NARGP_emukit_example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow.compat.v2 as tf\n",
    "import tensorflow_probability as tfp\n",
    "\n",
    "tfb = tfp.bijectors\n",
    "tfd = tfp.distributions\n",
    "tfk = tfp.math.psd_kernels\n",
    "tf.enable_v2_behavior()\n",
    "\n",
    "from tensorflow_probability.python.math.psd_kernels.internal import util\n",
    "from typing import List\n",
    "\n",
    "# Configure plot defaults\n",
    "plt.rcParams[\"axes.facecolor\"] = \"white\"\n",
    "plt.rcParams[\"grid.color\"] = \"#666666\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(0)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Non-linear Multi-Fidelity Gaussian Processes using TensorFlow Probability\n",
    "\n",
    "In this tutorial, we delve into the implementation of Non-linear Additive Regression Gaussian Processes (NARGP) using TensorFlow Probability (TFP).\n",
    "\n",
    "The foundation for this tutorial stems from the NARGP paper: [Perdikaris 2017](https://royalsocietypublishing.org/doi/10.1098/rspa.2016.0751).\n",
    "\n",
    "Presently, the NARGP implementation can be found in the `emukit` library, which uses on the `GPy` library for Gaussian process.\n",
    "\n",
    "Within this notebook, we will extend TensorFlow Probability's capabilities from standard Gaussian processes to encompass the more advanced NARGP. NARGP stands out as a deep Gaussian process model specifically designed for multi-fidelity modeling.\n",
    "\n",
    "The core concept of NARGP involves an approximation wherein instead of training a single deep GP, we train two distinct Gaussian processes:\n",
    "\n",
    "- Low-Fidelity GP: This GP is responsible for handling the low-fidelity data. Although these data points are of lower quality, they are abundant and more easily obtainable.\n",
    "- High-Fidelity GP: The output from the Low-Fidelity GP serves as input for this GP, which subsequently corrects for the fidelity differences. High-fidelity data points are of superior quality but are typically limited in quantity.\n",
    "\n",
    "This segregation between low and high fidelity data introduces a unique challenge, which is the focus of multi-fidelity modeling: developing a model capable of predicting high-fidelity outputs by synergizing information from both low and high fidelity data sources.\n",
    "\n",
    "Throughout this tutorial, we will:\n",
    "\n",
    "- Guide you step-by-step in adapting TensorFlow Probability's Gaussian process capabilities to construct and train the NARGP model.\n",
    "\n",
    "For theoretical and in-depth understanding of NARGP, I would recommend reading the original Perdikaris paper or reading emukit's website.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# original code\n",
    "#  https://github.com/EmuKit/emukit/blob/main/emukit/test_functions/forrester.py\n",
    "\n",
    "\n",
    "def forrester(x, sd=0):\n",
    "    \"\"\"\n",
    "    Forrester function\n",
    "    :param x: input vector to be evaluated\n",
    "    :param sd: standard deviation of noise parameter\n",
    "    :return: outputs of the function\n",
    "    \"\"\"\n",
    "    x = x.reshape((len(x), 1))\n",
    "    n = x.shape[0]\n",
    "    fval = ((6 * x - 2) ** 2) * np.sin(12 * x - 4)\n",
    "    if sd == 0:\n",
    "        noise = np.zeros(n).reshape(n, 1)\n",
    "    else:\n",
    "        noise = np.random.normal(0, sd, n).reshape(n, 1)\n",
    "    return fval.reshape(n, 1) + noise\n",
    "\n",
    "\n",
    "def forrester_low(x, sd=0):\n",
    "    \"\"\"\n",
    "    Low fidelity forrester function approximation:\n",
    "    :param x: input vector to be evaluated\n",
    "    :param sd: standard deviation of observation noise at low fidelity\n",
    "    :return: outputs of the function\n",
    "    \"\"\"\n",
    "    high_fidelity = forrester(x, 0)\n",
    "    return (\n",
    "        0.5 * high_fidelity\n",
    "        + 10 * (x[:, [0]] - 0.5)\n",
    "        + 5\n",
    "        + np.random.randn(x.shape[0], 1) * sd\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "high_fidelity = forrester\n",
    "low_fidelity = forrester_low\n",
    "\n",
    "x_plot = np.linspace(0, 1, 200)[:, None]\n",
    "y_plot_l = low_fidelity(x_plot)\n",
    "y_plot_h = high_fidelity(x_plot)\n",
    "\n",
    "x_train_l = np.atleast_2d(np.random.rand(40)).T\n",
    "x_train_h = np.atleast_2d(np.random.permutation(x_train_l)[:13])\n",
    "y_train_l = low_fidelity(x_train_l)\n",
    "y_train_h = high_fidelity(x_train_h)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We usually have more HF training points than the LF training points.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((13, 1), (40, 1))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_h.shape, y_train_l.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'High and low fidelity functions')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot the original functions\n",
    "\n",
    "plt.figure(figsize=(12, 8))\n",
    "plt.plot(x_plot, y_plot_l, \"C0\")\n",
    "plt.plot(x_plot, y_plot_h, \"C1\")\n",
    "plt.scatter(x_train_l, y_train_l, color=\"C0\", s=40)\n",
    "plt.scatter(x_train_h, y_train_h, color=\"C1\", s=40)\n",
    "plt.ylabel(\"f (x)\")\n",
    "plt.xlabel(\"x\")\n",
    "plt.legend([\"Low fidelity\", \"High fidelity\"])\n",
    "plt.title(\"High and low fidelity functions\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Normalize with sample mean\n",
    "\n",
    "Gaussian process will always use the zero mean if not specified.\n",
    "It means the GP prediction will regress to the zero mean when there is no data.\n",
    "Thus, this is essential to subtract the sample mean to make sure the GP prior is placed\n",
    "at the center of the data **when we want to use GP to do extrapolation**.\n",
    "\n",
    "Note that it is not so much an issue if we only want the interpolation values.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Normalize with the mean for each fidelity\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# [LF]\n",
    "y_train_l_mean = y_train_l.mean()\n",
    "y_train_l_old = y_train_l\n",
    "y_train_l = y_train_l - y_train_l_mean\n",
    "# [HF]\n",
    "y_train_h_mean = y_train_h.mean()\n",
    "y_train_h_old = y_train_h\n",
    "y_train_h = y_train_h - y_train_h_mean"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multi-fidelity with Non-linear Autoregressive Model\n",
    "\n",
    "The `ActiveDimension` class: Here we define a kernel class to mimic the `active_dims` feature\n",
    "in `GPy`. In `GPy`, `active_dims` means what are the input dimensions you want to consider\n",
    "to pass into the kernel. It is useful when you have multiple input features, but you only want\n",
    "the kernel takes some dimensions of the features.\n",
    "\n",
    "Reference: https://github.com/tensorflow/probability/issues/862\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ActiveDimension(tfk.FeatureTransformed):\n",
    "    def __init__(\n",
    "        self,\n",
    "        kernel,\n",
    "        active_dim: List,\n",
    "        validate_args=False,\n",
    "        parameters=None,\n",
    "        name=\"ActiveDim\",\n",
    "    ) -> None:\n",
    "        self._transformation_fn = (\n",
    "            lambda x, feature_ndims, param_expansion_ndims: tf.gather(\n",
    "                x,\n",
    "                active_dim,\n",
    "                axis=-1,\n",
    "            )\n",
    "        )\n",
    "\n",
    "        super(ActiveDimension, self).__init__(\n",
    "            kernel=kernel,\n",
    "            transformation_fn=self._transformation_fn,\n",
    "            validate_args=validate_args,\n",
    "            parameters=parameters,\n",
    "            name=name,\n",
    "        )\n",
    "\n",
    "        # Keep the kernel so that we can get the batch shape\n",
    "        self._kernel = kernel\n",
    "\n",
    "    @property\n",
    "    def batch_shape(self):\n",
    "        return self._kernel.batch_shape\n",
    "\n",
    "    def batch_shape_tensor(self):\n",
    "        return self._kernel.batch_shape_tensor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "NARGP is a composite kernel. We want the input x passed to kernel with `base_dims_list`,\n",
    "and the output from low-fidelity Gaussian process passed to `out_dims_list`.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_input_dims = 1  # this is a single input (x) problem\n",
    "n_output_dim = 1  # this is a single output (y) problem\n",
    "base_dims_list = list(range(n_input_dims))\n",
    "out_dims_list = list(range(n_input_dims, n_input_dims + n_output_dim))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_gp_lf(amplitude, length_scale, observation_noise_variance):\n",
    "    \"\"\"\n",
    "    Low-fidelity:\n",
    "    Defines the conditional dist. of GP outputs, given kernel parameters.\n",
    "    \"\"\"\n",
    "\n",
    "    # Create the covariance kernel, which will be shared between the prior (which we\n",
    "    # use for maximum likelihood training) and the posterior (which we use for\n",
    "    # posterior predictive sampling)\n",
    "    kernel = tfk.ExponentiatedQuadratic(amplitude, length_scale)\n",
    "\n",
    "    # Create the GP prior distribution, which we will use to train the model\n",
    "    # parameters.\n",
    "    return tfd.GaussianProcess(\n",
    "        kernel=kernel,\n",
    "        index_points=x_train_l,\n",
    "        observation_noise_variance=observation_noise_variance,\n",
    "    )\n",
    "\n",
    "\n",
    "# 1: low-fidelity; 0: high-fidelity\n",
    "def build_gp_nargp(\n",
    "    scale_amplitude_0,  # HF\n",
    "    scale_length_scale_0,  # HF\n",
    "    interaction_amplitude_0,  # HF\n",
    "    interaction_length_scale_0,  # HF\n",
    "    bias_amplitude_0,  # HF\n",
    "    bias_length_scale_0,  # HF\n",
    "    linear_bias_amplitude_0,  # HF\n",
    "    linear_slope_amplitude_0,  # HF\n",
    "    linear_shift_0,  # HF\n",
    "    amplitude_1,  # LF\n",
    "    length_scale_1,  # LF\n",
    "    observation_noise_variance_0,  # HF\n",
    "    observation_noise_variance_1,  # LF\n",
    "):\n",
    "    \"\"\"\n",
    "    High-fidelity:\n",
    "    Defines the conditional dist. of GP outputs, given kernel parameters.\n",
    "\n",
    "    NARGP models the high-fidelity as\n",
    "        f_HF = rho( f_LF, x ) + delta(x)\n",
    "    which can be simplified as\n",
    "        f_HF = g( f_LF, x )\n",
    "    with a composite kernel\n",
    "        K_g = K_rho * K_interaction + K_bias\n",
    "    \"\"\"\n",
    "    # HF kernel\n",
    "    scale_kernel = ActiveDimension(\n",
    "        tfk.ExponentiatedQuadratic(scale_amplitude_0, scale_length_scale_0),\n",
    "        active_dim=out_dims_list,\n",
    "    )\n",
    "    scale_kernel += ActiveDimension(\n",
    "        tfk.Linear(\n",
    "            bias_amplitude=linear_bias_amplitude_0,\n",
    "            slope_amplitude=linear_slope_amplitude_0,\n",
    "            shift=linear_shift_0,\n",
    "        ),\n",
    "        active_dim=out_dims_list,\n",
    "    )\n",
    "    interaction_kernel = ActiveDimension(\n",
    "        tfk.ExponentiatedQuadratic(interaction_amplitude_0, interaction_length_scale_0),\n",
    "        active_dim=base_dims_list,\n",
    "    )\n",
    "    bias_kernel = ActiveDimension(\n",
    "        tfk.ExponentiatedQuadratic(bias_amplitude_0, bias_length_scale_0),\n",
    "        active_dim=base_dims_list,\n",
    "    )\n",
    "    # NARGP kernel\n",
    "    kernel_h = interaction_kernel * scale_kernel + bias_kernel\n",
    "\n",
    "    # LF kernel\n",
    "    kernel_l1 = tfk.ExponentiatedQuadratic(amplitude_1, length_scale_1)\n",
    "    ## Mean predictions from the LF GP\n",
    "    gprm_l1 = tfd.GaussianProcessRegressionModel(\n",
    "        kernel=kernel_l1,\n",
    "        index_points=x_train_h,\n",
    "        observation_index_points=x_train_l,\n",
    "        observations=y_train_l[:, 0],\n",
    "        observation_noise_variance=observation_noise_variance_1,\n",
    "        predictive_noise_variance=0.0,\n",
    "    )\n",
    "    m1 = gprm_l1.mean()\n",
    "\n",
    "    ## Index points for H include [X, mean from L1]\n",
    "    index_points_h = tf.concat([x_train_h, m1[:, None]], axis=1)\n",
    "\n",
    "    # Create the GP prior distribution, which we will use to train the model\n",
    "    # parameters.\n",
    "    return tfd.GaussianProcess(\n",
    "        kernel=kernel_h,\n",
    "        index_points=index_points_h,\n",
    "        observation_noise_variance=observation_noise_variance_0,\n",
    "    )\n",
    "\n",
    "\n",
    "# Define the joint distribution of low-fidelity GP\n",
    "gp_joint_model_lf = tfd.JointDistributionNamed(\n",
    "    {\n",
    "        \"amplitude\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"length_scale\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"observation_noise_variance\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"observations\": build_gp_lf,\n",
    "    }\n",
    ")\n",
    "# Define the joint distribution of NARGP\n",
    "gp_joint_model_hf = tfd.JointDistributionNamed(\n",
    "    {\n",
    "        \"scale_amplitude_0\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"scale_length_scale_0\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"interaction_amplitude_0\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"interaction_length_scale_0\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"bias_amplitude_0\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"bias_length_scale_0\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"linear_bias_amplitude_0\": tfd.LogNormal(\n",
    "            loc=0.0, scale=np.float64(1.0)\n",
    "        ),  # linear\n",
    "        \"linear_slope_amplitude_0\": tfd.LogNormal(\n",
    "            loc=0.0, scale=np.float64(1.0)\n",
    "        ),  # linear\n",
    "        \"linear_shift_0\": tfd.Normal(loc=0.0, scale=np.float64(1.0)),  # linear\n",
    "        \"amplitude_1\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"length_scale_1\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"observation_noise_variance_0\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"observation_noise_variance_1\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"observations\": build_gp_nargp,\n",
    "    }\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /Users/jibanmac/Documents/conda/tensorflow-loaf/lib/python3.9/site-packages/tensorflow_probability/python/distributions/distribution.py:342: calling GaussianProcess.__init__ (from tensorflow_probability.python.distributions.gaussian_process) with jitter is deprecated and will be removed after 2021-05-10.\n",
      "Instructions for updating:\n",
      "`jitter` is deprecated; please use `marginal_fn` directly.\n",
      "WARNING:tensorflow:From /Users/jibanmac/Documents/conda/tensorflow-loaf/lib/python3.9/site-packages/tensorflow/python/util/deprecation.py:576: calling GaussianProcess.__init__ (from tensorflow_probability.python.distributions.gaussian_process) with always_yield_multivariate_normal is deprecated and will be removed after 2023-07-01.\n",
      "Instructions for updating:\n",
      "`always_yield_multivariate_normal` is deprecated. This arg is now ignoredand will be removed after 2023-07-01. A `GaussianProcess` evaluated at asingle index point now always has event shape `[1]` (the previous behaviorfor `always_yield_multivariate_normal=True`). To reproduce the previous behavior of `always_yield_multivariate_normal=False`, squeeze the rightmost singleton dimension from the output of `mean`, `sample`, etc.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /Users/jibanmac/Documents/conda/tensorflow-loaf/lib/python3.9/site-packages/tensorflow_probability/python/distributions/distribution.py:342: calling GaussianProcessRegressionModel.__init__ (from tensorflow_probability.python.distributions.gaussian_process_regression_model) with always_yield_multivariate_normal is deprecated and will be removed after 2023-07-01.\n",
      "Instructions for updating:\n",
      "`always_yield_multivariate_normal` is deprecated. This arg is now ignoredand will be removed after 2023-07-01. A `GaussianProcessRegressionModel`evaluated at a single index point now always has event shape `[1]` (theprevious behavior for `always_yield_multivariate_normal=True`). Toreproduce the previous behavior of`always_yield_multivariate_normal=False`, squeeze the rightmost singletondimension from the output of `mean`, `sample`, etc.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "({'observation_noise_variance': TensorShape([]),\n",
       "  'length_scale': TensorShape([]),\n",
       "  'amplitude': TensorShape([]),\n",
       "  'observations': TensorShape([])},\n",
       " {'scale_length_scale_0': TensorShape([]),\n",
       "  'scale_amplitude_0': TensorShape([]),\n",
       "  'observation_noise_variance_1': TensorShape([]),\n",
       "  'observation_noise_variance_0': TensorShape([]),\n",
       "  'linear_slope_amplitude_0': TensorShape([]),\n",
       "  'linear_shift_0': TensorShape([]),\n",
       "  'linear_bias_amplitude_0': TensorShape([]),\n",
       "  'length_scale_1': TensorShape([]),\n",
       "  'interaction_length_scale_0': TensorShape([]),\n",
       "  'interaction_amplitude_0': TensorShape([]),\n",
       "  'bias_length_scale_0': TensorShape([]),\n",
       "  'bias_amplitude_0': TensorShape([]),\n",
       "  'amplitude_1': TensorShape([]),\n",
       "  'observations': TensorShape([])})"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Test the batch shape\n",
    "gp_joint_model_lf.batch_shape, gp_joint_model_hf.batch_shape"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set the learnable parameters and train based on the maximum marginal likelihoods of two GPs.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create the trainable model parameters, which we'll subsequently optimize.\n",
    "# Note that we constrain them to be strictly positive.\n",
    "constrain_positive = tfb.Shift(np.finfo(np.float64).tiny)(tfb.Exp())\n",
    "# identify bijector for parameters range from -infty to infty\n",
    "identity = tfb.Identity()\n",
    "\n",
    "\n",
    "scale_amplitude_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"scale_amplitude_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "scale_length_scale_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"scale_length_scale_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "interaction_amplitude_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"interaction_amplitude_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "interaction_length_scale_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"interaction_length_scale_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "bias_amplitude_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"bias_amplitude_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "bias_length_scale_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"bias_length_scale_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "\n",
    "# Linear kernel\n",
    "linear_bias_amplitude_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"linear_bias_amplitude_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "linear_slope_amplitude_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"linear_slope_amplitude_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "linear_shift_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=np.float64(0.0),\n",
    "    bijector=identity,\n",
    "    name=\"linear_shift_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "\n",
    "\n",
    "# L1 parameters\n",
    "amplitude_var_1 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0, bijector=constrain_positive, name=\"amplitude_1\", dtype=np.float64\n",
    ")\n",
    "\n",
    "length_scale_var_1 = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"length_scale_1\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "\n",
    "# Noise\n",
    "observation_noise_variance_var_0 = tfp.util.TransformedVariable(\n",
    "    initial_value=1e-5,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"observation_noise_variance_var_0\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "observation_noise_variance_var_1 = tfp.util.TransformedVariable(\n",
    "    initial_value=1e-5,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"observation_noise_variance_var_1\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "\n",
    "# Store the trainable parameters for later use\n",
    "# [NARGP]\n",
    "trainable_variables = [\n",
    "    v.trainable_variables[0]\n",
    "    for v in [\n",
    "        scale_amplitude_var_0,\n",
    "        scale_length_scale_var_0,\n",
    "        interaction_amplitude_var_0,\n",
    "        interaction_length_scale_var_0,\n",
    "        bias_amplitude_var_0,\n",
    "        bias_length_scale_var_0,\n",
    "        observation_noise_variance_var_0,\n",
    "        linear_bias_amplitude_var_0,\n",
    "        linear_slope_amplitude_var_0,\n",
    "        linear_shift_var_0,\n",
    "    ]\n",
    "]\n",
    "\n",
    "# [LF only]\n",
    "trainable_variables_LF = [\n",
    "    v.trainable_variables[0]\n",
    "    for v in [\n",
    "        amplitude_var_1,\n",
    "        length_scale_var_1,\n",
    "        observation_noise_variance_var_1,\n",
    "    ]\n",
    "]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's test if we can compute the log prob\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "log_prob of sample: -70.31736004732093\n"
     ]
    }
   ],
   "source": [
    "x_l = gp_joint_model_lf.sample()\n",
    "lp_l = gp_joint_model_lf.log_prob(x_l)\n",
    "print(\"log_prob of sample: {}\".format(lp_l))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "log_prob of sample: -31.325224402091454\n"
     ]
    }
   ],
   "source": [
    "x_h = gp_joint_model_hf.sample()\n",
    "lp_h = gp_joint_model_hf.log_prob(x_h)\n",
    "\n",
    "print(\"log_prob of sample: {}\".format(lp_h))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To condition the model on our observed data, we'll define a `target_log_prob` function, which takes the (still to be inferred) kernel hyperparameters.\n",
    "\n",
    "For NARGP, we train LF and HF separately. When training HF, we expect the LF is already trained.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def target_log_prob_LF(\n",
    "    amplitude_1,\n",
    "    length_scale_1,\n",
    "    observation_noise_variance_1,\n",
    "):\n",
    "    \"\"\"\n",
    "    Loglike(LF)\n",
    "    \"\"\"\n",
    "\n",
    "    log_prob_lf = gp_joint_model_lf.log_prob(\n",
    "        {\n",
    "            \"amplitude\": amplitude_1,\n",
    "            \"length_scale\": length_scale_1,\n",
    "            \"observation_noise_variance\": observation_noise_variance_1,\n",
    "            \"observations\": y_train_l[:, 0],\n",
    "        }\n",
    "    )\n",
    "\n",
    "    return log_prob_lf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def target_log_prob(\n",
    "    scale_amplitude_0,  # HF\n",
    "    scale_length_scale_0,  # HF\n",
    "    interaction_amplitude_0,  # HF\n",
    "    interaction_length_scale_0,  # HF\n",
    "    bias_amplitude_0,  # HF\n",
    "    bias_length_scale_0,  # HF\n",
    "    observation_noise_variance_0,\n",
    "    linear_bias_amplitude_0,\n",
    "    linear_slope_amplitude_0,\n",
    "    linear_shift_0,\n",
    "    amplitude_1,\n",
    "    length_scale_1,\n",
    "    observation_noise_variance_1,\n",
    "):\n",
    "    \"\"\"\n",
    "    Loglike(HF)\n",
    "    \"\"\"\n",
    "\n",
    "    log_prob_hf = gp_joint_model_hf.log_prob(\n",
    "        {\n",
    "            \"scale_amplitude_0\": scale_amplitude_0,\n",
    "            \"scale_length_scale_0\": scale_length_scale_0,\n",
    "            \"interaction_amplitude_0\": interaction_amplitude_0,\n",
    "            \"interaction_length_scale_0\": interaction_length_scale_0,\n",
    "            \"bias_amplitude_0\": bias_amplitude_0,\n",
    "            \"bias_length_scale_0\": bias_length_scale_0,\n",
    "            \"linear_bias_amplitude_0\": linear_bias_amplitude_0,\n",
    "            \"linear_slope_amplitude_0\": linear_slope_amplitude_0,\n",
    "            \"linear_shift_0\": linear_shift_0,\n",
    "            \"amplitude_1\": amplitude_1,\n",
    "            \"length_scale_1\": length_scale_1,\n",
    "            \"observation_noise_variance_0\": observation_noise_variance_0,\n",
    "            \"observation_noise_variance_1\": observation_noise_variance_1,\n",
    "            \"observations\": y_train_h[:, 0],\n",
    "        }\n",
    "    )\n",
    "\n",
    "    return log_prob_hf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training Strategy\n",
    "\n",
    "Our training approach comprises two distinct steps, each designed to enhance the efficiency and accuracy of our model's learning process.\n",
    "\n",
    "- Step 1: Initial Training with Controlled Observational Noise\n",
    "\n",
    "In the first step, we strategically set the observational_noise parameter to a low value. By doing so, we mitigate the risk of this parameter dominating the likelihood estimation. This is particularly crucial when dealing with potential noise interference that might hinder the optimal learning of kernel parameters.\n",
    "\n",
    "- Step 2: Kernel Hyperparameter Optimization with Noise\n",
    "\n",
    "Having established a foundation in the initial step, we proceed to the second stage of training. Here, we relax the previous constraint and enable the training process to account for the observational noise as well.\n",
    "\n",
    "In this way, our two-step training strategy optimizes the learning process of both kernel hyperparameters and observational noise.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Gradients do not exist for variables ['observation_noise_variance_var_1:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss` argument?\n",
      "WARNING:tensorflow:Gradients do not exist for variables ['observation_noise_variance_var_1:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss` argument?\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-08-13 18:12:36.700800: W tensorflow/tsl/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n"
     ]
    }
   ],
   "source": [
    "# Now we optimize the model parameters.\n",
    "num_iters = 5000\n",
    "optimizer = tf.keras.optimizers.legacy.Adam(learning_rate=5e-3)\n",
    "\n",
    "# Fix the noise will increase the stability of the hyperkernel parameters training\n",
    "fixed_noise = 1e-7\n",
    "\n",
    "\n",
    "# [Low-fidelity] Fixed noise\n",
    "@tf.function(autograph=False, jit_compile=False)\n",
    "def train_model_LFonly_fixed():\n",
    "    with tf.GradientTape() as tape:\n",
    "        loss = -target_log_prob_LF(\n",
    "            amplitude_1=amplitude_var_1,\n",
    "            length_scale_1=length_scale_var_1,\n",
    "            observation_noise_variance_1=fixed_noise,\n",
    "        )\n",
    "    grads = tape.gradient(loss, trainable_variables_LF)\n",
    "    optimizer.apply_gradients(zip(grads, trainable_variables_LF))\n",
    "    return loss\n",
    "\n",
    "\n",
    "# [LF only] Fixed noise\n",
    "lls_noise_lfonly_fixed = np.zeros(num_iters, np.float64)\n",
    "for i in range(num_iters):\n",
    "    loss = train_model_LFonly_fixed()\n",
    "    lls_noise_lfonly_fixed[i] = loss\n",
    "\n",
    "num_iters = 10000\n",
    "optimizer = tf.keras.optimizers.legacy.Adam(learning_rate=5e-3)\n",
    "\n",
    "\n",
    "# [Low-fidelity] Relaxed noise\n",
    "@tf.function(autograph=False, jit_compile=False)\n",
    "def train_model_LFonly():\n",
    "    with tf.GradientTape() as tape:\n",
    "        loss = -target_log_prob_LF(\n",
    "            amplitude_1=amplitude_var_1,\n",
    "            length_scale_1=length_scale_var_1,\n",
    "            observation_noise_variance_1=observation_noise_variance_var_1,\n",
    "        )\n",
    "    grads = tape.gradient(loss, trainable_variables_LF)\n",
    "    optimizer.apply_gradients(zip(grads, trainable_variables_LF))\n",
    "    return loss\n",
    "\n",
    "\n",
    "# [LF only] Relaxed noise\n",
    "lls_noise_lfonly = np.zeros(num_iters, np.float64)\n",
    "for i in range(num_iters):\n",
    "    loss = train_model_LFonly()\n",
    "    lls_noise_lfonly[i] = loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Once we have the LF GP trained, the get the maximum a posteriori values of LF GP, and pass\n",
    "to the NARGP kernel.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Gradients do not exist for variables ['observation_noise_variance_var_0:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss` argument?\n",
      "WARNING:tensorflow:Gradients do not exist for variables ['observation_noise_variance_var_0:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss` argument?\n",
      "Trained parameters:\n",
      "LF amplitude: 7.696185923991638\n",
      "LF length_scale: 0.18907100395180312\n",
      "HF scale amplitude: 0.0981308073047468\n",
      "HF scale length_scale: 0.8847643712588855\n",
      "HF interaction amplitude: 1.6108405755760362\n",
      "HF interaction length_scale: 0.5948590397890419\n",
      "HF bias amplitude: 0.3739050984903757\n",
      "HF bias length_scale: 0.3745649867452564\n",
      "Linear kernel:\n",
      "----\n",
      "HF linear bias amplitude: 3.318865418538283\n",
      "HF linear slope amplitude: 0.669373593296993\n",
      "HF linear shift: 0.18675208993630918\n",
      "Noise:\n",
      "----\n",
      "HF observation_noise_variance: 0.02172950365781868\n",
      "LF observation_noise_variance: 6.343398621379098e-06\n"
     ]
    }
   ],
   "source": [
    "# [Fix Low-fidelity node] Perdikaris et al.\n",
    "# So the likelihood from the LF node is fixed.\n",
    "## L1\n",
    "fixed_noise_L1 = observation_noise_variance_var_1._value().numpy()\n",
    "fixed_amplitude_1 = amplitude_var_1._value().numpy()\n",
    "fixed_length_scale_1 = length_scale_var_1._value().numpy()\n",
    "\n",
    "\n",
    "# Now we optimize the model parameters.\n",
    "num_iters = 5000\n",
    "optimizer = tf.keras.optimizers.legacy.Adam(learning_rate=5e-3)\n",
    "\n",
    "\n",
    "# Fixed noise\n",
    "@tf.function(autograph=False, jit_compile=False)\n",
    "def train_model_fixed_noise():\n",
    "    with tf.GradientTape() as tape:\n",
    "        loss = -target_log_prob(\n",
    "            scale_amplitude_0=scale_amplitude_var_0,\n",
    "            scale_length_scale_0=scale_length_scale_var_0,\n",
    "            interaction_amplitude_0=interaction_amplitude_var_0,\n",
    "            interaction_length_scale_0=interaction_length_scale_var_0,\n",
    "            bias_amplitude_0=bias_amplitude_var_0,\n",
    "            bias_length_scale_0=bias_length_scale_var_0,\n",
    "            linear_bias_amplitude_0=linear_bias_amplitude_var_0,\n",
    "            linear_slope_amplitude_0=linear_slope_amplitude_var_0,\n",
    "            linear_shift_0=linear_shift_var_0,\n",
    "            amplitude_1=fixed_amplitude_1,\n",
    "            length_scale_1=fixed_length_scale_1,\n",
    "            observation_noise_variance_0=fixed_noise,\n",
    "            observation_noise_variance_1=fixed_noise_L1,\n",
    "        )\n",
    "    grads = tape.gradient(loss, trainable_variables)\n",
    "    optimizer.apply_gradients(zip(grads, trainable_variables))\n",
    "    return loss\n",
    "\n",
    "\n",
    "# Fixed noise\n",
    "lls_noise_fixed = np.zeros(num_iters, np.float64)\n",
    "for i in range(num_iters):\n",
    "    loss = train_model_fixed_noise()\n",
    "    lls_noise_fixed[i] = loss\n",
    "\n",
    "# Now we optimize the model parameters.\n",
    "num_iters = 10000\n",
    "optimizer = tf.keras.optimizers.legacy.Adam(learning_rate=3e-3)\n",
    "\n",
    "\n",
    "# Not fix noise\n",
    "# Use `tf.function` to trace the loss for more efficient evaluation.\n",
    "@tf.function(autograph=False, jit_compile=False)\n",
    "def train_model():\n",
    "    with tf.GradientTape() as tape:\n",
    "        loss = -target_log_prob(\n",
    "            scale_amplitude_0=scale_amplitude_var_0,\n",
    "            scale_length_scale_0=scale_length_scale_var_0,\n",
    "            interaction_amplitude_0=interaction_amplitude_var_0,\n",
    "            interaction_length_scale_0=interaction_length_scale_var_0,\n",
    "            bias_amplitude_0=bias_amplitude_var_0,\n",
    "            bias_length_scale_0=bias_length_scale_var_0,\n",
    "            linear_bias_amplitude_0=linear_bias_amplitude_var_0,\n",
    "            linear_slope_amplitude_0=linear_slope_amplitude_var_0,\n",
    "            linear_shift_0=linear_shift_var_0,\n",
    "            amplitude_1=fixed_amplitude_1,\n",
    "            length_scale_1=fixed_length_scale_1,\n",
    "            observation_noise_variance_0=observation_noise_variance_var_0,\n",
    "            observation_noise_variance_1=fixed_noise_L1,\n",
    "        )\n",
    "    grads = tape.gradient(loss, trainable_variables)\n",
    "    optimizer.apply_gradients(zip(grads, trainable_variables))\n",
    "    return loss\n",
    "\n",
    "\n",
    "# Relaxed noise\n",
    "lls_ = np.zeros(num_iters, np.float64)\n",
    "for i in range(num_iters):\n",
    "    loss = train_model()\n",
    "    lls_[i] = loss\n",
    "\n",
    "\n",
    "print(\"Trained parameters:\")\n",
    "print(\"LF amplitude: {}\".format(amplitude_var_1._value().numpy()))\n",
    "print(\"LF length_scale: {}\".format(length_scale_var_1._value().numpy()))\n",
    "print(\"HF scale amplitude: {}\".format(scale_amplitude_var_0._value().numpy()))\n",
    "print(\"HF scale length_scale: {}\".format(scale_length_scale_var_0._value().numpy()))\n",
    "print(\n",
    "    \"HF interaction amplitude: {}\".format(interaction_amplitude_var_0._value().numpy())\n",
    ")\n",
    "print(\n",
    "    \"HF interaction length_scale: {}\".format(\n",
    "        interaction_length_scale_var_0._value().numpy()\n",
    "    )\n",
    ")\n",
    "print(\"HF bias amplitude: {}\".format(bias_amplitude_var_0._value().numpy()))\n",
    "print(\"HF bias length_scale: {}\".format(bias_length_scale_var_0._value().numpy()))\n",
    "\n",
    "print(\"Linear kernel:\")\n",
    "print(\"----\")\n",
    "print(\n",
    "    \"HF linear bias amplitude: {}\".format(linear_bias_amplitude_var_0._value().numpy())\n",
    ")\n",
    "print(\n",
    "    \"HF linear slope amplitude: {}\".format(\n",
    "        linear_slope_amplitude_var_0._value().numpy()\n",
    "    )\n",
    ")\n",
    "print(\"HF linear shift: {}\".format(linear_shift_var_0._value().numpy()))\n",
    "\n",
    "print(\"Noise:\")\n",
    "print(\"----\")\n",
    "print(\n",
    "    \"HF observation_noise_variance: {}\".format(\n",
    "        observation_noise_variance_var_0._value().numpy()\n",
    "    )\n",
    ")\n",
    "print(\n",
    "    \"LF observation_noise_variance: {}\".format(\n",
    "        observation_noise_variance_var_1._value().numpy()\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x400 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the loss evolution\n",
    "fig, ax = plt.subplots(1, 4, figsize=(16, 4))\n",
    "\n",
    "ax[0].plot(lls_noise_lfonly_fixed)\n",
    "ax[0].set_title(\"Low-fidelity: Fix noise\")\n",
    "ax[0].set_xlabel(\"Training iteration\")\n",
    "ax[0].set_ylabel(\"Log marginal likelihood\")\n",
    "\n",
    "ax[1].plot(lls_noise_lfonly)\n",
    "ax[1].set_title(\"Low-fidelity: Train noise\")\n",
    "ax[1].set_xlabel(\"Training iteration\")\n",
    "ax[1].set_ylabel(\"Log marginal likelihood\")\n",
    "\n",
    "ax[2].plot(lls_noise_fixed)\n",
    "ax[2].set_title(\"High-fidelity: Fix noise\")\n",
    "ax[2].set_xlabel(\"Training iteration\")\n",
    "ax[2].set_ylabel(\"Log marginal likelihood\")\n",
    "\n",
    "ax[3].plot(lls_)\n",
    "ax[3].set_title(\"High-fidelity: Train noise\")\n",
    "ax[3].set_xlabel(\"Training iteration\")\n",
    "ax[3].set_ylabel(\"Log marginal likelihood\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predictions\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Having trained the model, we'd like to sample from the posterior conditioned\n",
    "# on observations. We'd like the samples to be at points other than the training\n",
    "# inputs.\n",
    "predictive_index_points_ = np.linspace(0, 1, 200, dtype=np.float64)\n",
    "# Reshape to [200, 1] -- 1 is the dimensionality of the feature space.\n",
    "predictive_index_points_ = predictive_index_points_[..., np.newaxis]\n",
    "\n",
    "\n",
    "# HF kernel\n",
    "optimized_scale_kernel = ActiveDimension(\n",
    "    tfk.ExponentiatedQuadratic(scale_amplitude_var_0, scale_length_scale_var_0),\n",
    "    active_dim=out_dims_list,\n",
    ")\n",
    "optimized_scale_kernel += ActiveDimension(\n",
    "    tfk.Linear(\n",
    "        bias_amplitude=linear_bias_amplitude_var_0,\n",
    "        slope_amplitude=linear_slope_amplitude_var_0,\n",
    "        shift=linear_shift_var_0,\n",
    "    ),\n",
    "    active_dim=out_dims_list,\n",
    ")\n",
    "\n",
    "optimized_interaction_kernel = ActiveDimension(\n",
    "    tfk.ExponentiatedQuadratic(\n",
    "        interaction_amplitude_var_0, interaction_length_scale_var_0\n",
    "    ),\n",
    "    active_dim=base_dims_list,\n",
    ")\n",
    "optimized_bias_kernel = ActiveDimension(\n",
    "    tfk.ExponentiatedQuadratic(bias_amplitude_var_0, bias_length_scale_var_0),\n",
    "    active_dim=base_dims_list,\n",
    ")\n",
    "# NARGP kernel\n",
    "optimized_kernel_h = (\n",
    "    optimized_interaction_kernel * optimized_scale_kernel + optimized_bias_kernel\n",
    ")\n",
    "\n",
    "\n",
    "# LF kernel\n",
    "optimized_kernel_1 = tfk.ExponentiatedQuadratic(amplitude_var_1, length_scale_var_1)\n",
    "\n",
    "## Passing information from low-fidelity node\n",
    "# L1 index points\n",
    "index_optimized_gprm_l1 = tfd.GaussianProcessRegressionModel(\n",
    "    kernel=optimized_kernel_1,\n",
    "    index_points=x_train_h,\n",
    "    observation_index_points=x_train_l,\n",
    "    observations=y_train_l[:, 0],\n",
    "    observation_noise_variance=observation_noise_variance_var_1,\n",
    "    predictive_noise_variance=np.float64(0.0),\n",
    ")\n",
    "# L1 predictive points\n",
    "optimized_gprm_l1 = tfd.GaussianProcessRegressionModel(\n",
    "    kernel=optimized_kernel_1,\n",
    "    index_points=predictive_index_points_,\n",
    "    observation_index_points=x_train_l,\n",
    "    observations=y_train_l[:, 0],\n",
    "    observation_noise_variance=observation_noise_variance_var_1,\n",
    "    predictive_noise_variance=np.float64(0.0),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To properly capture the uncertainty in the LF GP, we do a Monte Carlo sampling in the LF\n",
    "GP to create multiple index points for NARGP kernel.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Combine information\n",
    "## Using Irene Ji's full covariance method\n",
    "num_samples = 200\n",
    "ntest = predictive_index_points_.shape[0]\n",
    "\n",
    "index_m1 = index_optimized_gprm_l1.sample(num_samples)\n",
    "m1 = optimized_gprm_l1.sample(num_samples)\n",
    "\n",
    "tmp_m = np.zeros((num_samples, ntest))\n",
    "tmp_v = np.zeros((num_samples, ntest))\n",
    "tmp_sample = np.zeros((num_samples, ntest))\n",
    "# This way you also account for the uncertainty from L1 node\n",
    "for j in range(num_samples):\n",
    "    index_points_h = tf.concat([x_train_h, index_m1[j, :][:, None]], axis=1)\n",
    "    predictive_index_points_h = tf.concat(\n",
    "        [predictive_index_points_, m1[j, :][:, None]], axis=1\n",
    "    )\n",
    "\n",
    "    # Create the GP prior distribution, which we will use to train the model\n",
    "    # parameters.\n",
    "    nargp_gprm = tfd.GaussianProcessRegressionModel(\n",
    "        kernel=optimized_kernel_h,\n",
    "        index_points=predictive_index_points_h,\n",
    "        observation_index_points=index_points_h,\n",
    "        observations=y_train_h[:, 0],\n",
    "        observation_noise_variance=observation_noise_variance_var_0,\n",
    "        predictive_noise_variance=np.float64(0.0),\n",
    "    )\n",
    "\n",
    "    tmp_m[j, :] = nargp_gprm.mean().numpy()\n",
    "    tmp_v[j, :] = nargp_gprm.variance().numpy()\n",
    "    tmp_sample[j, :] = nargp_gprm.sample().numpy()\n",
    "\n",
    "mu_final = np.mean(tmp_m, axis=0)\n",
    "v_final = np.mean(tmp_v, axis=0) + np.var(tmp_m, axis=0)\n",
    "y_pred = mu_final[:, None]\n",
    "var_pred = np.abs(v_final[:, None])\n",
    "\n",
    "# LF only\n",
    "l1_gprm = tfd.GaussianProcessRegressionModel(\n",
    "    kernel=optimized_kernel_1,\n",
    "    index_points=predictive_index_points_,\n",
    "    observation_index_points=x_train_l,\n",
    "    observations=y_train_l[:, 0],\n",
    "    observation_noise_variance=observation_noise_variance_var_1,\n",
    "    predictive_noise_variance=np.float64(0.0),\n",
    ")\n",
    "\n",
    "l1_samples = l1_gprm.sample(num_samples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/sn/b7hd1bmd569cjwgmtww4mv440000gn/T/ipykernel_80537/459058000.py:46: MatplotlibDeprecationWarning: The legendHandles attribute was deprecated in Matplotlib 3.7 and will be removed two minor releases later. Use legend_handles instead.\n",
      "  for lh in leg.legendHandles:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the true function, observations, and posterior samples.\n",
    "plt.figure(figsize=(12, 8))\n",
    "\n",
    "# NARGP\n",
    "for i in range(num_samples):\n",
    "    plt.plot(\n",
    "        predictive_index_points_,\n",
    "        tmp_sample[i, :] + y_train_h_mean,\n",
    "        c=\"C2\",\n",
    "        alpha=0.1,\n",
    "        label=\"Posterior Sample\" if i == 0 else None,\n",
    "    )\n",
    "# Posterior mean - the uncertainty passed using Monte Carlo sampling\n",
    "plt.plot(\n",
    "    predictive_index_points_,\n",
    "    y_pred + y_train_h_mean,\n",
    "    c=\"C4\",\n",
    "    label=\"Posterior Mean\",\n",
    ")\n",
    "\n",
    "# HF\n",
    "for i in range(num_samples):\n",
    "    plt.plot(\n",
    "        predictive_index_points_,\n",
    "        l1_samples[i, :] + y_train_l_mean,\n",
    "        c=\"C3\",\n",
    "        alpha=0.05,\n",
    "        label=\"L1 Posterior Sample\" if i == 0 else None,\n",
    "    )\n",
    "\n",
    "plt.plot(\n",
    "    predictive_index_points_,\n",
    "    np.median(l1_samples, axis=0) + y_train_l_mean,\n",
    "    c=\"C3\",\n",
    "    label=\"L1 Median Posterior\",\n",
    ")\n",
    "\n",
    "plt.plot(x_plot, y_plot_h, label=\"True fn (HF)\", color=\"C1\")\n",
    "plt.plot(x_plot, y_plot_l, label=\"True fn (LF)\", color=\"C0\")\n",
    "\n",
    "plt.scatter(x_train_l, y_train_l_old, color=\"C0\", s=40, label=\"Observations (LF)\")\n",
    "plt.scatter(x_train_h, y_train_h_old, color=\"C1\", s=40, label=\"Observations (HF)\")\n",
    "\n",
    "\n",
    "leg = plt.legend(loc=\"upper left\")\n",
    "for lh in leg.legendHandles:\n",
    "    lh.set_alpha(1)\n",
    "plt.xlabel(r\"Index points ($\\mathbb{R}^1$)\")\n",
    "plt.ylabel(\"Observation space\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compare to HF only model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.\n"
     ]
    }
   ],
   "source": [
    "def build_gp(amplitude, length_scale, observation_noise_variance):\n",
    "    \"\"\"Defines the conditional dist. of GP outputs, given kernel parameters.\"\"\"\n",
    "\n",
    "    # Create the covariance kernel, which will be shared between the prior (which we\n",
    "    # use for maximum likelihood training) and the posterior (which we use for\n",
    "    # posterior predictive sampling)\n",
    "    kernel = tfk.ExponentiatedQuadratic(amplitude, length_scale)\n",
    "\n",
    "    # Create the GP prior distribution, which we will use to train the model\n",
    "    # parameters.\n",
    "    return tfd.GaussianProcess(\n",
    "        kernel=kernel,\n",
    "        index_points=x_train_h,\n",
    "        observation_noise_variance=observation_noise_variance,\n",
    "    )\n",
    "\n",
    "\n",
    "gp_joint_model = tfd.JointDistributionNamed(\n",
    "    {\n",
    "        \"amplitude\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"length_scale\": tfd.LogNormal(loc=0.0, scale=np.float64(1.0)),\n",
    "        \"observation_noise_variance\": tfd.LogNormal(loc=0.0, scale=np.float64(0.025)),\n",
    "        \"observations\": build_gp,\n",
    "    }\n",
    ")\n",
    "\n",
    "# Create the trainable model parameters, which we'll subsequently optimize.\n",
    "# Note that we constrain them to be strictly positive.\n",
    "\n",
    "constrain_positive = tfb.Shift(np.finfo(np.float64).tiny)(tfb.Exp())\n",
    "\n",
    "amplitude_var = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0, bijector=constrain_positive, name=\"amplitude\", dtype=np.float64\n",
    ")\n",
    "\n",
    "length_scale_var = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"length_scale\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "\n",
    "observation_noise_variance_var = tfp.util.TransformedVariable(\n",
    "    initial_value=1.0,\n",
    "    bijector=constrain_positive,\n",
    "    name=\"observation_noise_variance_var\",\n",
    "    dtype=np.float64,\n",
    ")\n",
    "\n",
    "trainable_variables_hf_only = [\n",
    "    v.trainable_variables[0]\n",
    "    for v in [amplitude_var, length_scale_var, observation_noise_variance_var]\n",
    "]\n",
    "\n",
    "\n",
    "def target_log_prob_hf_only(amplitude, length_scale, observation_noise_variance):\n",
    "    return gp_joint_model.log_prob(\n",
    "        {\n",
    "            \"amplitude\": amplitude,\n",
    "            \"length_scale\": length_scale,\n",
    "            \"observation_noise_variance\": observation_noise_variance,\n",
    "            \"observations\": y_train_h[:, 0],\n",
    "        }\n",
    "    )\n",
    "\n",
    "\n",
    "# Now we optimize the model parameters.\n",
    "num_iters = 1000\n",
    "optimizer = tf.optimizers.Adam(learning_rate=0.01)\n",
    "\n",
    "\n",
    "# Use `tf.function` to trace the loss for more efficient evaluation.\n",
    "@tf.function(autograph=False, jit_compile=False)\n",
    "def train_model():\n",
    "    with tf.GradientTape() as tape:\n",
    "        loss = -target_log_prob_hf_only(\n",
    "            amplitude_var, length_scale_var, observation_noise_variance_var\n",
    "        )\n",
    "    grads = tape.gradient(loss, trainable_variables_hf_only)\n",
    "    optimizer.apply_gradients(zip(grads, trainable_variables_hf_only))\n",
    "    return loss\n",
    "\n",
    "\n",
    "# Store the likelihood values during training, so we can plot the progress\n",
    "lls_ = np.zeros(num_iters, np.float64)\n",
    "for i in range(num_iters):\n",
    "    loss = train_model()\n",
    "    lls_[i] = loss\n",
    "\n",
    "\n",
    "# Having trained the model, we'd like to sample from the posterior conditioned\n",
    "# on observations. We'd like the samples to be at points other than the training\n",
    "# inputs.\n",
    "\n",
    "# HF only kernel\n",
    "optimized_kernel = tfk.ExponentiatedQuadratic(amplitude_var, length_scale_var)\n",
    "\n",
    "h_only_gprm = tfd.GaussianProcessRegressionModel(\n",
    "    kernel=optimized_kernel,\n",
    "    index_points=predictive_index_points_,\n",
    "    observation_index_points=x_train_h,\n",
    "    observations=y_train_h[:, 0],\n",
    "    observation_noise_variance=observation_noise_variance_var,\n",
    "    predictive_noise_variance=np.float64(0.0),\n",
    ")\n",
    "\n",
    "h_only_samples = h_only_gprm.sample(num_samples)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Mean absolute error, a metric used in https://adapt-python.github.io/adapt/examples/Multi_fidelity.html\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "MAE = np.mean(np.abs((y_pred + y_train_h_mean) - y_plot_h))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/sn/b7hd1bmd569cjwgmtww4mv440000gn/T/ipykernel_80537/2601353123.py:51: MatplotlibDeprecationWarning: The legendHandles attribute was deprecated in Matplotlib 3.7 and will be removed two minor releases later. Use legend_handles instead.\n",
      "  for lh in leg.legendHandles:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the true function, observations, and posterior samples.\n",
    "plt.figure(figsize=(12, 8))\n",
    "\n",
    "lower, median, upper = np.quantile(tmp_sample, [0.1, 0.5, 0.9], axis=0)\n",
    "\n",
    "# NARGP\n",
    "plt.plot(\n",
    "    predictive_index_points_,\n",
    "    y_pred + y_train_h_mean,\n",
    "    c=\"C2\",\n",
    "    label=\"NARGP: Posterior Mean\",\n",
    ")\n",
    "plt.fill_between(\n",
    "    predictive_index_points_[:, 0],\n",
    "    lower + y_train_h_mean,\n",
    "    upper + y_train_h_mean,\n",
    "    color=\"C2\",\n",
    "    alpha=0.1,\n",
    "    label=\"NARGP Posterior\",\n",
    ")\n",
    "\n",
    "\n",
    "lower_hf, median_hf, upper_hf = np.quantile(\n",
    "    h_only_samples.numpy(), [0.1, 0.5, 0.9], axis=0\n",
    ")\n",
    "\n",
    "# HF\n",
    "plt.plot(\n",
    "    predictive_index_points_,\n",
    "    median_hf + y_train_h_mean,\n",
    "    c=\"C3\",\n",
    "    label=\"HF Median Posterior\",\n",
    ")\n",
    "plt.fill_between(\n",
    "    predictive_index_points_[:, 0],\n",
    "    lower_hf + y_train_h_mean,\n",
    "    upper_hf + y_train_h_mean,\n",
    "    color=\"C3\",\n",
    "    alpha=0.1,\n",
    "    label=\"HF Posterior\",\n",
    ")\n",
    "\n",
    "plt.plot(x_plot, y_plot_h, label=\"True fn (HF)\", color=\"C1\")\n",
    "plt.plot(x_plot, y_plot_l, label=\"True fn (LF)\", color=\"C0\")\n",
    "\n",
    "plt.scatter(x_train_l, y_train_l_old, color=\"C0\", s=40, label=\"Observations (LF)\")\n",
    "plt.scatter(x_train_h, y_train_h_old, color=\"C1\", s=40, label=\"Observations (HF)\")\n",
    "\n",
    "\n",
    "leg = plt.legend(loc=\"upper left\")\n",
    "for lh in leg.legendHandles:\n",
    "    lh.set_alpha(1)\n",
    "\n",
    "plt.title(\n",
    "    \"(Multi-fideliy) Non-linear Autoregressive Gaussian Process : MAE = {:.3g}\".format(\n",
    "        MAE\n",
    "    )\n",
    ")\n",
    "plt.xlabel(r\"Index points ($\\mathbb{R}^1$)\")\n",
    "plt.ylabel(\"Observation space\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.16"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "d1b01947c0149e39ac179d3b99d6783022ec59b0ff1ef1456a9a1f74f9dd6eec"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}