270HB/JP+H69db2Qv0E2Jos9AlsSgoLXtubJb4Xbgbpz585up1A96d2sdu3asmbNGq25bebb43Xfvn2drtOwYUN9f8CAAbZ5qNyG+YCyaZRxO5ZhYz4qrBEREQVshycZAVnOXbp0kTp16ki9evU01XvlyhVbUMUDQvHixbUcGTByV3R0tEycOFFatWql2fHbtm2T2bNn6/sFCxbUyR5qfSPFXbFixQw5BiIiIsvV+vaW9u3ba6WtkSNHaoWwmjVryqpVq2wVxo4cOZIsi6BRo0Yyf/58GTFihLz88stSoUIFrfFdrVo1Hx4FERFRxggybqW3kgCFymQRERHaE5u3KpO1Gb9K4hPS14yN/EN4WIgsG9rC17tBvjAlFyuTZYXKZDFXfBJnWPOFiIjIwhioiYiILIyBmoiIyMIYqImIiCyMgZqIiMjCGKiJiIgsjIGaiIjIwhioiYiILIyBmoiIyMIYqImIiCyMgZqIiMjCGKiJiIgsjIGaiIjIwhioiYiILIyBmoiIyMIYqImIiCyMgZqIiMjCGKiJiIgsjIGaiIjIwhioiYiILIyBmoiIyMIYqImIiCyMgZqIiMjCGKiJiIgsjIGaiIjIwhioiYiILIyBmoiIyMIYqImIiCyMgZqIiMjCGKiJiIgsjIGaiIjIwhioiYiILIyBmoiIyMIYqImIiCyMgZqIiMjCGKiJiIgszBKBesaMGVKmTBnJnj271K9fX7Zs2ZLq8osWLZJKlSrp8tWrV5eVK1fa3ktISJCXXnpJ5+fKlUuKFSsmnTt3luPHj2fCkRAREQVYoF64cKEMGjRIYmNjZceOHRIVFSXNmzeXU6dOOV1+48aN0rFjR+nevbvs3LlT2rZtq9Pu3bv1/atXr+p2XnnlFf1/8eLFsn//fmnTpk0mHxkREdGtCzIMwxAfQgq6bt26Mn36dH2dlJQkJUuWlH79+snQoUNTLN++fXu5cuWKLF++3DavQYMGUrNmTZk1a5bTz9i6davUq1dP/vrrLylVqlSa+3Tx4kWJiIiQCxcuSN68ecUb2oxfJfEJiV7ZFllTeFiILBvawte7Qb4wJZfIzas894EsNKdIzBWvbc6TOOPTFPWNGzdk+/bt0rRp0/92KDhYX2/atMnpOphvvzwgBe5qecCJCAoKknz58nlx74mIiDJeqPjQmTNnJDExUYoUKZJsPl7v27fP6TonT550ujzmO3P9+nUts0Z2uaunlvj4eJ3sn3TM1D0mbwgSQycKXPh+vXW9kL8JtkJJImWoYAQFr23Nk98KnwbqjIaKZe3atRPk7r/zzjsulxs3bpyMHj06xfzTp09roPeGErmTJOEmA3UgCwtNclm3ggJcjpoiif897FMACgkX8eL9fenSJf8I1IUKFZKQkBCJi4tLNh+vixYt6nQdzHdneTNIo1z6u+++S7UMYNiwYVqhzT5FjXLywoULe62M+u/LwRKfwEAdyMLDgiUyMtLXu0G+cG0Xy6izQhl1pPfub7RacvujxYeyZcsmtWvXljVr1mjNbTM7AK/79u3rdJ2GDRvq+wMGDLDNW716tc53DNIHDhyQtWvXSsGCBVPdj/DwcJ0cobwckzf8m/Ed5JVtkTXh+/XW9UL+BtmYLPYIbEkICl7bmie/FT7P+kZKtkuXLlKnTh2tmf32229rre6uXbvq+2gDXbx4cc2ehpiYGImOjpaJEydKq1atZMGCBbJt2zaZPXu2LUg//vjj2jQLNcNRBm6WXxcoUEAfDoiIiPyFzwM1mluhLHjkyJEaUNHMatWqVbYKY0eOHEn25NGoUSOZP3++jBgxQl5++WWpUKGCLF26VKpVq6bvHzt2TJYtW6Z/Y1v2kLpu0qRJph4fERGRX7ejtiK2o6b0YDvqLIztqANfaBZtR01ERESpY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILs0SgnjFjhpQpU0ayZ88u9evXly1btqS6/KJFi6RSpUq6fPXq1WXlypXJ3jcMQ0aOHCm33Xab5MiRQ5o2bSoHDhzI4KMgIiIKwEC9cOFCGTRokMTGxsqOHTskKipKmjdvLqdOnXK6/MaNG6Vjx47SvXt32blzp7Rt21an3bt325Z58803ZerUqTJr1iz56aefJFeuXLrN69evZ+KRERER3bogA8lPH0IKum7dujJ9+nR9nZSUJCVLlpR+/frJ0KFDUyzfvn17uXLliixfvtw2r0GDBlKzZk0NzDicYsWKyQsvvCCDBw/W9y9cuCBFihSRefPmSYcOHdLcp4sXL0pERISulzdvXq8cZ7uJq+V6QqJXtkXWlD0sRD59oZmvd4N8YWakSMIVnvtAFpZLpLfzBGR6eBJnQsWHbty4Idu3b5dhw4bZ5gUHB2tW9aZNm5yug/lIgdtDannp0qX69+HDh+XkyZO6DRNOBh4IsK6zQB0fH6+TCScOzp8/rw8O3jC7e12vbIesDdcMZUFP/u7rPaDM4MX7G4Ea3Ekr+zRQnzlzRhITEzW1aw+v9+3b53QdBGFny2O++b45z9UyjsaNGyejR49OMb906dIeHhEREZH7Ll26pIlJywZqq0CK3j6VjlT0uXPnpGDBghIUFOTTffNXeFpEEcbRo0e9VnxARNbA+/vWISWNII2i2rT4NFAXKlRIQkJCJC4uLtl8vC5atKjTdTA/teXN/zEPtb7tl0E5tjPh4eE62cuXL186j4rsIUgzUBMFJt7ftyatlLQlan1ny5ZNateuLWvWrEmWmsXrhg0bOl0H8+2Xh9WrV9uWv/322zVY2y+Dpz/U/na1TSIiIqvyedY3spy7dOkiderUkXr16snbb7+ttbq7du2q73fu3FmKFy+u5cgQExMj0dHRMnHiRGnVqpUsWLBAtm3bJrNnz9b3kVU9YMAAGTNmjFSoUEED9yuvvKLZC2jGRURE5E98HqjR3Or06dPaQQkqeyF7etWqVbbKYEeOHNGa4KZGjRrJ/PnzZcSIEfLyyy9rMEaN72rVqtmWefHFFzXYP/fcc1oL9+6779ZtooMUyhwoSkDbeMciBSLyf7y/s1g7aiIiIrJwz2RERETkGgM1ERGRhTFQExERWRgDNfm9Jk2aaE1/Ikr/fYMRDNHqxmr+/PNPbc2za9cuyaoYqAMMas5jQJOyZctqzUz0DvbQQw+laHvua+g6dvLkyTpMKWrj58+fX1q2bCk//vijr3eNyC8888wzGsAcp4MHD2bI5129elV7cSxXrpzes4ULF9amsl988YXlg72/83nzLPLuk+ddd92lvapNmDBBg2BCQoJ8/fXX0qdPH5f9p2OZsLCwTPsq0NAAg6N8++23up/333+/dkqDccnxlI/xxtnmnShtLVq0kLlz5yabhwCaEXr27KkdR02bNk2qVKkiZ8+e1WGH8T9lMDTPosDQsmVLo3jx4sbly5dTvPfPP//Y/sbXPnPmTOOhhx4ycubMacTGxup8zCtbtqwRFhZm3HHHHcYHH3xgWycpKUmXK1mypJEtWzbjtttuM/r162d7f8aMGUb58uWN8PBwIzIy0njsscdc7ueCBQt0H5YtW5bivUcffdQoWLCg7RjwmVFRUbovpUuXNvLmzWu0b9/euHjxom2d6OhoIyYmRv8ePXq0UbVq1RTbxTZGjBjh1nkk8gddunQxHn74Ybffwz2Ce8XZfQO4vyZPnuzy8yIiIox58+a5fB/bw31tP8GZM2eMDh06GMWKFTNy5MhhVKtWzZg/f36ydRMTE4033njDKFeunP6+4HdmzJgx+t7hw4d1Wzt37tTXN2/eNLp27WpUrFjR+Ouvv4ysgFnfAQKDiKBTF6Scc+XKleJ9x77LR40aJY888oj8+uuv0q1bN1myZIn2+oZxvHfv3i3PP/+89g63du1aXf7zzz/XrOr//e9/cuDAAe1kBil2QM9w/fv3l1dffVX279+v+9G4cWOX+4oOa+644w7NkneEz8cTOrqFNf3xxx/6eRiDHNP69etl/PjxTreNY9m7d69s3brVNm/nzp3yyy+/2Hq7IyLPoWvmlStX6kASzixevFhKlCihvwMnTpzQCa5fv65dRa9YsUJ/W9AR1dNPPy1btmyxrYssddzT6EXyt99+098IxxEQAcMRP/HEE1pevWHDBilVqpRkBcz6DhAol0JiuVKlSm4t/+STTyYLXB07dtQyr969e9u6dt28ebO89dZbcu+992oPcbhRMc43sslxg6DLV8B7eDho3bq15MmTR4cHrVWrlsvP/v3336Vy5cpO3zPnYxn7/t/nzZun2wbc5Chzf/3111Osjx8KjE+O7MC6df8dAxx/oywN5fZEgQQPrrlz57a9Rj0PFB1lBHTT/NRTT+moglFRUdrj4+OPP67FbVCgQAEdZAn3qf2gSugCevDgwbbXqEOD4rhPP/1Uf0MQ+KdMmSLTp0/X7qQB5eDYvr3Lly9rt9EI1khAuDugRSBgijpAeNrBHPpWt4dUqHnDmfAa8wFPsdeuXdNg16NHD02B37x5U99r1qyZBme8hyD68ccfa8UTb+0vKqiYQRowKtqpU6dcLo/9++STT/RJ/saNG/p0jpQ2UaDBQzRSl+Y0derUDPss5JIdOnRIH5IRoPfs2SP33HOPvPbaa2lWHMUyyIFDMMeDBQI1HvABvzEIvqirkpqOHTtq19DffPNNlgrSwEAdINDnOWp8uqow5shZ9nhqUHsc2dozZ86UHDlyaMobNy4qoiGI7tixQ4Mjgij6bccTN/pZdwbZ3uYDgCNzPpYxOVZ0w3Eile0KstRR4x0PE19++aXuI35YiAIN7uPy5cvbJnNoX4yP4PgwjPvgVuFeRHB+6aWXNGAimxtBGA/ErqDCKFLMWAcpYTxQINfLXAe/J+548MEHtQhr06ZNktUwUAcIPKni4kfNaTx1OnIVNO2znB2bRuE1aneacEMhCOKpfd26dXrDoIwbQkNDNVv8zTff1JsJNdC/++47p5+FGt8o50YQdYRR0ZC1hlR6emFfkIWGLG9M+Dx3fwyIAgFqfptlxKaMaIeM3wfkrCH3yhy6GClox9+Rhx9+WDp16qQP8Mh5sy/aQiID92daTUh79eql5dht2rTReipZCcuoAwiCNLKrUe6DJ90aNWroTYSKWe+8847LVCwMGTJE2rVrp2XLCLgIoqgcgiZUgDJi3ID169eXnDlzykcffaQ3F7K8UU6GLDGksNEeGhVOkOKtWLGi089C4EQ5GoKpY/OsZcuW6XuepvgdPfvss7bybrbNpqzmvvvu03vrgw8+kIYNG+r9iopcqdUdSQuaTiL7GcVmeJhGpS+MYIjs97x589qKqb7//nu9x5GrVahQIQ3En332mTblwu/DpEmTJC4uzpYIQJtspLYx6iECPX7DMKIista7d++ebB9Qvo3fIdSH+eqrr1KUYwcsX1c7J+86fvy40adPH21qgWYOaK7Vpk0bY+3atbZl8LUvWbIkxbqpNc/C8vXr19fmUbly5TIaNGhgfPvtt/rehg0btGlG/vz5tflFjRo1jIULF6a6nwkJCcaECRO0KRX2E9tt3ry58cMPPyRbzmyeZQ9NSHB8rpqZmO655x6nTbWIAr15FowcOdIoUqSINqsaOHCg0bdv31tqnjV27FijYcOGRoECBYzs2bPrb0X//v21+ZVp06ZNev+jmaYZXs6ePav7mTt3bm26iWaSnTt3TrbvaJ6F5ljYB/z+lCpVSj/PWfMsmDhxopEnTx7jxx9/NLICDnNJAQnPI3iSR1k6arATEfkrZn1TwEG22YIFC7Q7VbadJiJ/x0BNAScyMlLLxtDuE2ViRET+jIGaJKu3KScisjI2zyIiIrIwBmoiIiILY6AmIiKyMAZqIiIiC2OgJiIisjAGaiIiIgtjoCYiIrIwBmoiIiILY6AmIiIS6/p/CeeXEI/E8MkAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 500x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "labels = ['Cross Only', 'Full Stack']\n",
    "maes   = [mae_cross, mae_full]\n",
    "colors = ['steelblue', 'darkorange']\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(5, 4))\n",
    "bars = ax.bar(labels, maes, color=colors, width=0.5, edgecolor='white', linewidth=1.5)\n",
    "for bar, v in zip(bars, maes):\n",
    "    ax.text(bar.get_x() + bar.get_width()/2, v + 0.001, f'{v:.4f}',\n",
    "            ha='center', va='bottom', fontsize=11, fontweight='bold')\n",
    "ax.set_title('MAE by Attention Stack Configuration', fontsize=13)\n",
    "ax.set_ylabel('Mean Absolute Error')\n",
    "ax.set_ylim(0, max(maes) * 1.25)\n",
    "ax.grid(True, axis='y', alpha=0.3)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "416dc394",
   "metadata": {},
   "source": [
    "## Recommendation Guide\n",
    "\n",
    "**Choose based on your data characteristics:**\n",
    "\n",
    "- **Simple Forecasting** → Use Cross-Attention only\n",
    "- **Seasonal/Trend Pattern** → Use Cross + Hierarchical\n",
    "- **Complex Multi-scale** → Use Full Stack\n",
    "- **Limited Memory/Speed** → Use Cross-Attention or Cross+Hierarchical"
   ]
  }
 ],
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.20"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}