NorESM_GP_model-pytorch-version.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "GP model for NOAA AI 2021 Hackathon 'Climate Bench'.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/iris/config.py:139: UserWarning: Ignoring config item 'Resources':'test_data_dir' (section:option) as '/home/h05/itwl/projects/git/iris-test-data/test_data' is not a valid directory path.\n",
      "  warnings.warn(msg.format(section, option, c_path))\n"
     ]
    }
   ],
   "source": [
    "# Environment setup\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import xarray as xr\n",
    "from eofs.xarray import Eof\n",
    "import torch\n",
    "import gpytorch\n",
    "import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Utilities for normalizing the emissions data\n",
    "min_co2 = 0.\n",
    "max_co2 = 2400\n",
    "def normalize_co2(data):\n",
    "    return data / max_co2\n",
    "\n",
    "def un_normalize_co2(data):\n",
    "    return data * max_co2\n",
    "\n",
    "min_ch4 = 0.\n",
    "max_ch4 = 0.6\n",
    "def normalize_ch4(data):\n",
    "    return data / max_ch4\n",
    "\n",
    "def un_normalize_ch4(data):\n",
    "    return data * max_ch4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot inputs\n",
    "data_path = \"./data/train_val/\"\n",
    "from glob import glob\n",
    "\n",
    "inputs = glob(data_path + \"inputs_s*.nc\")\n",
    "SECONDS_IN_YEAR = 60*60*24*365 #s\n",
    "\n",
    "fig, axes = plt.subplots(2, 2, figsize=(12,12))\n",
    "\n",
    "for input in inputs:\n",
    "    label=input.split('_')[1][:-3]\n",
    "    X = xr.open_dataset(input)\n",
    "    x = range(2015, 2101)\n",
    "\n",
    "    weights = np.cos(np.deg2rad(X.latitude))\n",
    "    \n",
    "    axes[0, 0].plot(x, X['CO2'].data, label=label)\n",
    "    axes[0, 0].set_ylabel(\"Cumulative anthropogenic CO2 \\nemissions since 1850 (GtCO2)\")\n",
    "    axes[0, 1].plot(x, X['CH4'].data, label=label)\n",
    "    axes[0, 1].set_ylabel(\"Anthropogenic CH4 \\nemissions (GtCH4 / year)\")\n",
    "    # FIXME: Not sure where this factor of 1000 comes from...! Maybe the CEDS data is really g/m-2/s?\n",
    "    axes[1, 0].plot(x, X['SO2'].weighted(weights).sum(['latitude', 'longitude']).data*SECONDS_IN_YEAR*1e-9, label=label)\n",
    "    axes[1, 0].set_ylabel(\"Anthropogenic SO2 \\nemissions (GtSO2 / year)\")\n",
    "    axes[1, 1].plot(x, X['BC'].weighted(weights).sum(['latitude', 'longitude']).data*SECONDS_IN_YEAR*1e-9, label=label)\n",
    "    axes[1, 1].set_ylabel(\"Anthropogenic BC \\nemissions (GtBC / year)\")\n",
    "\n",
    "axes[0, 0].set_title('CO2')\n",
    "axes[0, 1].set_title('CH4')\n",
    "axes[1, 0].set_title('SO2')\n",
    "axes[1, 1].set_title('BC')\n",
    "axes[0, 0].legend()\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-6-63d8f46b56cd>:110: FutureWarning: arrays to stack must be passed as a \"sequence\" type such as list or tuple. Support for non-sequence iterables such as generators is deprecated as of NumPy 1.16 and will raise an error in the future.\n",
      "  y_train_tas = np.vstack((cube_to_vector(create_predicatand_data(file)[\"tas\"]) for file in files))\n",
      "<ipython-input-6-63d8f46b56cd>:111: FutureWarning: arrays to stack must be passed as a \"sequence\" type such as list or tuple. Support for non-sequence iterables such as generators is deprecated as of NumPy 1.16 and will raise an error in the future.\n",
      "  y_train_pr = np.vstack((cube_to_vector(create_predicatand_data(file)[\"pr\"]) for file in files))\n",
      "<ipython-input-6-63d8f46b56cd>:112: FutureWarning: arrays to stack must be passed as a \"sequence\" type such as list or tuple. Support for non-sequence iterables such as generators is deprecated as of NumPy 1.16 and will raise an error in the future.\n",
      "  y_train_dtf = np.vstack((cube_to_vector(create_predicatand_data(file)[\"diurnal_temperature_range\"]) for file in files))\n",
      "<ipython-input-6-63d8f46b56cd>:113: FutureWarning: arrays to stack must be passed as a \"sequence\" type such as list or tuple. Support for non-sequence iterables such as generators is deprecated as of NumPy 1.16 and will raise an error in the future.\n",
      "  y_train_pr90 = np.vstack((cube_to_vector(create_predicatand_data(file)[\"pr90\"]) for file in files))\n"
     ]
    }
   ],
   "source": [
    "## Set up training data\n",
    "\n",
    "#adapted from Kai and Theorodore:\n",
    "def cube_to_vector(data_array):\n",
    "    \"transforms 3 dimensional xarray (time, lat, lon) to 2 dimensional numpy array (time, lat*lon)\"\n",
    "    time_range = data_array[\"time\"].shape[0]\n",
    "    \n",
    "    return data_array.values.reshape(time_range,96*144)\n",
    "\n",
    "\n",
    "def moving_average(a, n) :\n",
    "    ret = np.cumsum(a, dtype=float)\n",
    "    ret[n:] = ret[n:] - ret[:-n]\n",
    "    return ret[n - 1:] / n\n",
    "\n",
    "def create_predictor_data(data_set, n_eofs=5):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        data_set (str): name of dataset\n",
    "        n_eofs (int): number of eofs to create for aerosol variables\n",
    "    \n",
    "    \"\"\"\n",
    "    X = xr.open_dataset(data_path + \"inputs_{}.nc\".format(data_set)).compute()\n",
    "    \n",
    "    if data_set == \"hist-aer\":\n",
    "        X = X.rename_vars({\"CO4\":\"CO2\"})\n",
    "        X = X.sel(time=slice(1850,2100))\n",
    "\n",
    "    if data_set == \"hist-GHG\":\n",
    "        X = X.sel(time=slice(1850,2014))\n",
    "    \n",
    "    if \"ssp\" in data_set or data_set == \"hist-aer\":\n",
    "        # Create an EOF solver to do the EOF analysis. Square-root of cosine of\n",
    "        # latitude weights are applied before the computation of EOFs.\n",
    "        bc_solver = Eof(X['BC'])\n",
    "\n",
    "        # Retrieve the leading EOF, expressed as the correlation between the leading\n",
    "        # PC time series and the input SST anomalies at each grid point, and the\n",
    "        # leading PC time series itself.\n",
    "        bc_eofs = bc_solver.eofsAsCorrelation(neofs=n_eofs)\n",
    "        bc_pcs = bc_solver.pcs(npcs=n_eofs, pcscaling=1)\n",
    "\n",
    "        # Create an EOF solver to do the EOF analysis. Square-root of cosine of\n",
    "        # latitude weights are applied before the computation of EOFs.\n",
    "        so2_solver = Eof(X['SO2'])\n",
    "\n",
    "        # Retrieve the leading EOF, expressed as the correlation between the leading\n",
    "        # PC time series and the input SST anomalies at each grid point, and the\n",
    "        # leading PC time series itself.\n",
    "        so2_eofs = so2_solver.eofsAsCorrelation(neofs=n_eofs)\n",
    "        so2_pcs = so2_solver.pcs(npcs=n_eofs, pcscaling=1)\n",
    "\n",
    "        # Convert the Principle Components of the aerosol emissions (calculated above) in to Pandas DataFrames\n",
    "        bc_df = bc_pcs.to_dataframe().unstack('mode')\n",
    "        bc_df.columns = [f\"BC_{i}\" for i in range(n_eofs)]\n",
    "\n",
    "        so2_df = so2_pcs.to_dataframe().unstack('mode')\n",
    "        so2_df.columns = [f\"SO2_{i}\" for i in range(n_eofs)]\n",
    "    else:\n",
    "        # all values are zero, fill up eofs so we have same inputs as for other datasets\n",
    "        timesteps = X[\"time\"].shape[0]\n",
    "        zeros = np.zeros(shape=(timesteps, n_eofs))\n",
    "        bc_df = pd.DataFrame(zeros, columns=[f\"BC_{i}\" for i in range(n_eofs)],index=X[\"BC\"].coords['time'].data)\n",
    "        so2_df = pd.DataFrame(zeros, columns=[f\"SO2_{i}\" for i in range(n_eofs)],index=X[\"BC\"].coords['time'].data)\n",
    "\n",
    "    # Bring the emissions data back together again and normalise\n",
    "    inputs = pd.DataFrame({\n",
    "        \"CO2\": normalize_co2(X[\"CO2\"].data),\n",
    "        \"CH4\": normalize_ch4(X[\"CH4\"].data)\n",
    "    }, index=X[\"CO2\"].coords['time'].data)\n",
    "\n",
    "    #5-year running mean of CO2\n",
    "    co2_1 = normalize_co2(X[\"CO2\"].data)\n",
    "    co2_5yr=moving_average(co2_1,5)\n",
    "    nan_array = np.empty((1,4))\n",
    "    nan_array[:] = np.NaN\n",
    "    co2_5yr = np.append(nan_array, co2_5yr)\n",
    "    \n",
    "    co2_5 = pd.DataFrame({\"CO2_5YR\": co2_5yr}, index=X[\"CO2\"].coords['time'].data)\n",
    "    \n",
    "    # Combine with aerosol EOFs\n",
    "    inputs=pd.concat([inputs, bc_df, so2_df, co2_5], axis=1)\n",
    "    \n",
    "    # replace nan values with 0\n",
    "    inputs = inputs.replace(np.nan, 0)\n",
    "    \n",
    "    return inputs\n",
    "\n",
    "def create_predicatand_data(data_set):\n",
    "    Y = xr.open_dataset(data_path + \"outputs_{}.nc\".format(data_set)).mean(\"member\")\n",
    "    if data_set == \"hist-aer\":\n",
    "        Y = Y.sel(time=slice(1850,2100))\n",
    "    # Convert the precip values to mm/day\n",
    "    Y[\"pr\"] *= 86400\n",
    "    Y[\"pr90\"] *= 86400\n",
    "    return Y\n",
    "\n",
    "def create_predicatand_data(data_set):\n",
    "    Y = xr.open_dataset(data_path + \"outputs_{}.nc\".format(data_set)).mean(\"member\")\n",
    "    # Convert the precip values to mm/day\n",
    "    Y[\"pr\"] *= 86400\n",
    "    Y[\"pr90\"] *= 86400\n",
    "        \n",
    "    return Y\n",
    "\n",
    "files = [\"ssp126\",\"ssp585\",\"historical\",\"1pctCO2\",\"hist-GHG\",\"hist-aer\"]\n",
    "\n",
    "# create training data\n",
    "X_train = pd.concat([create_predictor_data(file) for file in files])\n",
    "y_train_tas = np.vstack((cube_to_vector(create_predicatand_data(file)[\"tas\"]) for file in files))\n",
    "y_train_pr = np.vstack((cube_to_vector(create_predicatand_data(file)[\"pr\"]) for file in files))\n",
    "y_train_dtf = np.vstack((cube_to_vector(create_predicatand_data(file)[\"diurnal_temperature_range\"]) for file in files))\n",
    "y_train_pr90 = np.vstack((cube_to_vector(create_predicatand_data(file)[\"pr90\"]) for file in files)) \n",
    "\n",
    "files2 = [\"ssp370\"]\n",
    "\n",
    "# create test data\n",
    "X_test = pd.concat([create_predictor_data(file) for file in files2])\n",
    "test_Y = xr.open_dataset(data_path + 'outputs_ssp370.nc').compute()\n",
    "\n",
    "tas_truth = test_Y[\"tas\"].mean('member')\n",
    "pr_truth = test_Y[\"pr\"].mean('member') * 86400\n",
    "pr90_truth = test_Y[\"pr90\"].mean('member') * 86400\n",
    "dtr_truth = test_Y[\"diurnal_temperature_range\"].mean('member')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make a kernel\n",
    "With gpytorch, we need to explicitely construct a kernel instance. This is the real strength of gpytorch, we can be much finer in our choices of kernels as we have much more flexibility."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ProductKernel(\n",
       "  (kernels): ModuleList(\n",
       "    (0): LinearKernel(\n",
       "      (raw_variance_constraint): Positive()\n",
       "    )\n",
       "    (1): RBFKernel(\n",
       "      (raw_lengthscale_constraint): Positive()\n",
       "    )\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Instantiate a linear kernel\n",
    "linear_kernel = gpytorch.kernels.LinearKernel()\n",
    "\n",
    "# Instantiate a RBF kernel\n",
    "rbf_kernel = gpytorch.kernels.RBFKernel()\n",
    "\n",
    "# To take the product kernel (like you did), we can then just write their product\n",
    "kernel = linear_kernel * rbf_kernel\n",
    "\n",
    "# This is what the resulting object looks like\n",
    "kernel"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make a GP\n",
    "We then need to explicitely construct a class that defines the GP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MyGP(gpytorch.models.ExactGP):\n",
    "    \"\"\"Our custom GP class\n",
    "    \n",
    "    Args:\n",
    "        X_train (torch.Tensor): that's your training inputs, but converted as pytorch tensors\n",
    "        y_train (torch.Tensor): training targets as pytorch tensors also\n",
    "        kernel (gpytorch.kernels.Kernel): an instance of gpytorch kernel like the one we have\n",
    "            defined above, but could be any other really\n",
    "        likelihood (gpytorch.likelihoods.Likelihood): this is a module that will allow to \n",
    "            define the observation process behavior, it is not something ESEm has I believe.\n",
    "            You can ignore it for now.\n",
    "    \"\"\"\n",
    "    def __init__(self, X_train, y_train, kernel, likelihood):\n",
    "        # Feed some arguments to the parent class\n",
    "        super().__init__(X_train, y_train, likelihood)\n",
    "        \n",
    "        # Define the mean of our GP, we use a zero mean for simplicity here\n",
    "        self.mean = gpytorch.means.ZeroMean()\n",
    "        \n",
    "        # Set our kernel as a class attribute\n",
    "        self.kernel = kernel\n",
    "\n",
    "    def forward(self, x):\n",
    "        \"\"\"This method defines how the GP prior is explicitely determined given input\n",
    "            sample, i.e. given `x`, what distribution follows `f(x)`\n",
    "        \n",
    "        Args:\n",
    "            x (torch.Tensor): input samples we want to evaluate the GP over\n",
    "        \"\"\"\n",
    "        # Compute the mean vector corresponding to the input samples - m(x)\n",
    "        mean_x = self.mean(x)\n",
    "        \n",
    "        # Compute the covariance matrix of the input samples - k(x, x)\n",
    "        covar_x = self.kernel(x)\n",
    "        \n",
    "        # Encapsulate as a multivariate normal distribution N(m(x), k(x, x))\n",
    "        prior_distribution = gpytorch.distributions.MultivariateNormal(mean_x, covar_x)\n",
    "        return prior_distribution"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can now instantiate an instance of our GP."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert training data as pytorch tensors\n",
    "X_train_as_tensor = torch.from_numpy(X_train.values).float()\n",
    "y_train_as_tensor = torch.from_numpy(y_train_tas).float()\n",
    "\n",
    "# Utility function to instantiate a GP\n",
    "def create_gp(X_train, y_train, kernel):\n",
    "    # Instantiate a likelihood instance - again, ignore this for now, I'll explain later why we need it\n",
    "    likelihood = gpytorch.likelihoods.GaussianLikelihood()\n",
    "\n",
    "    # Instantiate our GP model\n",
    "    model = MyGP(X_train=X_train,\n",
    "                 y_train=y_train,\n",
    "                 kernel=kernel,\n",
    "                 likelihood=likelihood)\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since GPs are scalar-valued and the output Y is multidimensional, we need to instantiate one GP for each output dimensions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the number of output dimensions \n",
    "ndim = y_train_as_tensor.size(-1)\n",
    "\n",
    "# Instantiate ndim GPs sharing the same kernel - each one predicts one dimension only\n",
    "models = [create_gp(X_train_as_tensor, y_train_as_tensor[:,i], kernel) for i in range(ndim)]\n",
    "\n",
    "# Encapsulate them as a list of independent models\n",
    "model = gpytorch.models.IndependentModelList(*models)\n",
    "\n",
    "# Encapsulate their likelihoods as a list of independent likelihoods\n",
    "likelihood = gpytorch.likelihoods.LikelihoodList(*[x.likelihood for x in models])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train the GP\n",
    "\n",
    "With pytorch, there is no `.train()` magic method that does all the work for you. Instead, you need to specify the loss and explicitely write gradient descent steps."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Say to the model and its likelihood that we'd like to be in training mode\n",
    "model.train()\n",
    "likelihood.train()\n",
    "\n",
    "# Define an optimization strategy over the model parameters and a learning rate (lr)\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=0.1)\n",
    "\n",
    "# \"Loss\" for GPs - the marginal log likelihood\n",
    "mll = gpytorch.mlls.SumMarginalLogLikelihood(likelihood, model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "42dfe30859404a458ff3d4b34a07a306",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Epochs'), FloatProgress(value=0.0, max=50.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "ERROR:root:Internal Python error in the inspect module.\n",
      "Below is the traceback from this internal error.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Traceback (most recent call last):\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/interactiveshell.py\", line 3418, in run_code\n",
      "    exec(code_obj, self.user_global_ns, self.user_ns)\n",
      "  File \"<ipython-input-23-5f72fa2edff2>\", line 14, in <module>\n",
      "    loss = -mll(output, model.train_targets)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\", line 30, in __call__\n",
      "    outputs = self.forward(*inputs, **kwargs)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/mlls/sum_marginal_log_likelihood.py\", line 33, in forward\n",
      "    sum_mll = sum(mll(output, target) for mll, output, target in zip(self.mlls, outputs, targets))\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/mlls/sum_marginal_log_likelihood.py\", line 33, in <genexpr>\n",
      "    sum_mll = sum(mll(output, target) for mll, output, target in zip(self.mlls, outputs, targets))\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\", line 30, in __call__\n",
      "    outputs = self.forward(*inputs, **kwargs)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/mlls/exact_marginal_log_likelihood.py\", line 62, in forward\n",
      "    res = output.log_prob(target)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/distributions/multivariate_normal.py\", line 168, in log_prob\n",
      "    covar = covar.evaluate_kernel()\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/added_diag_lazy_tensor.py\", line 187, in evaluate_kernel\n",
      "    added_diag_lazy_tsr = self.representation_tree()(*self.representation())\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_tensor.py\", line 1516, in representation_tree\n",
      "    return LazyTensorRepresentationTree(self)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_tensor_representation_tree.py\", line 13, in __init__\n",
      "    representation_size = len(arg.representation())\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_evaluated_kernel_tensor.py\", line 317, in representation\n",
      "    return self.evaluate_kernel().representation()\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/utils/memoize.py\", line 59, in g\n",
      "    return _add_to_cache(self, cache_name, method(self, *args, **kwargs), *args, kwargs_pkl=kwargs_pkl)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_evaluated_kernel_tensor.py\", line 284, in evaluate_kernel\n",
      "    res = self.kernel(x1, x2, diag=False, last_dim_is_batch=self.last_dim_is_batch, **self.params)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\", line 398, in __call__\n",
      "    res = lazify(super(Kernel, self).__call__(x1_, x2_, last_dim_is_batch=last_dim_is_batch, **params))\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\", line 30, in __call__\n",
      "    outputs = self.forward(*inputs, **kwargs)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\", line 518, in forward\n",
      "    next_term = kern(x1, x2, diag=diag, **params)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\", line 398, in __call__\n",
      "    res = lazify(super(Kernel, self).__call__(x1_, x2_, last_dim_is_batch=last_dim_is_batch, **params))\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\", line 30, in __call__\n",
      "    outputs = self.forward(*inputs, **kwargs)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/rbf_kernel.py\", line 86, in forward\n",
      "    return RBFCovariance.apply(\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/functions/rbf_covariance.py\", line 14, in forward\n",
      "    unitless_sq_dist = sq_dist_func(x1_, x2_)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/rbf_kernel.py\", line 90, in <lambda>\n",
      "    lambda x1, x2: self.covar_dist(\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\", line 327, in covar_dist\n",
      "    res = self.distance_module._sq_dist(x1, x2, postprocess, x1_eq_x2)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\", line 43, in _sq_dist\n",
      "    res = x1_.matmul(x2_.transpose(-2, -1))\n",
      "KeyboardInterrupt\n",
      "\n",
      "During handling of the above exception, another exception occurred:\n",
      "\n",
      "Traceback (most recent call last):\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/interactiveshell.py\", line 2045, in showtraceback\n",
      "    stb = value._render_traceback_()\n",
      "AttributeError: 'KeyboardInterrupt' object has no attribute '_render_traceback_'\n",
      "\n",
      "During handling of the above exception, another exception occurred:\n",
      "\n",
      "Traceback (most recent call last):\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\", line 1170, in get_records\n",
      "    return _fixed_getinnerframes(etb, number_of_lines_of_context, tb_offset)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\", line 316, in wrapped\n",
      "    return f(*args, **kwargs)\n",
      "  File \"/Users/dieze/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\", line 350, in _fixed_getinnerframes\n",
      "    records = fix_frame_records_filenames(inspect.getinnerframes(etb, context))\n",
      "  File \"/Users/dieze/.pyenv/versions/3.8.0/lib/python3.8/inspect.py\", line 1503, in getinnerframes\n",
      "    frameinfo = (tb.tb_frame,) + getframeinfo(tb, context)\n",
      "  File \"/Users/dieze/.pyenv/versions/3.8.0/lib/python3.8/inspect.py\", line 1461, in getframeinfo\n",
      "    filename = getsourcefile(frame) or getfile(frame)\n",
      "  File \"/Users/dieze/.pyenv/versions/3.8.0/lib/python3.8/inspect.py\", line 708, in getsourcefile\n",
      "    if getattr(getmodule(object, filename), '__loader__', None) is not None:\n",
      "  File \"/Users/dieze/.pyenv/versions/3.8.0/lib/python3.8/inspect.py\", line 747, in getmodule\n",
      "    if f == _filesbymodname.get(modname, None):\n",
      "KeyboardInterrupt\n"
     ]
    },
    {
     "ename": "TypeError",
     "evalue": "object of type 'NoneType' has no len()",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "    \u001b[0;31m[... skipping hidden 1 frame]\u001b[0m\n",
      "\u001b[0;32m<ipython-input-23-5f72fa2edff2>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     13\u001b[0m     \u001b[0;31m# Calculate loss\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 14\u001b[0;31m     \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m-\u001b[0m\u001b[0mmll\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_targets\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     15\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"B\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *inputs, **kwargs)\u001b[0m\n\u001b[1;32m     29\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m         \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     31\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/mlls/sum_marginal_log_likelihood.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, outputs, targets, *params)\u001b[0m\n\u001b[1;32m     32\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 33\u001b[0;31m             \u001b[0msum_mll\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmll\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mmll\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmlls\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moutputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtargets\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     34\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/mlls/sum_marginal_log_likelihood.py\u001b[0m in \u001b[0;36m<genexpr>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m     32\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 33\u001b[0;31m             \u001b[0msum_mll\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmll\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mmll\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmlls\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moutputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtargets\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     34\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *inputs, **kwargs)\u001b[0m\n\u001b[1;32m     29\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m         \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     31\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/mlls/exact_marginal_log_likelihood.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, function_dist, target, *params)\u001b[0m\n\u001b[1;32m     61\u001b[0m         \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlikelihood\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfunction_dist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 62\u001b[0;31m         \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0moutput\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlog_prob\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtarget\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     63\u001b[0m         \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_add_other_terms\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/distributions/multivariate_normal.py\u001b[0m in \u001b[0;36mlog_prob\u001b[0;34m(self, value)\u001b[0m\n\u001b[1;32m    167\u001b[0m         \u001b[0;31m# Get log determininant and first part of quadratic form\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 168\u001b[0;31m         \u001b[0mcovar\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcovar\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mevaluate_kernel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    169\u001b[0m         \u001b[0minv_quad\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlogdet\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcovar\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minv_quad_logdet\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minv_quad_rhs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdiff\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munsqueeze\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlogdet\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/added_diag_lazy_tensor.py\u001b[0m in \u001b[0;36mevaluate_kernel\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    186\u001b[0m         \"\"\"\n\u001b[0;32m--> 187\u001b[0;31m         \u001b[0madded_diag_lazy_tsr\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrepresentation_tree\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrepresentation\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    188\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0madded_diag_lazy_tsr\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_lazy_tensor\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0madded_diag_lazy_tsr\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_diag_tensor\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_tensor.py\u001b[0m in \u001b[0;36mrepresentation_tree\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m   1515\u001b[0m         \"\"\"\n\u001b[0;32m-> 1516\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mLazyTensorRepresentationTree\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1517\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_tensor_representation_tree.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, lazy_tsr)\u001b[0m\n\u001b[1;32m     12\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marg\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"representation\"\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mcallable\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrepresentation\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;31m# Is it a lazy tensor?\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 13\u001b[0;31m                 \u001b[0mrepresentation_size\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrepresentation\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     14\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mchildren\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mslice\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcounter\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcounter\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mrepresentation_size\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0marg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrepresentation_tree\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_evaluated_kernel_tensor.py\u001b[0m in \u001b[0;36mrepresentation\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    316\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 317\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mevaluate_kernel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrepresentation\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    318\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/utils/memoize.py\u001b[0m in \u001b[0;36mg\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m     58\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0m_is_in_cache\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcache_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs_pkl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mkwargs_pkl\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 59\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0m_add_to_cache\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcache_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmethod\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs_pkl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mkwargs_pkl\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     60\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0m_get_from_cache\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcache_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs_pkl\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mkwargs_pkl\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/lazy/lazy_evaluated_kernel_tensor.py\u001b[0m in \u001b[0;36mevaluate_kernel\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    283\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkernel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mactive_dims\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 284\u001b[0;31m             \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkernel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdiag\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlast_dim_is_batch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlast_dim_is_batch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    285\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkernel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mactive_dims\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtemp_active_dims\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, x1, x2, diag, last_dim_is_batch, **params)\u001b[0m\n\u001b[1;32m    397\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 398\u001b[0;31m                 \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlazify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mKernel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx1_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlast_dim_is_batch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mlast_dim_is_batch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    399\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *inputs, **kwargs)\u001b[0m\n\u001b[1;32m     29\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m         \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     31\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x1, x2, diag, **params)\u001b[0m\n\u001b[1;32m    517\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mkern\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkernels\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 518\u001b[0;31m             \u001b[0mnext_term\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mkern\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdiag\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdiag\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    519\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mx1_eq_x2\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, x1, x2, diag, last_dim_is_batch, **params)\u001b[0m\n\u001b[1;32m    397\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 398\u001b[0;31m                 \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlazify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mKernel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx1_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlast_dim_is_batch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mlast_dim_is_batch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    399\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *inputs, **kwargs)\u001b[0m\n\u001b[1;32m     29\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m         \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     31\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/rbf_kernel.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x1, x2, diag, **params)\u001b[0m\n\u001b[1;32m     85\u001b[0m             )\n\u001b[0;32m---> 86\u001b[0;31m         return RBFCovariance.apply(\n\u001b[0m\u001b[1;32m     87\u001b[0m             \u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/functions/rbf_covariance.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(ctx, x1, x2, lengthscale, sq_dist_func)\u001b[0m\n\u001b[1;32m     13\u001b[0m         \u001b[0mx2_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx2\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdiv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlengthscale\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 14\u001b[0;31m         \u001b[0munitless_sq_dist\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msq_dist_func\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx1_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2_\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     15\u001b[0m         \u001b[0;31m# clone because inplace operations will mess with what's saved for backward\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/rbf_kernel.py\u001b[0m in \u001b[0;36m<lambda>\u001b[0;34m(x1, x2)\u001b[0m\n\u001b[1;32m     89\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlengthscale\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 90\u001b[0;31m             lambda x1, x2: self.covar_dist(\n\u001b[0m\u001b[1;32m     91\u001b[0m                 \u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msquare_dist\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdiag\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdist_postprocess_func\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mpostprocess_rbf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpostprocess\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mparams\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\u001b[0m in \u001b[0;36mcovar_dist\u001b[0;34m(self, x1, x2, diag, last_dim_is_batch, square_dist, dist_postprocess_func, postprocess, **params)\u001b[0m\n\u001b[1;32m    326\u001b[0m         \u001b[0;32melif\u001b[0m \u001b[0msquare_dist\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 327\u001b[0;31m             \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdistance_module\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sq_dist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpostprocess\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx1_eq_x2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    328\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/gpytorch/kernels/kernel.py\u001b[0m in \u001b[0;36m_sq_dist\u001b[0;34m(self, x1, x2, postprocess, x1_eq_x2)\u001b[0m\n\u001b[1;32m     42\u001b[0m         \u001b[0mx2_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mx2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2_pad\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2_norm\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 43\u001b[0;31m         \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx1_\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx2_\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtranspose\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     44\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: ",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/interactiveshell.py\u001b[0m in \u001b[0;36mshowtraceback\u001b[0;34m(self, exc_tuple, filename, tb_offset, exception_only, running_compiled_code)\u001b[0m\n\u001b[1;32m   2044\u001b[0m                         \u001b[0;31m# in the engines. This should return a list of strings.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2045\u001b[0;31m                         \u001b[0mstb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mvalue\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_render_traceback_\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2046\u001b[0m                     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mAttributeError\u001b[0m: 'KeyboardInterrupt' object has no attribute '_render_traceback_'",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "    \u001b[0;31m[... skipping hidden 1 frame]\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/interactiveshell.py\u001b[0m in \u001b[0;36mshowtraceback\u001b[0;34m(self, exc_tuple, filename, tb_offset, exception_only, running_compiled_code)\u001b[0m\n\u001b[1;32m   2045\u001b[0m                         \u001b[0mstb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mvalue\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_render_traceback_\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2046\u001b[0m                     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2047\u001b[0;31m                         stb = self.InteractiveTB.structured_traceback(etype,\n\u001b[0m\u001b[1;32m   2048\u001b[0m                                             value, tb, tb_offset=tb_offset)\n\u001b[1;32m   2049\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\u001b[0m in \u001b[0;36mstructured_traceback\u001b[0;34m(self, etype, value, tb, tb_offset, number_of_lines_of_context)\u001b[0m\n\u001b[1;32m   1434\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1435\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtb\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1436\u001b[0;31m         return FormattedTB.structured_traceback(\n\u001b[0m\u001b[1;32m   1437\u001b[0m             self, etype, value, tb, tb_offset, number_of_lines_of_context)\n\u001b[1;32m   1438\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\u001b[0m in \u001b[0;36mstructured_traceback\u001b[0;34m(self, etype, value, tb, tb_offset, number_of_lines_of_context)\u001b[0m\n\u001b[1;32m   1334\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mmode\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mverbose_modes\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1335\u001b[0m             \u001b[0;31m# Verbose modes need a full traceback\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1336\u001b[0;31m             return VerboseTB.structured_traceback(\n\u001b[0m\u001b[1;32m   1337\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0metype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtb_offset\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnumber_of_lines_of_context\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1338\u001b[0m             )\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\u001b[0m in \u001b[0;36mstructured_traceback\u001b[0;34m(self, etype, evalue, etb, tb_offset, number_of_lines_of_context)\u001b[0m\n\u001b[1;32m   1191\u001b[0m         \u001b[0;34m\"\"\"Return a nice text document describing the traceback.\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1192\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1193\u001b[0;31m         formatted_exception = self.format_exception_as_a_whole(etype, evalue, etb, number_of_lines_of_context,\n\u001b[0m\u001b[1;32m   1194\u001b[0m                                                                tb_offset)\n\u001b[1;32m   1195\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\u001b[0m in \u001b[0;36mformat_exception_as_a_whole\u001b[0;34m(self, etype, evalue, etb, number_of_lines_of_context, tb_offset)\u001b[0m\n\u001b[1;32m   1149\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1150\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1151\u001b[0;31m         \u001b[0mlast_unique\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrecursion_repeat\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfind_recursion\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0morig_etype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mevalue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrecords\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1152\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1153\u001b[0m         \u001b[0mframes\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat_records\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrecords\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlast_unique\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrecursion_repeat\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/torch-3.8/lib/python3.8/site-packages/IPython/core/ultratb.py\u001b[0m in \u001b[0;36mfind_recursion\u001b[0;34m(etype, value, records)\u001b[0m\n\u001b[1;32m    449\u001b[0m     \u001b[0;31m# first frame (from in to out) that looks different.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    450\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mis_recursion_error\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0metype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrecords\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 451\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrecords\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    452\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    453\u001b[0m     \u001b[0;31m# Select filename, lineno, func_name to track frames with\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mTypeError\u001b[0m: object of type 'NoneType' has no len()"
     ]
    }
   ],
   "source": [
    "# Define number of gradient iterations and a visual iterator to track progress\n",
    "n_iterations = 50\n",
    "training_iter = tqdm.notebook.tqdm(range(n_iterations), desc=f\"Epochs\")\n",
    "\n",
    "for i in training_iter:\n",
    "    # Zero gradients from previous iteration\n",
    "    optimizer.zero_grad()\n",
    "\n",
    "    # Output from model over training data\n",
    "    output = model(*model.train_inputs)\n",
    "    print(\"A\")\n",
    "\n",
    "    # Calculate loss\n",
    "    loss = -mll(output, model.train_targets)\n",
    "    print(\"B\")\n",
    "\n",
    "    # Compute gradients accross parameters\n",
    "    loss.backward()\n",
    "    print(\"C\")\n",
    "\n",
    "    # Take actual gradient step for all model parameters (essentially kernel parameters)\n",
    "    optimizer.step()\n",
    "    print(\"D\")\n",
    "\n",
    "    # Print out current loss value in visual iterator\n",
    "    training_iter.set_postfix(loss=loss.item())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set into eval mode\n",
    "model.eval()\n",
    "likelihood.eval()\n",
    "\n",
    "# Convert testing inputs as pytorch tensors\n",
    "X_test_as_tensor = torch.from_numpy(X_test.values).float()\n",
    "\n",
    "# Make predictions (use the same test points)\n",
    "with torch.no_grad():\n",
    "    # This contains predictions for all models as a list\n",
    "    predictions = likelihood(*model(*[X_test_as_tensor] * ndim))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "torch-kernel",
   "language": "python",
   "name": "torch-3.8-kernel"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}