docs/source/quickstart.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Quickstart\n",
    "\n",
    "VIABEL currently supports both standard KL-based variational inference (KLVI) and chi-squared variational inference (CHIVI). Models are provided as Autograd-compatible log densities or can be constructed from PyStan fit objects.\n",
    "\n",
    "As a simple example, we consider Neal's funnel distribution in 2 dimensions so that we can visualize the results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import autograd.numpy as np\n",
    "import autograd.scipy.stats.norm as norm\n",
    "\n",
    "D = 2  # number of dimensions\n",
    "log_sigma_stdev = 1. # 1.35\n",
    "def log_density(x):\n",
    "    mu, log_sigma = x[:, 0], x[:, 1]\n",
    "    sigma_density = norm.logpdf(log_sigma, 0, log_sigma_stdev)\n",
    "    mu_density = norm.logpdf(mu, 0, np.exp(log_sigma))\n",
    "    return sigma_density + mu_density"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Black-box Variational Inference\n",
    "\n",
    "VIABEL's `bbvi` function provides reasonable defaults: the objective is the ELBO (i.e., the including Kullback-Leibler divergence), a mean-field Gaussian approximation family, and automated RMSProp with adaptive step reduction and stopping rule. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using cached StanModel for weighted_lin_regression\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "average loss = 1.669 | R hat converged|:  60%|██████    | 18125/30000 [00:29<00:19, 623.53it/s]       \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Convergence reached at iteration 18125\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "average loss = 0.86764 | R hat converged|:   5%|▌         | 651/11874 [00:00<00:16, 660.99it/s]    "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Convergence reached at iteration 651\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "\n",
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.03 seconds (Sampling)\n",
      "               0.07 seconds (Total)\n",
      "\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.04 seconds (Sampling)\n",
      "               0.08 seconds (Total)\n",
      "\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.05 seconds (Sampling)\n",
      "               0.09 seconds (Total)\n",
      "\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.06 seconds (Sampling)\n",
      "               0.1 seconds (Total)\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:pystan:20 of 2000 iterations ended with a divergence (1 %).\n",
      "WARNING:pystan:Try running with adapt_delta larger than 0.98 to remove the divergences.\n",
      "average loss = 0.72357 | R hat converged|:   6%|▌         | 682/11222 [00:01<00:16, 642.38it/s]    "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Convergence reached at iteration 682\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "\n",
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.02 seconds (Sampling)\n",
      "               0.06 seconds (Total)\n",
      "\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.01 seconds (Sampling)\n",
      "               0.05 seconds (Total)\n",
      "\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.02 seconds (Sampling)\n",
      "               0.06 seconds (Total)\n",
      "\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.03 seconds (Sampling)\n",
      "               0.07 seconds (Total)\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "average loss = 0.61402 | R hat converged|:  23%|██▎       | 2443/10539 [00:03<00:12, 644.84it/s]   "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Convergence reached at iteration 2443\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "\n",
      "Gradient evaluation took 0 seconds\n",
      "1000 transitions using 10 leapfrog steps per transition would take 0 seconds.\n",
      "Adjust your expectations accordingly!\n",
      "\n",
      "\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration:   1 / 1000 [  0%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 100 / 1000 [ 10%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 200 / 1000 [ 20%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 300 / 1000 [ 30%]  (Warmup)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 400 / 1000 [ 40%]  (Warmup)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 500 / 1000 [ 50%]  (Warmup)\n",
      "Iteration: 501 / 1000 [ 50%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 600 / 1000 [ 60%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 700 / 1000 [ 70%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.03 seconds (Warm-up)\n",
      "               0.03 seconds (Sampling)\n",
      "               0.06 seconds (Total)\n",
      "\n",
      "Iteration: 800 / 1000 [ 80%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.02 seconds (Sampling)\n",
      "               0.06 seconds (Total)\n",
      "\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 900 / 1000 [ 90%]  (Sampling)\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.04 seconds (Warm-up)\n",
      "               0.02 seconds (Sampling)\n",
      "               0.06 seconds (Total)\n",
      "\n",
      "Iteration: 1000 / 1000 [100%]  (Sampling)\n",
      "\n",
      " Elapsed Time: 0.05 seconds (Warm-up)\n",
      "               0.02 seconds (Sampling)\n",
      "               0.07 seconds (Total)\n",
      "\n",
      "Termination rule reached at iteration 21901\n",
      "Inefficiency Index: 1.1550993138293368\n"
     ]
    }
   ],
   "source": [
    "from viabel import bbvi, MFStudentT\n",
    "results = bbvi(D, log_density=log_density, learning_rate=0.5, n_iters=30000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then plot contours the from the approximation Gaussian (red) and the target funnel distribution (black)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "sns.set_style('white')\n",
    "sns.set_context('notebook', font_scale=2, rc={'lines.linewidth': 2})\n",
    "\n",
    "def plot_approx_and_exact_contours(log_density, approx, var_param,\n",
    "                                   xlim, ylim, cmap2='Reds'):\n",
    "    xlist = np.linspace(*xlim, 100)\n",
    "    ylist = np.linspace(*ylim, 100)\n",
    "    X, Y = np.meshgrid(xlist, ylist)\n",
    "    XY = np.concatenate([np.atleast_2d(X.ravel()), np.atleast_2d(Y.ravel())]).T\n",
    "    zs = np.exp(log_density(XY))\n",
    "    Z = zs.reshape(X.shape)\n",
    "    zsapprox = np.exp(approx.log_density(var_param, XY))\n",
    "    Zapprox = zsapprox.reshape(X.shape)\n",
    "    plt.contour(X, Y, Z, cmap='Greys', linestyles='solid')\n",
    "    plt.contour(X, Y, Zapprox, cmap=cmap2, linestyles='solid')\n",
    "    plt.xlim(xlim)\n",
    "    plt.ylim(ylim)\n",
    "    plt.show()\n",
    "\n",
    "plot_approx_and_exact_contours(log_density, results['objective'].approx, results['opt_param'], \n",
    "                               xlim=[-2, 2], ylim=[-3.25, 2])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Diagnostics\n",
    "\n",
    "VIABEL also has a suite of diagostics for variational inference. We can easily run these using the `vi_diagnostics` function, although low-level support is also provided. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pareto k is estimated to be khat = 0.77\n",
      "WARNING: khat > 0.7 means importance sampling is not feasible.\n",
      "WARNING: not running further diagnostics\n"
     ]
    }
   ],
   "source": [
    "from viabel import vi_diagnostics\n",
    "import warnings\n",
    "with warnings.catch_warnings():\n",
    "    warnings.simplefilter('ignore')\n",
    "    diagnostics = vi_diagnostics(results['opt_param'], objective=results['objective'], n_samples=100000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importance Sampling\n",
    "\n",
    "The Pareto-smoothed weights provide a fairly accuracy approximation. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "weights = np.exp(diagnostics['smoothed_log_weights'])\n",
    "all_samples = diagnostics['samples']\n",
    "\n",
    "# sample 1000 weights for visualization\n",
    "subset = np.random.choice(all_samples.shape[1], size=1000, p=weights)\n",
    "samples = all_samples[:,subset]\n",
    "plt.plot(samples[0], samples[1], '*', alpha=.3)\n",
    "plot_approx_and_exact_contours(log_density, results['objective'].approx, results['opt_param'], \n",
    "                               xlim=[-2, 2], ylim=[-3.25, 2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (Spyder)",
   "language": "python3",
   "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": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}