model_testing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\bujiw\\AppData\\Local\\Temp\\ipykernel_52432\\3889204627.py:33: The name tf.logging.set_verbosity is deprecated. Please use tf.compat.v1.logging.set_verbosity instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf2\n",
    "assert tf2.__version__ >= \"2.0\"\n",
    "import numpy as np\n",
    "from scipy.stats import norm\n",
    "import matplotlib.pyplot as plt\n",
    "import time\n",
    "from tqdm import tqdm_notebook\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import pandas as pd \n",
    "from tqdm import tqdm\n",
    "from multiprocessing import Pool\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.decomposition import PCA\n",
    "from tensorflow.keras.models import load_model\n",
    "from tensorflow.keras.losses import MeanSquaredError\n",
    "from concurrent.futures import ProcessPoolExecutor, as_completed\n",
    "from multiprocessing import Manager\n",
    "import random\n",
    "import os\n",
    "from scipy.integrate import quad_vec  # quad_vec allows to compute integrals accurately\n",
    "from scipy.stats import norm\n",
    "from scipy.stats import qmc # to perform Latin Hypercube Sampling (LHS) \n",
    "import pandas as pd \n",
    "\n",
    "def set_random_seed(seed=42):\n",
    "    tf.compat.v1.set_random_seed(seed)\n",
    "    np.random.seed(seed)\n",
    "    random.seed(seed)\n",
    "\n",
    "os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'  # Suppress INFO and WARNING logs\n",
    "tf = tf2.compat.v1\n",
    "tf.logging.set_verbosity(tf.logging.ERROR)\n",
    "set_random_seed()\n",
    "real_type = tf.float32\n",
    "\n",
    "## Load the dataset\n",
    "feller_data = pd.read_csv('data//feller_d2.csv')\n",
    "\n",
    "## Obtain 2nd order differentials and their PCs\n",
    "second_differentials = feller_data.iloc[:,-36:]\n",
    "normalized_zscore = (second_differentials - second_differentials.mean()) / second_differentials.std()\n",
    "pca_d2 = PCA(n_components=15)\n",
    "second_differential_label = pca_d2.fit_transform(normalized_zscore[:9000])\n",
    "\n",
    "## Train-test Split (90-10)\n",
    "feller_testing = feller_data.iloc[9000:]\n",
    "feller_data = feller_data.iloc[:9000]\n",
    "\n",
    "training_col = feller_data.columns[:8]\n",
    "training_target = feller_data.columns[9]\n",
    "network_inputs = feller_data[training_col].values\n",
    "network_outputs = feller_data[training_target].values\n",
    "\n",
    "func_names = [\"lm\", \"r\", \"tau\", \"theta\", \"sigma\", \"rho\", \"kappa\", \"v0\"]\n",
    "network_inputs = feller_data[func_names]\n",
    "option_prices = feller_data['P_hat']\n",
    "network_first_order = feller_data[[f\"diff wrt {col}\" for col in func_names]].values\n",
    "sec_order_names = []\n",
    "for i in func_names:\n",
    "    for j in func_names:\n",
    "        if os.path.exists(f\"data//d2_{i}_{j}.csv\"):\n",
    "            sec_order_names.append(f\"d2_{i}_{j}\")\n",
    "network_second_order = feller_data[[f\"{col}\" for col in sec_order_names]].values\n",
    "\n",
    "\n",
    "## Load the training records\n",
    "combined_df = pd.read_csv(\"results//learning_history.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_grad_model1(model, raw_inputs):\n",
    "    with tf.GradientTape(persistent=True) as tape:\n",
    "        tape.watch(raw_inputs) \n",
    "        option_price = model(raw_inputs)  \n",
    "    first_order_differential = tape.gradient(option_price, raw_inputs)\n",
    "    return option_price, first_order_differential\n",
    "\n",
    "def compute_grad_model2(model, raw_inputs):\n",
    "    with tf.GradientTape(persistent=True) as tape:\n",
    "        tape.watch(raw_inputs) \n",
    "        option_price = model(raw_inputs)\n",
    "    first_order_differential = tape.gradient(option_price, raw_inputs)\n",
    "    return option_price, first_order_differential\n",
    "def compute_grad_model3(model, raw_inputs):\n",
    "    with tf.GradientTape(persistent=True) as tape:\n",
    "        tape.watch(raw_inputs) \n",
    "        option_price, second_order_diff = model(raw_inputs)\n",
    "    first_order_differential = tape.gradient(option_price, raw_inputs)\n",
    "    return option_price, first_order_differential, second_order_diff\n",
    "\n",
    "def compute_grad_model4(model, raw_inputs):\n",
    "    with tf.GradientTape(persistent=True) as tape:\n",
    "        tape.watch(raw_inputs) \n",
    "        option_price, second_order_diff = model(raw_inputs)  \n",
    "    first_order_differential = tape.gradient(option_price, raw_inputs)\n",
    "    return option_price, first_order_differential, second_order_diff"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Performance of Model 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Loss: $0.181\n",
      "Price Loss: $0.069\n",
      "1st Order Differential Loss: $0.224\n"
     ]
    }
   ],
   "source": [
    "subset_0_1 = combined_df[(combined_df['source'] == 'history1')]\n",
    "starting_point = 0\n",
    "total_loss = subset_0_1[\"total_loss\"].values[starting_point:][-1]\n",
    "price_loss = subset_0_1[\"price_loss\"].values[starting_point:][-1]\n",
    "first_order_loss = subset_0_1[\"first_order_loss\"].values[starting_point:][-1]\n",
    "\n",
    "print(f\"Total Loss: ${total_loss:.3f}\")\n",
    "print(f\"Price Loss: ${price_loss:.3f}\")\n",
    "print(f\"1st Order Differential Loss: ${first_order_loss:.3f}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Performance of Model 2\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Loss: $0.078\n",
      "Price Loss: $0.070\n",
      "1st Order Differential Loss: $0.075\n",
      "Total Loss: $0.090\n",
      "Price Loss: $0.075\n",
      "1st Order Differential Loss: $0.030\n",
      "Total Loss: $0.088\n",
      "Price Loss: $0.076\n",
      "1st Order Differential Loss: $0.017\n"
     ]
    }
   ],
   "source": [
    "subset_0_1 = combined_df[(combined_df['source'] == 'history2')].groupby(['lambda1'])\n",
    "starting_point = 0\n",
    "lambda_keys = [0.1, 0.5, 0.7]\n",
    "\n",
    "for i in lambda_keys:\n",
    "    total_loss = subset_0_1.get_group(i)[\"total_loss\"].values[starting_point:][-1]\n",
    "    price_loss = subset_0_1.get_group(i)[\"price_loss\"].values[starting_point:][-1]\n",
    "    first_order_loss = subset_0_1.get_group(i)[\"first_order_loss\"].values[starting_point:][-1]\n",
    "    print(f\"Total Loss: ${total_loss:.3f}\")\n",
    "    print(f\"Price Loss: ${price_loss:.3f}\")\n",
    "    print(f\"1st Order Differential Loss: ${first_order_loss:.3f}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Performance of Model 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   Lambda1  Lambda2  Total Loss  Price Loss  1st Order Loss  2nd Order Loss\n",
      "0      0.1      0.1    0.224117    0.070678        0.069237        1.465150\n",
      "1      0.1      0.5    0.302010    0.071235        0.041287        0.453294\n",
      "2      0.1      0.7    0.289425    0.079906        0.039647        0.293649\n",
      "3      0.5      0.1    0.107951    0.074218        0.013145        0.271608\n",
      "4      0.5      0.5    0.169962    0.073988        0.012880        0.179066\n",
      "5      0.5      0.7    0.174212    0.073952        0.012480        0.134314\n",
      "6      0.7      0.1    0.094249    0.075086        0.008926        0.129141\n",
      "7      0.7      0.5    0.129983    0.075026        0.009103        0.097169\n",
      "8      0.7      0.7    0.137508    0.075077        0.009026        0.080161\n"
     ]
    }
   ],
   "source": [
    "subset_0_1 = combined_df[(combined_df['source'] == 'history4')].groupby(['lambda1', 'lambda2'])\n",
    "starting_point = 0\n",
    "lambda_keys = [0.1, 0.5, 0.7]\n",
    "\n",
    "table_data = []\n",
    "for i in lambda_keys:\n",
    "    for j in lambda_keys:\n",
    "        total_loss = subset_0_1.get_group((i, j))[\"total_loss\"].values[starting_point:][-1]\n",
    "        price_loss = subset_0_1.get_group((i, j))[\"price_loss\"].values[starting_point:][-1]\n",
    "        first_order_loss = subset_0_1.get_group((i, j))[\"first_order_loss\"].values[starting_point:][-1]\n",
    "        second_order_loss = subset_0_1.get_group((i, j))[\"second_order_loss\"].values[starting_point:][-1]\n",
    "        table_data.append([i,j,total_loss,price_loss,first_order_loss,second_order_loss])\n",
    "\n",
    "df = pd.DataFrame(table_data, columns=['Lambda1','Lambda2', 'Total Loss', 'Price Loss', '1st Order Loss', '2nd Order Loss'])\n",
    "print(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Performance of Model 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   Lambda1  Lambda2  Total Loss  Price Loss  1st Order Loss  2nd Order Loss\n",
      "0      0.1      0.1    0.156377    0.069493        0.095656        0.773187\n",
      "1      0.1      0.5    0.193802    0.070729        0.046152        0.236914\n",
      "2      0.1      0.7    0.187967    0.072102        0.033746        0.160700\n",
      "3      0.5      0.1    0.095559    0.073957        0.014795        0.142048\n",
      "4      0.5      0.5    0.120946    0.073790        0.013242        0.081070\n",
      "5      0.5      0.7    0.121179    0.073750        0.012878        0.058556\n",
      "6      0.7      0.1    0.087427    0.074938        0.009788        0.056373\n",
      "7      0.7      0.5    0.102068    0.074925        0.009545        0.040922\n",
      "8      0.7      0.7    0.102783    0.075139        0.009033        0.030459\n"
     ]
    }
   ],
   "source": [
    "subset_0_1 = combined_df[(combined_df['source'] == 'history3')].groupby(['lambda1', 'lambda2'])\n",
    "starting_point = 0\n",
    "lambda_keys = [0.1, 0.5, 0.7]\n",
    "\n",
    "table_data = []\n",
    "for i in lambda_keys:\n",
    "    for j in lambda_keys:\n",
    "        total_loss = subset_0_1.get_group((i, j))[\"total_loss\"].values[starting_point:][-1]\n",
    "        price_loss = subset_0_1.get_group((i, j))[\"price_loss\"].values[starting_point:][-1]\n",
    "        first_order_loss = subset_0_1.get_group((i, j))[\"first_order_loss\"].values[starting_point:][-1]\n",
    "        second_order_loss = subset_0_1.get_group((i, j))[\"second_order_loss\"].values[starting_point:][-1]\n",
    "        table_data.append([i,j, total_loss, price_loss, first_order_loss,second_order_loss])\n",
    "\n",
    "df = pd.DataFrame(table_data, columns=['Lambda1','Lambda2', 'Total Loss', 'Price Loss', '1st Order Loss', '2nd Order Loss'])\n",
    "print(df)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Training Record"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_1 = combined_df[(combined_df['source'] == 'history1')]\n",
    "model_2 = combined_df[(combined_df['source'] == 'history2')].groupby(['lambda1']).get_group(0.1)\n",
    "model_3 = combined_df[(combined_df['source'] == 'history3')].groupby(['lambda1', 'lambda2']).get_group((0.1,0.1))\n",
    "model_4 = combined_df[(combined_df['source'] == 'history4')].groupby(['lambda1', 'lambda2']).get_group((0.1,0.1))\n",
    "\n",
    "fig, ax = plt.subplots(2,1)\n",
    "\n",
    "model_color = ['r','b','g','orange']\n",
    "for k, model in enumerate([model_1,model_2,model_3,model_4]):\n",
    "    truncated_end = 20\n",
    "    ax[0].plot(range(truncated_end),model['price_loss'][:truncated_end],color=model_color[k])\n",
    "    ax[1].plot(range(1000),model['first_order_loss'][:1000],color=model_color[k])\n",
    "\n",
    "ax[0].legend([\"model 1\",\"model 2\",\"model 3\",\"model 4\"])\n",
    "ax[1].legend([\"model 1\",\"model 2\",\"model 3\",\"model 4\"])\n",
    "ax[0].set_title(f\"Price Error ($MSE_0$)\")\n",
    "ax[1].set_title(f\"First Order Differentials Error ($MSE_1$)\")\n",
    "ax[0].set_xlabel(\"epoch\")\n",
    "ax[1].set_xlabel(\"epoch\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"results//training_error.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Observations\n",
    "\n",
    "We can see that while attaining similar option prices, Model 2 significantly reduces the first order differential errors. This coheres the results in the original DML paper, stating that the twin network can improve accuracy in pricing and can also be used in financial applications, e.g. hedging, that require accurate approximation of the first order differentials. \n",
    "\n",
    "It also can be seen that by incorporating the second order differentials, Model 3 can attain similar level of pricing accuracy as Model 2, while outperforming in first order differential approximation. The only pitfall is the extra computation of the second order differentials and more combinations of Lagrange coefficients $\\lambda_1, \\lambda_2$ to test. Notice that we carefully separate the connection between first order and second order differentials in the network to avoid over-fitting. The results however still showed improvements. \n",
    "\n",
    "Regarding Diff-PCA, the results show that while reducing the dimension of training targets (second order differentials), Model 4 still performed comparably with Model 3. This verified the use of Diff-PCA. \n",
    "\n",
    "\n",
    "for the Price Error ($MSE_0$), by using more information (first and/or second order differentials), Model 2 and Model 3 converged the slowest. However, by effectively capturing significant signals in the second order differentials, Model 4 converges the fastest, even better than the original Model 1. On the other hand, for $MSE_1$, it can be seen that the use of second order differentials improved the convergence in first order differential predictions, with Model 4 being the best. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Testing Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "feller_data = pd.read_csv('data//feller_d2.csv')\n",
    "feller_testing = feller_data.iloc[9000:]\n",
    "func_names = [\"lm\", \"r\", \"tau\", \"theta\", \"sigma\", \"rho\", \"kappa\", \"v0\"]\n",
    "testing_network_inputs = feller_testing[func_names]\n",
    "testing_option_prices = feller_testing['P_hat']\n",
    "testing_first_order = feller_testing[[f\"diff wrt {col}\" for col in func_names]].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model 1: 0.069223,0.229037\n",
      "Model 2: 0.070697,0.077745\n",
      "Model 3: 0.071351,0.046381\n",
      "Model 4: 0.072941,0.042167\n"
     ]
    }
   ],
   "source": [
    "model1 = load_model(\"models//model1.h5\")\n",
    "predicted_price, predicted_first_order = compute_grad_model1(model1, tf.convert_to_tensor(testing_network_inputs))\n",
    "mse1 = np.average((predicted_first_order-testing_first_order)**2)\n",
    "mse0 = np.average((predicted_price-testing_option_prices)**2)\n",
    "print(f\"Model 1: {mse0:3f},{mse1:3f}\")\n",
    "model2 = load_model(\"models//model2_1.h5\")\n",
    "predicted_price, predicted_first_order = compute_grad_model2(model2, tf.convert_to_tensor(testing_network_inputs))\n",
    "mse1 = np.average((predicted_first_order-testing_first_order)**2)\n",
    "mse0 = np.average((predicted_price-testing_option_prices)**2)\n",
    "print(f\"Model 2: {mse0:3f},{mse1:3f}\")\n",
    "model3 = load_model(\"models//model3_1_5.h5\")\n",
    "predicted_price, predicted_first_order, _ = compute_grad_model3(model3, tf.convert_to_tensor(testing_network_inputs))\n",
    "mse1 = np.average((predicted_first_order-testing_first_order)**2)\n",
    "mse0 = np.average((predicted_price-testing_option_prices)**2)\n",
    "print(f\"Model 3: {mse0:3f},{mse1:3f}\")\n",
    "model4 = load_model(\"models//model4_1_5.h5\")\n",
    "predicted_price, predicted_first_order, _ = compute_grad_model4(model4, tf.convert_to_tensor(testing_network_inputs))\n",
    "mse1 = np.average((predicted_first_order-testing_first_order)**2)\n",
    "mse0 = np.average((predicted_price-testing_option_prices)**2)\n",
    "print(f\"Model 4: {mse0:3f},{mse1:3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that while performing at a similar level of accuracy in option prices, Models 2,3, and 4 can predict the first order differentials accurately, with Model 4 performing the best. The fact that Model 1 prices the best could be because of the smaller dataset in our setting. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n",
      "WARNING:absl:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n"
     ]
    }
   ],
   "source": [
    "def collect_2nd_model(model_num):\n",
    "    table_mse0 = []\n",
    "    table_mse1 = []\n",
    "    lambda_pairs = []\n",
    "    for i in lambda_keys:\n",
    "        for j in lambda_keys:\n",
    "            model = load_model(f\"models//model{model_num}_{int(i*10)}_{int(j*10)}.h5\")\n",
    "            predicted_price, predicted_first_order, _ = compute_grad_model3(model, tf.convert_to_tensor(testing_network_inputs))\n",
    "            mse1 = np.average((predicted_first_order-testing_first_order)**2)\n",
    "            mse0 = np.average((predicted_price-testing_option_prices)**2)\n",
    "            lambda_pairs.append((i,j))\n",
    "            table_mse0.append(mse0)\n",
    "            table_mse1.append(mse1)\n",
    "    return lambda_pairs, np.array(table_mse0), np.array(table_mse1)\n",
    "\n",
    "lambda_paris, mse0_model3, mse1_model3 = collect_2nd_model(3)\n",
    "_, mse0_model4, mse1_model4 = collect_2nd_model(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1,2, figsize=(10,5),sharex=True,sharey=True)\n",
    "vmax = np.max([mse0_model3,mse0_model4])\n",
    "vmin = np.min([mse0_model3,mse0_model4])\n",
    "img1 = ax[0].imshow(mse0_model3.reshape(3,3), vmax=vmax, vmin=vmin)\n",
    "img2 = ax[1].imshow(mse0_model4.reshape(3,3), vmax=vmax, vmin=vmin)\n",
    "data1 = mse0_model3.reshape(3,3)\n",
    "for i in range(data1.shape[0]):\n",
    "    for j in range(data1.shape[1]):\n",
    "        ax[0].text(j, i, f\"{data1[i, j]:.4f}\", ha='center', va='center', color='white', size=15)\n",
    "data1 = mse0_model4.reshape(3,3)\n",
    "for i in range(data1.shape[0]):\n",
    "    for j in range(data1.shape[1]):\n",
    "        ax[1].text(j, i, f\"{data1[i, j]:.4f}\", ha='center', va='center', color='white', size=15)\n",
    "\n",
    "ax[0].set_title(\"Model 3 (without Diff-PCA)\")\n",
    "ax[1].set_title(\"Model 4 (with Diff-PCA)\")\n",
    "ax[0].set_xticks([0,1,2])\n",
    "ax[0].set_xticklabels(['0.1','0.5','0.7'])\n",
    "ax[1].set_xticks([0,1,2])\n",
    "ax[1].set_xticklabels(['0.1','0.5','0.7'])\n",
    "ax[0].set_yticks([0,1,2])\n",
    "ax[0].set_yticklabels(['0.1','0.5','0.7'])\n",
    "ax[1].set_yticks([0,1,2])\n",
    "ax[1].set_yticklabels(['0.1','0.5','0.7'])\n",
    "ax[0].set_xlabel(f'$\\lambda_1$')\n",
    "ax[1].set_xlabel(f'$\\lambda_1$')\n",
    "ax[0].set_ylabel(f'$\\lambda_2$')\n",
    "ax[1].set_ylabel(f'$\\lambda_2$')\n",
    "fig.suptitle(\"Price Testing Error\")\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "fig, ax = plt.subplots(1,2, figsize=(10,5))\n",
    "vmax = np.max([mse1_model3,mse1_model4])\n",
    "vmin = np.min([mse1_model3,mse1_model4])\n",
    "img1 = ax[0].imshow(mse1_model3.reshape(3,3), vmax=vmax, vmin=vmin)\n",
    "img2 = ax[1].imshow(mse1_model4.reshape(3,3), vmax=vmax, vmin=vmin)\n",
    "data1 = mse1_model3.reshape(3,3)\n",
    "for i in range(data1.shape[0]):\n",
    "    for j in range(data1.shape[1]):\n",
    "        ax[0].text(j, i, f\"{data1[i, j]:.4f}\", ha='center', va='center', color='white', size=15)\n",
    "data1 = mse1_model4.reshape(3,3)\n",
    "for i in range(data1.shape[0]):\n",
    "    for j in range(data1.shape[1]):\n",
    "        ax[1].text(j, i, f\"{data1[i, j]:.4f}\", ha='center', va='center', color='white', size=15)\n",
    "\n",
    "ax[0].set_title(\"Model 3 (without Diff-PCA)\")\n",
    "ax[1].set_title(\"Model 4 (with Diff-PCA)\")\n",
    "ax[0].set_xticks([0,1,2])\n",
    "ax[0].set_xticklabels(['0.1','0.5','0.7'])\n",
    "ax[1].set_xticks([0,1,2])\n",
    "ax[1].set_xticklabels(['0.1','0.5','0.7'])\n",
    "ax[0].set_yticks([0,1,2])\n",
    "ax[0].set_yticklabels(['0.1','0.5','0.7'])\n",
    "ax[1].set_yticks([0,1,2])\n",
    "ax[1].set_yticklabels(['0.1','0.5','0.7'])\n",
    "ax[0].set_xlabel(f'$\\lambda_1$')\n",
    "ax[1].set_xlabel(f'$\\lambda_1$')\n",
    "ax[0].set_ylabel(f'$\\lambda_2$')\n",
    "ax[1].set_ylabel(f'$\\lambda_2$')\n",
    "fig.suptitle(\"Delta Testing Error\")\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Again, the use of Diff-PCA marginally impeded the performance of price prediction. However, due to order-reduced training, Model 4 outperformed Model 3 in out-of-sample prediction for first order differentials, implying that the method is suitable for applications that emphasize on accurate modelling on differentials (e.g. hedging)."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "sys-trading",
   "language": "python",
   "name": "sys-trading"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}