Formal analysis and implementation sanity check of noisy minibatch estimates for expectations under correlated distributions

estimating_expectations_under_correlated_distributions.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "7aabff81",
   "metadata": {},
   "source": [
    "# Estimating Expectations of IID Likelihoods under correlated distributions over likelihood parameters\n",
    "\n",
    "This is a tutorial notebook about estimating certain kinds of expectations.\n",
    "\n",
    "This notebook has two parts:\n",
    "\n",
    "* Part 1 introduces the problem and discusses theory about estimators\n",
    "* Part 2 is a practical implementation of the estimators\n",
    "\n",
    "# Part 1: Theory\n",
    "\n",
    "## Random Variables\n",
    "\n",
    "We have two $D$-dimensional r.v.s:\n",
    "\n",
    "* $\\mathbf{f} = [f_1, f_2, \\ldots f_D]$, where each $f_d \\in \\mathbb{R}$ is a scalar real\n",
    "* $\\mathbf{y} = [y_1, y_2, \\ldots y_D]$, where each $y_d \\in \\mathbb{R}$ is a scalar real\n",
    "\n",
    "## Model\n",
    "\n",
    "We'll consider the following model\n",
    "\n",
    "\\begin{align}\n",
    "    p(\\mathbf{f}) &= \\mathcal{N}( \\mathbf{0}_D, \\mathbf{\\Sigma} ),  \\qquad && \\text{where}~\n",
    "   \\mathbf{\\Sigma} ~\\text{is a DxD covariance matrix}\n",
    "    \\\\\n",
    "    p(\\mathbf{y}|\\mathbf{f}) &= \\mathcal{N}( \\mathbf{f}, \\tau^2 \\mathbf{I}_D ), \\qquad && \\text{where}~ \\tau>0 ~\\text{is a scalar standard deviation}\n",
    "\\end{align}\n",
    "\n",
    "We'll call this a correlated/iid model, because the f variable has a correlated distribution, but the y-given-f is iid.\n",
    "\n",
    "\n",
    "## Goal: Compute expectation of iid likelihood under correlated dist\n",
    "\n",
    "We wish to evaluate the following expectation:\n",
    "\n",
    "\\begin{align}\n",
    "    \\mathcal{J}(\\mathbf{y}, \\Sigma) \n",
    "    &= \\mathbb{E}_{\\mathbf{f} \\sim \\mathcal{N}(\n",
    "        \\mathbf{0}_D, \\mathbf{\\Sigma})} \n",
    "    \\left[ \n",
    "        \\log p(\\mathbf{y} | \\mathbf{f} )\n",
    "    \\right]\n",
    "\\end{align}\n",
    "\n",
    "Applying some elementary transformations, we can manipulate the expression as follows\n",
    "\\begin{align}\n",
    "    \\mathcal{J}(\\mathbf{y}, \\Sigma) \n",
    "    &= \\mathbb{E}_{\\mathbf{f} \\sim \\mathcal{N}(\n",
    "        \\mathbf{0}_D, \\mathbf{\\Sigma})} \n",
    "    \\left[ \n",
    "        \\log p(\\mathbf{y} | \\mathbf{f} )\n",
    "    \\right]\n",
    "    \\\\\n",
    "    &=  \\mathbb{E}_{\\mathbf{f} \\sim \\mathcal{N}(\n",
    "        \\mathbf{0}_D, \\mathbf{\\Sigma})} \n",
    "    \\left[ \n",
    "        \\sum_{d=1}^D \\log \\mathcal{N}( y_d | f_d, \\tau^2)\n",
    "    \\right]\n",
    "    && \\text{by defn of y-given-f likelihood}\n",
    "    \\\\\n",
    "    &= \\sum_{d=1}^D \n",
    "    \\mathbb{E}_{\\mathbf{f} \\sim \\mathcal{N}(\n",
    "        \\mathbf{0}_D, \\mathbf{\\Sigma})} \n",
    "    \\left[ \n",
    "        \\log \\mathcal{N}( y_d | f_d, \\tau^2)\n",
    "    \\right]\n",
    "    && \\text{integral of sum = sum of integrals}\n",
    "    \\\\\n",
    "        &= \\sum_{d=1}^D \n",
    "    \\mathbb{E}_{f_d \\sim \\mathcal{N}(0, \\Sigma_{dd})}\n",
    "    \\left[ \n",
    "        \\log \\mathcal{N}( y_d | f_d, \\tau^2)\n",
    "    \\right]\n",
    "    && \\text{because of Property A below}\n",
    "\\end{align}\n",
    "\n",
    "**Proposition A**: The expectation of any function of one entry of a vector is equal to the expectation under the marginal of that one entry\n",
    "\n",
    "Let $f_d$ denote a specific scalar entry of $\\mathbf{f}$, and let $\\mathbf{f}_{-d}$ denote all the other entries in vector $\\mathbf{f}$ except the one at index $d$. \n",
    "\n",
    "\\begin{align}\n",
    "    \\mathbb{E}_{p(\\mathbf{f})}[ g(f_d) ] \n",
    "    &= \n",
    "        \\int \\int  g(f_d) \\underbrace{p(f_d) p(\\mathbf{f}_{-d} | f_d)}_{\\text{equals}~ p(\\mathbf{f})~ \\text{by product rule}} \\,df_d \\,d\\mathbf{f}_{-d}\n",
    "    \\\\\n",
    "    &= \\int g(f_d) p(f_d) \\underbrace{\\int  p(f_{-d} | f_d) \\,df_{-d}}_{\\text{equals one by defn of pdf over}~\\mathbf{f}_{-d}} \\,df_d   \n",
    "    \\\\\n",
    "    &= \\mathbb{E}_{ p(f_d) }[ g(f_d) ]\n",
    "\\end{align}\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5ceb7d9",
   "metadata": {},
   "source": [
    "\n",
    "## A noisy estimator by subsampling dimensions\n",
    "\n",
    "Suppose $D$ is large... we may want to avoid computing every term in the sum above and instead try to make an unbiased estimator by sampling a subset of dimensions.\n",
    "\n",
    "Let integer-valued random variable $u \\in \\{1, 2, \\ldots D\\}$ denote a specific dimension selected.\n",
    "\n",
    "Given a sample of $u$, define an estimator for our expectation as\n",
    "\n",
    "\\begin{align}\n",
    "\\hat{J}( u, \\mathbf{y}, \\mathbf{\\Sigma}) = D \\mathbb{E}_{f_u \\sim \\mathcal{N}(0, \\Sigma_{uu})}\n",
    "    \\left[ \n",
    "        \\log \\mathcal{N}( y_u | f_u, \\tau^2)\n",
    "    \\right]\n",
    "\\end{align}\n",
    "\n",
    "We'd like to show that this noisy estimator is UNBIASED. That is, its expected value is exactly equal to the true objective $J$.\n",
    "\n",
    "\\begin{align}\n",
    "\\text{WANT TO SHOW:}~~ \\mathbb{E}_{u \\sim \\text{Unif}(\\{1, \\ldots D\\})}[ \\hat{J}( u, \\mathbf{y}, \\mathbf{\\Sigma}) ] = J( \\mathbf{y}, \\Sigma)\n",
    "\\end{align}\n",
    "\n",
    "Let's proceed by manipulating the expectation over $u$ via standard rules \n",
    "\n",
    "\\begin{align}\n",
    "\\mathbb{E}_{u \\sim \\text{Unif}(\\{1, \\ldots D\\})}[ \\hat{J}( u, \\mathbf{y}, \\mathbf{\\Sigma}) ] \n",
    "    &= \\sum_{d=1}^D p(u=d) \\hat{J}(d, \\mathbf{y}, \\mathbf{\\Sigma}), \\qquad &&\\text{by defn of expectation over r.v. u}\n",
    "    \\\\\n",
    "    &= \\sum_{d=1}^D \\frac{1}{D} \\cdot \\left( D \\mathbb{E}_{f_d \\sim \\mathcal{N}(0, \\Sigma_{dd})}\n",
    "    \\left[ \n",
    "        \\log \\mathcal{N}( y_d | f_d, \\tau^2)\n",
    "    \\right] \\right), && \\text{by defn of u's Uniform PMF}\n",
    "    \\\\\n",
    "    &= \\sum_{d=1}^D \\mathbb{E}_{f_d \\sim \\mathcal{N}(0, \\Sigma_{dd})}\n",
    "    \\left[ \n",
    "        \\log \\mathcal{N}( y_d | f_d, \\tau^2)\n",
    "    \\right]\n",
    "    \\\\\n",
    "    &= J( \\mathbf{y}, \\mathbf{\\Sigma} ) && \\text{by defn of $J$}\n",
    "\\end{align}\n",
    "\n",
    "We have thus proven that we can subsample a specific dimension, compute a noisy estimator of our intended expectation by summing over only the selected dimension, and rescaling by a factor of $D$.\n",
    "\n",
    "\n",
    "## Extension to minibatch sampling\n",
    "\n",
    "Similar logic holds for sampling a subset of $B$ dimensions uniformly at random with replacement, where $B < D$.\n",
    "\n",
    "Define the sampling procedure as:\n",
    "\n",
    "\\begin{align}\n",
    "    u_1 &\\sim \\text{Unif}( \\{1, \\ldots D\\} )\n",
    "    \\\\\n",
    "    u_2 &\\sim \\text{Unif}( \\{1, \\ldots D\\} )\n",
    "    \\\\\n",
    "    \\vdots\n",
    "    \\\\\n",
    "    u_B &\\sim \\text{Unif}( \\{1, \\ldots D\\} )\n",
    "\\end{align}\n",
    "\n",
    "The minibatch estimator of $J$ is thus:\n",
    "\n",
    "\\begin{align}\n",
    "\\hat{J}( \\mathbf{u}_{1:B}, \\mathbf{y}, \\mathbf{\\Sigma})\n",
    "    = \\frac{D}{B} \\sum_{b=1}^B \\mathbb{E}_{f_{u_b} \\sim \\mathcal{N}(0, \\Sigma_{u_b,u_b})}\n",
    "    \\left[ \n",
    "        \\log \\mathcal{N}( y_{u_b} | f_{u_b}, \\tau^2)\n",
    "    \\right]\n",
    "\\end{align}\n",
    "\n",
    "It is straightforward (following the same logic from the single-dimension noisy estimator case) to show that this estimator is also UNBIASED."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b20f554d",
   "metadata": {},
   "source": [
    "## Why does this matter?\n",
    "\n",
    "We have shown that we can avoid expensive computation of the objective term $J$ by instead doing noisy estimates with single-entry or minibatch sampling.\n",
    "\n",
    "We've shown that the estimator is formally UNBIASED, which means its mean is \"in the right place\"\n",
    "\n",
    "What we haven't shown is how noisy the estimator is. \n",
    "We should be concerned about using minibatches in practice... the noise may lead to worse estimates."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5c7525c",
   "metadata": {},
   "source": [
    "# Part 2: Implementation\n",
    "\n",
    "We will compare 3 estimators for the objective $J$:\n",
    "\n",
    "* A Monte-Carlo estimator of the exact function $J$ using joint samples of $\\mathbb{f}$\n",
    "* A Monte-Carlo estimator of the exact function $J$ using marginal samples of each entry $f_d$\n",
    "* A Monte-Carlo estimator of the approximation $\\hat{J}$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "190131f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "afd71095",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mhughes/miniconda3/envs/spr_2021s_env/lib/python3.9/site-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.16.5 and <1.23.0 is required for this version of SciPy (detected version 1.23.1\n",
      "  warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n"
     ]
    }
   ],
   "source": [
    "import scipy.stats;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6f64fe1d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import seaborn as sns\n",
    "sns.set_style('whitegrid');\n",
    "sns.set(font_scale=1.33)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d96c24cf",
   "metadata": {},
   "source": [
    "### Hyperparameters\n",
    "\n",
    "We'll set D to be 3 (so we have nontrivial number of dims)\n",
    "\n",
    "We'll create a strongly correlated cov matrix Sigma\n",
    "\n",
    "We'll set the y-given-f std dev tau to be rather small"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a7f8f905",
   "metadata": {},
   "outputs": [],
   "source": [
    "D = 3\n",
    "Sigma_DD = .01 * np.eye(D) + 0.99\n",
    "tau = 0.1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef80f6cd",
   "metadata": {},
   "source": [
    "### Utility func to draw samples of f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "94abbbfb",
   "metadata": {},
   "outputs": [],
   "source": [
    "def draw_samples_of_f(n_samples, prng, Sigma_DD=Sigma_DD):\n",
    "    f_ND = prng.multivariate_normal(\n",
    "        np.zeros(D),\n",
    "        Sigma_DD, size=n_samples)\n",
    "    return f_ND\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d49b46f6",
   "metadata": {},
   "source": [
    "### Utility func to draw samples of y and f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "a7cdea31",
   "metadata": {},
   "outputs": [],
   "source": [
    "def draw_samples_of_y_and_f(n_samples, prng):\n",
    "    f_ND = draw_samples_of_f(n_samples, prng)\n",
    "    y_ND = tau * prng.randn(n_samples, D) + f_ND\n",
    "    return y_ND, f_ND"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ebcb2c9",
   "metadata": {},
   "source": [
    "### Show samples y "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "480870b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "prng = np.random.RandomState(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "cce72054",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_ND, f_ND = draw_samples_of_y_and_f(1000, prng)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "baa6bea0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 540x540 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.pairplot( pd.DataFrame(y_ND, columns=['y1','y2','y3']));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "88b2eb14",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-1.53443664, -1.78328796, -1.75201151],\n",
       "       [-2.10617572, -2.28553213, -2.44382124],\n",
       "       [-0.95319295, -0.92770474, -1.09444496],\n",
       "       [-0.31384097, -0.50057997, -0.39028486],\n",
       "       [-0.73326255, -0.94369179, -0.89444694],\n",
       "       [-0.21679044, -0.39113088, -0.44421848],\n",
       "       [-0.38213229, -0.1166083 , -0.65183733],\n",
       "       [-0.62962879, -0.64401357, -0.76730551],\n",
       "       [-2.27729516, -2.41547372, -2.13724376],\n",
       "       [ 0.50866712, -0.00444496,  0.19712329]])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's pick 10 values of our y-vector to use in experiments\n",
    "\n",
    "M = 10\n",
    "y_MD = y_ND[:M]\n",
    "\n",
    "y_MD"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "64ae9b20",
   "metadata": {},
   "source": [
    "## Option A: MC estimation of J using JOINT samples\n",
    "\n",
    "$$\n",
    "J(\\mathbf{y}, \\mathbf{\\Sigma}) = \\mathbb{E}_{\\mathcal{N}(\\mathbf{f}|\\mathbf{0}_D, \\mathbf{\\Sigma})}[ \\log p(\\mathbf{y}|\\mathbf{f}) ]\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4ef02b48",
   "metadata": {},
   "outputs": [],
   "source": [
    "def estimate_exact_J_via_joint_montecarlo(y_ND, n_samples, prng):\n",
    "    ''' Estimate objective J via MonteCarlo samples of f and exact J definition\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "    J_N : 1D array, size (N,)\n",
    "        Each entry equals the value of J for the corresponding row of input y_ND\n",
    "    '''\n",
    "    f_SD = draw_samples_of_f(n_samples, prng)\n",
    "    f_1DS = f_SD.T[np.newaxis,:]\n",
    "    \n",
    "    logpdf_NDS = scipy.stats.norm.logpdf(\n",
    "        y_ND[:,:,np.newaxis], f_1DS, tau)\n",
    "    J_N = logpdf_NDS.sum(axis=1).mean(axis=1)\n",
    "    return J_N"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "228d1eae",
   "metadata": {},
   "source": [
    "### Using 50 monte-carlo samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "78252b3f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "J(y_1), ... J(y_10) using rand state 101\n",
      "[-570.09192727 -904.81453669 -308.28804268 -201.53334864 -274.91107254\n",
      " -197.65112855 -206.47130172 -238.69900473 -903.13185806 -208.78500001]\n",
      "J(y_1), ... J(y_10) using rand state 202\n",
      "[-531.53133547 -884.26090589 -248.74716317 -124.11472458 -211.37229443\n",
      " -119.33084784 -129.43478119 -169.86945149 -880.61697067 -113.31565169]\n"
     ]
    }
   ],
   "source": [
    "for rand_state in [101, 202]:\n",
    "    prng = np.random.RandomState(rand_state)\n",
    "    print(\"J(y_1), ... J(y_10) using rand state %d\" % rand_state)\n",
    "    print(estimate_exact_J_via_joint_montecarlo(y_MD, 50, prng))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6e70188",
   "metadata": {},
   "source": [
    "### Using 500000 monte-carlo samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ec1bf428",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "J(y_1), ... J(y_10) using rand state 101\n",
      "[-576.95927708 -928.56122322 -294.84404423 -171.34958811 -257.86005484\n",
      " -166.12814371 -175.49541809 -216.37701107 -926.38008539 -160.9354694 ]\n",
      "J(y_1), ... J(y_10) using rand state 202\n",
      "[-576.01864905 -927.42930734 -294.13274702 -170.82833211 -257.1901543\n",
      " -165.62201148 -174.98204861 -215.76632902 -925.25374827 -160.61498885]\n"
     ]
    }
   ],
   "source": [
    "for rand_state in [101, 202]:\n",
    "    prng = np.random.RandomState(rand_state)\n",
    "    print(\"J(y_1), ... J(y_10) using rand state %d\" % rand_state)\n",
    "    print(estimate_exact_J_via_joint_montecarlo(y_MD, 500000, prng))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36d7eaec",
   "metadata": {},
   "source": [
    "## Option B: MC estimation of J using MARGINAL samples\n",
    "\n",
    "$$\n",
    "J(\\mathbf{y}, \\mathbf{\\Sigma}) = \\sum_{d=1}^D\n",
    "    \\mathbb{E}_{\\mathcal{N}(f_d|0, \\Sigma_{dd})}[ \\log p( y_d | f_d) ]\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "fc5471a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def estimate_exact_J_via_marginal_montecarlo(y_ND, n_samples, prng):\n",
    "    ''' Estimate objective J via MonteCarlo samples of f and exact J definition\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "    J_N : 1D array, size (N,)\n",
    "        Each entry equals the value of J for the corresponding row of input y_ND\n",
    "    '''\n",
    "    # Draw vector f using diagonal entries Sigma_dd of Sigma\n",
    "    # Look, no correlations!\n",
    "    DiagSigma_DD = np.diag(np.diag(Sigma_DD))\n",
    "    f_SD = draw_samples_of_f(n_samples, prng, Sigma_DD=DiagSigma_DD)\n",
    "    f_1DS = f_SD.T[np.newaxis,:]\n",
    "    \n",
    "    # After sampling f, this is the same logic as above for joint MC estimation\n",
    "    logpdf_NDS = scipy.stats.norm.logpdf(\n",
    "        y_ND[:,:,np.newaxis], f_1DS, tau)\n",
    "    J_N = logpdf_NDS.sum(axis=1).mean(axis=1)\n",
    "    return J_N"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d18ce544",
   "metadata": {},
   "source": [
    "### Using 500000 monte-carlo samples\n",
    "\n",
    "We should get some estimates that are very similar to the previous tests from Option A"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7106b964",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "J(y_1), ... J(y_10) using rand state 101\n",
      "[-575.53122886 -926.63550103 -294.0403738  -171.03335896 -257.15138263\n",
      " -165.86056846 -175.240645   -215.83788905 -924.44791067 -161.14257098]\n",
      "J(y_1), ... J(y_10) using rand state 202\n",
      "[-575.83488702 -927.16765046 -294.04520666 -170.81664131 -257.11728741\n",
      " -165.61584566 -174.97662361 -215.71943691 -924.99708671 -160.68102818]\n"
     ]
    }
   ],
   "source": [
    "for rand_state in [101, 202]:\n",
    "    prng = np.random.RandomState(rand_state)\n",
    "    print(\"J(y_1), ... J(y_10) using rand state %d\" % rand_state)\n",
    "    print(estimate_exact_J_via_marginal_montecarlo(y_MD, 500000, prng))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7efd3b13",
   "metadata": {},
   "source": [
    "## Option C: Noisy estimator using single-dim subsample\n",
    "\n",
    "This estimator works as follows:\n",
    "$$\n",
    "    u \\sim \\text{Unif}( 1, 2, \\ldots D)\n",
    "    \\\\ \\notag\n",
    "    \\hat{J}(u, \\mathbf{y}, \\mathbf{\\Sigma}) = \n",
    "    D \\cdot \\mathbb{E}_{f_u \\sim \\mathcal{N}(0, \\Sigma_{uu})} [ \\log p(y_u | f_u) ]\n",
    "$$\n",
    "\n",
    "Rescaling by $D$ here is crucial for correctness."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "ec175b96",
   "metadata": {},
   "outputs": [],
   "source": [
    "def estimate_noisy_J_via_montecarlo(y_ND, n_samples, prng):\n",
    "    catu_S = prng.choice(np.arange(D), size=n_samples)\n",
    "    assert catu_S.size == n_samples\n",
    "\n",
    "    stddev_sigma_D = np.sqrt(np.diag(Sigma_DD))\n",
    "    stddev_sigma_S = stddev_sigma_D[catu_S]\n",
    "\n",
    "    # Draw each f from its marginal N(0, \\Sigma_{uu})\n",
    "    # using the trick of a standard norm sample multipled by stddev\n",
    "    f_S = prng.randn(n_samples) * stddev_sigma_S\n",
    "\n",
    "    # Build array yy_NS of the entries of y that correspond to selected dims\n",
    "    onehotu_SD = np.zeros((n_samples, D), dtype=np.float64)\n",
    "    onehotu_SD[(np.arange(n_samples), catu_S)] = 1\n",
    "    yy_NS = np.dot(y_ND, onehotu_SD.T)\n",
    "    assert yy_NS.shape == (y_ND.shape[0], n_samples)\n",
    "    \n",
    "    # Now evaluate the normal log pdf at each row and each sample\n",
    "    logpdf_NS = scipy.stats.norm.logpdf(\n",
    "        yy_NS, f_S[np.newaxis,:], tau)\n",
    "    hatJ_N = D * logpdf_NS.mean(axis=1) # rescaling by D\n",
    "    return hatJ_N"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9bfcb93a",
   "metadata": {},
   "source": [
    "### Bakeoff: Compare the 3 estimators of J \n",
    "\n",
    "We'll try several different:\n",
    "\n",
    "* number of MC samples (denoted $S$)\n",
    "* repetitions (to see how estimators vary across different random seeds)\n",
    "\n",
    "At each of these settings, we'll compute estimates of $J$ at the $M$ different $y$ vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "355a0711",
   "metadata": {},
   "outputs": [],
   "source": [
    "row_dict_list = list()\n",
    "\n",
    "for S in [10, 100, 1000, 10000, 100000, 333333, 1000000]:\n",
    "    for rep in [1, 2, 3, 4, 5]:\n",
    "        prngA = np.random.RandomState(rep)\n",
    "        prngB = np.random.RandomState(rep)\n",
    "        prngC = np.random.RandomState(rep)\n",
    "\n",
    "        JA_M = estimate_exact_J_via_joint_montecarlo(y_MD, S, prngA)\n",
    "        JB_M = estimate_exact_J_via_marginal_montecarlo(y_MD, S, prngB)\n",
    "        JC_M = estimate_noisy_J_via_montecarlo(y_MD, S, prngC)\n",
    "\n",
    "        res_dict = dict(\n",
    "            S=S, rep=rep,\n",
    "            )\n",
    "        for m in range(M):\n",
    "            res_dict['JA_for_y%02d' % m] = JA_M[m]\n",
    "            res_dict['JB_for_y%02d' % m] = JB_M[m]\n",
    "            res_dict['JC_for_y%02d' % m] = JC_M[m]\n",
    "        row_dict_list.append(res_dict)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "c948cb96",
   "metadata": {},
   "outputs": [],
   "source": [
    "res_df = pd.DataFrame(row_dict_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "2c3a17c5",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>S</th>\n",
       "      <th>rep</th>\n",
       "      <th>JA_for_y00</th>\n",
       "      <th>JB_for_y00</th>\n",
       "      <th>JC_for_y00</th>\n",
       "      <th>JA_for_y01</th>\n",
       "      <th>JB_for_y01</th>\n",
       "      <th>JC_for_y01</th>\n",
       "      <th>JA_for_y02</th>\n",
       "      <th>JB_for_y02</th>\n",
       "      <th>...</th>\n",
       "      <th>JC_for_y06</th>\n",
       "      <th>JA_for_y07</th>\n",
       "      <th>JB_for_y07</th>\n",
       "      <th>JC_for_y07</th>\n",
       "      <th>JA_for_y08</th>\n",
       "      <th>JB_for_y08</th>\n",
       "      <th>JC_for_y08</th>\n",
       "      <th>JA_for_y09</th>\n",
       "      <th>JB_for_y09</th>\n",
       "      <th>JC_for_y09</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>-504.599944</td>\n",
       "      <td>-544.264204</td>\n",
       "      <td>-540.585473</td>\n",
       "      <td>-824.147746</td>\n",
       "      <td>-879.313596</td>\n",
       "      <td>-836.685851</td>\n",
       "      <td>-259.039507</td>\n",
       "      <td>-273.744322</td>\n",
       "      <td>...</td>\n",
       "      <td>-203.522563</td>\n",
       "      <td>-197.383789</td>\n",
       "      <td>-201.740845</td>\n",
       "      <td>-228.025428</td>\n",
       "      <td>-823.205262</td>\n",
       "      <td>-895.201292</td>\n",
       "      <td>-936.007681</td>\n",
       "      <td>-192.159264</td>\n",
       "      <td>-163.901870</td>\n",
       "      <td>-191.383041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>-633.396705</td>\n",
       "      <td>-407.102428</td>\n",
       "      <td>-763.572720</td>\n",
       "      <td>-1023.030494</td>\n",
       "      <td>-702.024901</td>\n",
       "      <td>-1191.571636</td>\n",
       "      <td>-305.437541</td>\n",
       "      <td>-198.918735</td>\n",
       "      <td>...</td>\n",
       "      <td>-304.240798</td>\n",
       "      <td>-206.984719</td>\n",
       "      <td>-150.761926</td>\n",
       "      <td>-330.839373</td>\n",
       "      <td>-1019.580342</td>\n",
       "      <td>-697.788793</td>\n",
       "      <td>-1135.197254</td>\n",
       "      <td>-94.234316</td>\n",
       "      <td>-178.685754</td>\n",
       "      <td>-155.668188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>-603.768243</td>\n",
       "      <td>-487.469855</td>\n",
       "      <td>-410.639103</td>\n",
       "      <td>-965.081168</td>\n",
       "      <td>-816.863299</td>\n",
       "      <td>-692.788195</td>\n",
       "      <td>-309.709773</td>\n",
       "      <td>-233.215351</td>\n",
       "      <td>...</td>\n",
       "      <td>-155.763340</td>\n",
       "      <td>-226.136786</td>\n",
       "      <td>-166.354461</td>\n",
       "      <td>-155.581827</td>\n",
       "      <td>-962.587568</td>\n",
       "      <td>-815.990242</td>\n",
       "      <td>-695.239242</td>\n",
       "      <td>-156.287394</td>\n",
       "      <td>-142.790655</td>\n",
       "      <td>-212.520458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>-482.904526</td>\n",
       "      <td>-583.419262</td>\n",
       "      <td>-432.683418</td>\n",
       "      <td>-805.221630</td>\n",
       "      <td>-934.070839</td>\n",
       "      <td>-762.216208</td>\n",
       "      <td>-234.226755</td>\n",
       "      <td>-289.321111</td>\n",
       "      <td>...</td>\n",
       "      <td>-87.338233</td>\n",
       "      <td>-171.028031</td>\n",
       "      <td>-208.950520</td>\n",
       "      <td>-117.412501</td>\n",
       "      <td>-804.899889</td>\n",
       "      <td>-950.563449</td>\n",
       "      <td>-744.999605</td>\n",
       "      <td>-160.775222</td>\n",
       "      <td>-151.329368</td>\n",
       "      <td>-106.433889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>-709.752998</td>\n",
       "      <td>-554.744192</td>\n",
       "      <td>-682.566045</td>\n",
       "      <td>-1129.075044</td>\n",
       "      <td>-904.294981</td>\n",
       "      <td>-1051.320059</td>\n",
       "      <td>-348.191704</td>\n",
       "      <td>-275.186319</td>\n",
       "      <td>...</td>\n",
       "      <td>-317.761440</td>\n",
       "      <td>-233.947298</td>\n",
       "      <td>-199.635835</td>\n",
       "      <td>-338.572527</td>\n",
       "      <td>-1126.260826</td>\n",
       "      <td>-887.455564</td>\n",
       "      <td>-1015.689541</td>\n",
       "      <td>-72.842793</td>\n",
       "      <td>-160.340910</td>\n",
       "      <td>-267.303353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>100</td>\n",
       "      <td>1</td>\n",
       "      <td>-566.219846</td>\n",
       "      <td>-599.499154</td>\n",
       "      <td>-583.822087</td>\n",
       "      <td>-914.516777</td>\n",
       "      <td>-964.468152</td>\n",
       "      <td>-946.682944</td>\n",
       "      <td>-288.387351</td>\n",
       "      <td>-301.385293</td>\n",
       "      <td>...</td>\n",
       "      <td>-136.290110</td>\n",
       "      <td>-211.662654</td>\n",
       "      <td>-216.112790</td>\n",
       "      <td>-190.054496</td>\n",
       "      <td>-912.415363</td>\n",
       "      <td>-960.092263</td>\n",
       "      <td>-969.243499</td>\n",
       "      <td>-160.932174</td>\n",
       "      <td>-141.976006</td>\n",
       "      <td>-99.873156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>100</td>\n",
       "      <td>2</td>\n",
       "      <td>-653.867152</td>\n",
       "      <td>-563.618621</td>\n",
       "      <td>-574.426610</td>\n",
       "      <td>-1020.510590</td>\n",
       "      <td>-906.375594</td>\n",
       "      <td>-942.854252</td>\n",
       "      <td>-353.980681</td>\n",
       "      <td>-293.780362</td>\n",
       "      <td>...</td>\n",
       "      <td>-144.550750</td>\n",
       "      <td>-267.594518</td>\n",
       "      <td>-220.392153</td>\n",
       "      <td>-193.276546</td>\n",
       "      <td>-1017.962878</td>\n",
       "      <td>-901.113374</td>\n",
       "      <td>-924.798213</td>\n",
       "      <td>-189.009752</td>\n",
       "      <td>-179.689207</td>\n",
       "      <td>-124.477771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>100</td>\n",
       "      <td>3</td>\n",
       "      <td>-527.393471</td>\n",
       "      <td>-606.609245</td>\n",
       "      <td>-566.645229</td>\n",
       "      <td>-862.765895</td>\n",
       "      <td>-967.301501</td>\n",
       "      <td>-919.197835</td>\n",
       "      <td>-264.767375</td>\n",
       "      <td>-316.253357</td>\n",
       "      <td>...</td>\n",
       "      <td>-153.320607</td>\n",
       "      <td>-194.913134</td>\n",
       "      <td>-233.096642</td>\n",
       "      <td>-198.138268</td>\n",
       "      <td>-860.448629</td>\n",
       "      <td>-964.243889</td>\n",
       "      <td>-933.449665</td>\n",
       "      <td>-164.605414</td>\n",
       "      <td>-161.773831</td>\n",
       "      <td>-134.993448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>100</td>\n",
       "      <td>4</td>\n",
       "      <td>-552.963947</td>\n",
       "      <td>-616.276417</td>\n",
       "      <td>-578.325265</td>\n",
       "      <td>-893.022381</td>\n",
       "      <td>-980.645857</td>\n",
       "      <td>-941.544693</td>\n",
       "      <td>-284.595018</td>\n",
       "      <td>-316.230209</td>\n",
       "      <td>...</td>\n",
       "      <td>-173.230261</td>\n",
       "      <td>-212.182040</td>\n",
       "      <td>-230.712288</td>\n",
       "      <td>-212.434045</td>\n",
       "      <td>-891.315362</td>\n",
       "      <td>-980.657077</td>\n",
       "      <td>-915.997349</td>\n",
       "      <td>-174.222637</td>\n",
       "      <td>-156.215860</td>\n",
       "      <td>-150.429715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>100</td>\n",
       "      <td>5</td>\n",
       "      <td>-569.297038</td>\n",
       "      <td>-564.259051</td>\n",
       "      <td>-528.128732</td>\n",
       "      <td>-916.814200</td>\n",
       "      <td>-914.834462</td>\n",
       "      <td>-875.515050</td>\n",
       "      <td>-291.854327</td>\n",
       "      <td>-281.897050</td>\n",
       "      <td>...</td>\n",
       "      <td>-126.246040</td>\n",
       "      <td>-215.522220</td>\n",
       "      <td>-203.666385</td>\n",
       "      <td>-171.150431</td>\n",
       "      <td>-914.850892</td>\n",
       "      <td>-915.046792</td>\n",
       "      <td>-867.352899</td>\n",
       "      <td>-166.412423</td>\n",
       "      <td>-149.138767</td>\n",
       "      <td>-132.801446</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1000</td>\n",
       "      <td>1</td>\n",
       "      <td>-562.016913</td>\n",
       "      <td>-581.866984</td>\n",
       "      <td>-604.004425</td>\n",
       "      <td>-909.908136</td>\n",
       "      <td>-935.859304</td>\n",
       "      <td>-965.948466</td>\n",
       "      <td>-284.347752</td>\n",
       "      <td>-297.064635</td>\n",
       "      <td>...</td>\n",
       "      <td>-183.186599</td>\n",
       "      <td>-207.842773</td>\n",
       "      <td>-217.324658</td>\n",
       "      <td>-227.922659</td>\n",
       "      <td>-907.752921</td>\n",
       "      <td>-933.742012</td>\n",
       "      <td>-963.278154</td>\n",
       "      <td>-158.106786</td>\n",
       "      <td>-157.960330</td>\n",
       "      <td>-155.297833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1000</td>\n",
       "      <td>2</td>\n",
       "      <td>-606.706797</td>\n",
       "      <td>-571.064737</td>\n",
       "      <td>-590.301581</td>\n",
       "      <td>-967.279768</td>\n",
       "      <td>-920.530537</td>\n",
       "      <td>-946.468805</td>\n",
       "      <td>-314.049703</td>\n",
       "      <td>-291.912708</td>\n",
       "      <td>...</td>\n",
       "      <td>-189.501974</td>\n",
       "      <td>-230.849224</td>\n",
       "      <td>-214.826884</td>\n",
       "      <td>-227.885652</td>\n",
       "      <td>-964.965416</td>\n",
       "      <td>-916.366417</td>\n",
       "      <td>-931.064139</td>\n",
       "      <td>-161.471651</td>\n",
       "      <td>-164.052031</td>\n",
       "      <td>-172.331390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1000</td>\n",
       "      <td>3</td>\n",
       "      <td>-585.709580</td>\n",
       "      <td>-564.689133</td>\n",
       "      <td>-567.664480</td>\n",
       "      <td>-941.429815</td>\n",
       "      <td>-912.562477</td>\n",
       "      <td>-913.941178</td>\n",
       "      <td>-298.619050</td>\n",
       "      <td>-286.309300</td>\n",
       "      <td>...</td>\n",
       "      <td>-178.872108</td>\n",
       "      <td>-217.965238</td>\n",
       "      <td>-209.754116</td>\n",
       "      <td>-216.617971</td>\n",
       "      <td>-939.291905</td>\n",
       "      <td>-910.851347</td>\n",
       "      <td>-915.146265</td>\n",
       "      <td>-156.167795</td>\n",
       "      <td>-160.375636</td>\n",
       "      <td>-167.461381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1000</td>\n",
       "      <td>4</td>\n",
       "      <td>-581.655949</td>\n",
       "      <td>-574.379631</td>\n",
       "      <td>-589.358752</td>\n",
       "      <td>-934.163602</td>\n",
       "      <td>-927.494195</td>\n",
       "      <td>-943.102934</td>\n",
       "      <td>-298.441650</td>\n",
       "      <td>-289.053238</td>\n",
       "      <td>...</td>\n",
       "      <td>-178.105494</td>\n",
       "      <td>-219.473433</td>\n",
       "      <td>-209.691271</td>\n",
       "      <td>-222.129912</td>\n",
       "      <td>-932.154064</td>\n",
       "      <td>-925.999704</td>\n",
       "      <td>-943.683674</td>\n",
       "      <td>-162.494299</td>\n",
       "      <td>-152.689273</td>\n",
       "      <td>-161.430565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1000</td>\n",
       "      <td>5</td>\n",
       "      <td>-557.680345</td>\n",
       "      <td>-573.923961</td>\n",
       "      <td>-571.223055</td>\n",
       "      <td>-906.388499</td>\n",
       "      <td>-924.579181</td>\n",
       "      <td>-925.423794</td>\n",
       "      <td>-278.928720</td>\n",
       "      <td>-292.729738</td>\n",
       "      <td>...</td>\n",
       "      <td>-172.165369</td>\n",
       "      <td>-201.986951</td>\n",
       "      <td>-214.644501</td>\n",
       "      <td>-210.141622</td>\n",
       "      <td>-904.261501</td>\n",
       "      <td>-923.346429</td>\n",
       "      <td>-915.005224</td>\n",
       "      <td>-151.067394</td>\n",
       "      <td>-160.029925</td>\n",
       "      <td>-153.395357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>10000</td>\n",
       "      <td>1</td>\n",
       "      <td>-576.524419</td>\n",
       "      <td>-576.504283</td>\n",
       "      <td>-586.295566</td>\n",
       "      <td>-927.958920</td>\n",
       "      <td>-928.088192</td>\n",
       "      <td>-941.266684</td>\n",
       "      <td>-294.604317</td>\n",
       "      <td>-294.211545</td>\n",
       "      <td>...</td>\n",
       "      <td>-176.640137</td>\n",
       "      <td>-216.222099</td>\n",
       "      <td>-215.735577</td>\n",
       "      <td>-219.811130</td>\n",
       "      <td>-925.808509</td>\n",
       "      <td>-926.045927</td>\n",
       "      <td>-937.554263</td>\n",
       "      <td>-161.017099</td>\n",
       "      <td>-160.395470</td>\n",
       "      <td>-160.367822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>10000</td>\n",
       "      <td>2</td>\n",
       "      <td>-578.788448</td>\n",
       "      <td>-575.481293</td>\n",
       "      <td>-579.969955</td>\n",
       "      <td>-930.757935</td>\n",
       "      <td>-926.505687</td>\n",
       "      <td>-932.437494</td>\n",
       "      <td>-296.274060</td>\n",
       "      <td>-294.592438</td>\n",
       "      <td>...</td>\n",
       "      <td>-176.488624</td>\n",
       "      <td>-217.620719</td>\n",
       "      <td>-216.501513</td>\n",
       "      <td>-217.649576</td>\n",
       "      <td>-928.511815</td>\n",
       "      <td>-923.786772</td>\n",
       "      <td>-931.099879</td>\n",
       "      <td>-161.628855</td>\n",
       "      <td>-161.901414</td>\n",
       "      <td>-159.494635</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>10000</td>\n",
       "      <td>3</td>\n",
       "      <td>-581.903733</td>\n",
       "      <td>-569.686028</td>\n",
       "      <td>-576.070964</td>\n",
       "      <td>-936.404169</td>\n",
       "      <td>-919.268266</td>\n",
       "      <td>-927.816766</td>\n",
       "      <td>-296.319629</td>\n",
       "      <td>-289.641746</td>\n",
       "      <td>...</td>\n",
       "      <td>-175.444753</td>\n",
       "      <td>-216.315772</td>\n",
       "      <td>-212.239326</td>\n",
       "      <td>-215.750634</td>\n",
       "      <td>-934.236934</td>\n",
       "      <td>-917.111570</td>\n",
       "      <td>-925.184229</td>\n",
       "      <td>-156.383104</td>\n",
       "      <td>-160.240203</td>\n",
       "      <td>-160.231947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>10000</td>\n",
       "      <td>4</td>\n",
       "      <td>-569.865408</td>\n",
       "      <td>-573.794274</td>\n",
       "      <td>-575.102009</td>\n",
       "      <td>-919.818881</td>\n",
       "      <td>-925.034754</td>\n",
       "      <td>-926.199753</td>\n",
       "      <td>-289.688754</td>\n",
       "      <td>-292.164721</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.231817</td>\n",
       "      <td>-212.093011</td>\n",
       "      <td>-213.851336</td>\n",
       "      <td>-215.152917</td>\n",
       "      <td>-917.650316</td>\n",
       "      <td>-923.158839</td>\n",
       "      <td>-925.984416</td>\n",
       "      <td>-159.221467</td>\n",
       "      <td>-158.643343</td>\n",
       "      <td>-159.240305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>10000</td>\n",
       "      <td>5</td>\n",
       "      <td>-573.544258</td>\n",
       "      <td>-578.902726</td>\n",
       "      <td>-577.510750</td>\n",
       "      <td>-924.271802</td>\n",
       "      <td>-931.013185</td>\n",
       "      <td>-929.653767</td>\n",
       "      <td>-292.482619</td>\n",
       "      <td>-296.186915</td>\n",
       "      <td>...</td>\n",
       "      <td>-176.771049</td>\n",
       "      <td>-214.476997</td>\n",
       "      <td>-217.451654</td>\n",
       "      <td>-217.409681</td>\n",
       "      <td>-922.104504</td>\n",
       "      <td>-928.823691</td>\n",
       "      <td>-925.992972</td>\n",
       "      <td>-160.367160</td>\n",
       "      <td>-161.223882</td>\n",
       "      <td>-162.316482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>100000</td>\n",
       "      <td>1</td>\n",
       "      <td>-574.297784</td>\n",
       "      <td>-577.051516</td>\n",
       "      <td>-575.518226</td>\n",
       "      <td>-925.254959</td>\n",
       "      <td>-928.860162</td>\n",
       "      <td>-927.012102</td>\n",
       "      <td>-292.960297</td>\n",
       "      <td>-294.731272</td>\n",
       "      <td>...</td>\n",
       "      <td>-173.994741</td>\n",
       "      <td>-214.833960</td>\n",
       "      <td>-216.158328</td>\n",
       "      <td>-214.984632</td>\n",
       "      <td>-923.080399</td>\n",
       "      <td>-926.652467</td>\n",
       "      <td>-924.942234</td>\n",
       "      <td>-160.376802</td>\n",
       "      <td>-160.379071</td>\n",
       "      <td>-159.853806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>100000</td>\n",
       "      <td>2</td>\n",
       "      <td>-576.667510</td>\n",
       "      <td>-573.992584</td>\n",
       "      <td>-571.956151</td>\n",
       "      <td>-928.599385</td>\n",
       "      <td>-924.766107</td>\n",
       "      <td>-921.767141</td>\n",
       "      <td>-294.144316</td>\n",
       "      <td>-292.729540</td>\n",
       "      <td>...</td>\n",
       "      <td>-173.685564</td>\n",
       "      <td>-215.500443</td>\n",
       "      <td>-214.682201</td>\n",
       "      <td>-213.882756</td>\n",
       "      <td>-926.434496</td>\n",
       "      <td>-922.731147</td>\n",
       "      <td>-921.030253</td>\n",
       "      <td>-159.547601</td>\n",
       "      <td>-160.528426</td>\n",
       "      <td>-160.416527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>100000</td>\n",
       "      <td>3</td>\n",
       "      <td>-576.171304</td>\n",
       "      <td>-574.729337</td>\n",
       "      <td>-573.782895</td>\n",
       "      <td>-927.881100</td>\n",
       "      <td>-925.910395</td>\n",
       "      <td>-924.195729</td>\n",
       "      <td>-293.930867</td>\n",
       "      <td>-293.131589</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.390515</td>\n",
       "      <td>-215.406601</td>\n",
       "      <td>-214.891622</td>\n",
       "      <td>-214.875478</td>\n",
       "      <td>-925.701048</td>\n",
       "      <td>-923.696591</td>\n",
       "      <td>-922.726926</td>\n",
       "      <td>-159.793604</td>\n",
       "      <td>-160.115921</td>\n",
       "      <td>-161.004783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>100000</td>\n",
       "      <td>4</td>\n",
       "      <td>-577.726847</td>\n",
       "      <td>-573.166664</td>\n",
       "      <td>-574.244118</td>\n",
       "      <td>-929.945356</td>\n",
       "      <td>-923.631285</td>\n",
       "      <td>-924.809421</td>\n",
       "      <td>-294.864300</td>\n",
       "      <td>-292.360224</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.989405</td>\n",
       "      <td>-216.070411</td>\n",
       "      <td>-214.486692</td>\n",
       "      <td>-215.361236</td>\n",
       "      <td>-927.766293</td>\n",
       "      <td>-921.518576</td>\n",
       "      <td>-922.797429</td>\n",
       "      <td>-159.683653</td>\n",
       "      <td>-160.793394</td>\n",
       "      <td>-161.592596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>100000</td>\n",
       "      <td>5</td>\n",
       "      <td>-575.809622</td>\n",
       "      <td>-576.379276</td>\n",
       "      <td>-576.034436</td>\n",
       "      <td>-927.458318</td>\n",
       "      <td>-927.895551</td>\n",
       "      <td>-927.697352</td>\n",
       "      <td>-293.649321</td>\n",
       "      <td>-294.359473</td>\n",
       "      <td>...</td>\n",
       "      <td>-175.045348</td>\n",
       "      <td>-215.156932</td>\n",
       "      <td>-215.946649</td>\n",
       "      <td>-215.742074</td>\n",
       "      <td>-925.282463</td>\n",
       "      <td>-925.643291</td>\n",
       "      <td>-925.460761</td>\n",
       "      <td>-159.627647</td>\n",
       "      <td>-160.707985</td>\n",
       "      <td>-159.997986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>333333</td>\n",
       "      <td>1</td>\n",
       "      <td>-575.251485</td>\n",
       "      <td>-576.239894</td>\n",
       "      <td>-575.507605</td>\n",
       "      <td>-926.560265</td>\n",
       "      <td>-927.738229</td>\n",
       "      <td>-926.755714</td>\n",
       "      <td>-293.489031</td>\n",
       "      <td>-294.238732</td>\n",
       "      <td>...</td>\n",
       "      <td>-175.067989</td>\n",
       "      <td>-215.176192</td>\n",
       "      <td>-215.825248</td>\n",
       "      <td>-215.662971</td>\n",
       "      <td>-924.387800</td>\n",
       "      <td>-925.567247</td>\n",
       "      <td>-924.678354</td>\n",
       "      <td>-160.178004</td>\n",
       "      <td>-160.544591</td>\n",
       "      <td>-160.593056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>333333</td>\n",
       "      <td>2</td>\n",
       "      <td>-576.087727</td>\n",
       "      <td>-575.306313</td>\n",
       "      <td>-574.381483</td>\n",
       "      <td>-927.535516</td>\n",
       "      <td>-926.546086</td>\n",
       "      <td>-925.291259</td>\n",
       "      <td>-294.159986</td>\n",
       "      <td>-293.693804</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.062404</td>\n",
       "      <td>-215.774996</td>\n",
       "      <td>-215.423810</td>\n",
       "      <td>-214.745888</td>\n",
       "      <td>-925.351327</td>\n",
       "      <td>-924.326289</td>\n",
       "      <td>-923.344278</td>\n",
       "      <td>-160.571405</td>\n",
       "      <td>-160.507672</td>\n",
       "      <td>-160.333651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>333333</td>\n",
       "      <td>3</td>\n",
       "      <td>-573.684754</td>\n",
       "      <td>-576.485423</td>\n",
       "      <td>-575.293270</td>\n",
       "      <td>-924.347570</td>\n",
       "      <td>-928.205933</td>\n",
       "      <td>-926.428182</td>\n",
       "      <td>-292.688411</td>\n",
       "      <td>-294.249862</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.494161</td>\n",
       "      <td>-214.717292</td>\n",
       "      <td>-215.710653</td>\n",
       "      <td>-215.295989</td>\n",
       "      <td>-922.177929</td>\n",
       "      <td>-926.100969</td>\n",
       "      <td>-924.597686</td>\n",
       "      <td>-160.721634</td>\n",
       "      <td>-159.996550</td>\n",
       "      <td>-160.618385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>333333</td>\n",
       "      <td>4</td>\n",
       "      <td>-576.488753</td>\n",
       "      <td>-574.526585</td>\n",
       "      <td>-573.624745</td>\n",
       "      <td>-928.097821</td>\n",
       "      <td>-925.414481</td>\n",
       "      <td>-924.356793</td>\n",
       "      <td>-294.357882</td>\n",
       "      <td>-293.281373</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.022539</td>\n",
       "      <td>-215.886933</td>\n",
       "      <td>-215.186336</td>\n",
       "      <td>-214.517232</td>\n",
       "      <td>-925.917515</td>\n",
       "      <td>-923.283548</td>\n",
       "      <td>-922.222678</td>\n",
       "      <td>-160.440703</td>\n",
       "      <td>-160.788010</td>\n",
       "      <td>-160.442088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>333333</td>\n",
       "      <td>5</td>\n",
       "      <td>-577.541562</td>\n",
       "      <td>-576.092478</td>\n",
       "      <td>-576.306150</td>\n",
       "      <td>-929.532601</td>\n",
       "      <td>-927.461851</td>\n",
       "      <td>-927.667825</td>\n",
       "      <td>-294.969300</td>\n",
       "      <td>-294.156462</td>\n",
       "      <td>...</td>\n",
       "      <td>-175.327471</td>\n",
       "      <td>-216.294648</td>\n",
       "      <td>-215.815823</td>\n",
       "      <td>-216.074305</td>\n",
       "      <td>-927.364280</td>\n",
       "      <td>-925.250315</td>\n",
       "      <td>-925.753047</td>\n",
       "      <td>-160.231887</td>\n",
       "      <td>-160.859092</td>\n",
       "      <td>-160.816860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>1000000</td>\n",
       "      <td>1</td>\n",
       "      <td>-575.804492</td>\n",
       "      <td>-575.933666</td>\n",
       "      <td>-576.422700</td>\n",
       "      <td>-927.222742</td>\n",
       "      <td>-927.289643</td>\n",
       "      <td>-927.925308</td>\n",
       "      <td>-293.909065</td>\n",
       "      <td>-294.096814</td>\n",
       "      <td>...</td>\n",
       "      <td>-175.067821</td>\n",
       "      <td>-215.538330</td>\n",
       "      <td>-215.757228</td>\n",
       "      <td>-215.955805</td>\n",
       "      <td>-925.049539</td>\n",
       "      <td>-925.130374</td>\n",
       "      <td>-925.705481</td>\n",
       "      <td>-160.373591</td>\n",
       "      <td>-160.692106</td>\n",
       "      <td>-160.790978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>1000000</td>\n",
       "      <td>2</td>\n",
       "      <td>-576.345343</td>\n",
       "      <td>-575.622218</td>\n",
       "      <td>-575.679025</td>\n",
       "      <td>-927.809252</td>\n",
       "      <td>-926.866017</td>\n",
       "      <td>-927.047273</td>\n",
       "      <td>-294.395957</td>\n",
       "      <td>-293.962440</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.852378</td>\n",
       "      <td>-216.001790</td>\n",
       "      <td>-215.685728</td>\n",
       "      <td>-215.562747</td>\n",
       "      <td>-925.630072</td>\n",
       "      <td>-924.680993</td>\n",
       "      <td>-924.863904</td>\n",
       "      <td>-160.771157</td>\n",
       "      <td>-160.774451</td>\n",
       "      <td>-160.420239</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>1000000</td>\n",
       "      <td>3</td>\n",
       "      <td>-574.795840</td>\n",
       "      <td>-576.124152</td>\n",
       "      <td>-575.475049</td>\n",
       "      <td>-925.893284</td>\n",
       "      <td>-927.625433</td>\n",
       "      <td>-926.650895</td>\n",
       "      <td>-293.281813</td>\n",
       "      <td>-294.127444</td>\n",
       "      <td>...</td>\n",
       "      <td>-174.922671</td>\n",
       "      <td>-215.080754</td>\n",
       "      <td>-215.707634</td>\n",
       "      <td>-215.590657</td>\n",
       "      <td>-923.721794</td>\n",
       "      <td>-925.493003</td>\n",
       "      <td>-924.515728</td>\n",
       "      <td>-160.412710</td>\n",
       "      <td>-160.380114</td>\n",
       "      <td>-160.829327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>1000000</td>\n",
       "      <td>4</td>\n",
       "      <td>-576.257743</td>\n",
       "      <td>-575.724558</td>\n",
       "      <td>-575.478194</td>\n",
       "      <td>-927.699038</td>\n",
       "      <td>-927.004579</td>\n",
       "      <td>-926.705098</td>\n",
       "      <td>-294.337719</td>\n",
       "      <td>-294.047789</td>\n",
       "      <td>...</td>\n",
       "      <td>-175.094499</td>\n",
       "      <td>-215.955899</td>\n",
       "      <td>-215.755421</td>\n",
       "      <td>-215.678559</td>\n",
       "      <td>-925.516991</td>\n",
       "      <td>-924.792517</td>\n",
       "      <td>-924.275483</td>\n",
       "      <td>-160.760485</td>\n",
       "      <td>-160.786151</td>\n",
       "      <td>-160.966874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>1000000</td>\n",
       "      <td>5</td>\n",
       "      <td>-575.625466</td>\n",
       "      <td>-576.482836</td>\n",
       "      <td>-576.626751</td>\n",
       "      <td>-927.002951</td>\n",
       "      <td>-928.066221</td>\n",
       "      <td>-928.324998</td>\n",
       "      <td>-293.785086</td>\n",
       "      <td>-294.381684</td>\n",
       "      <td>...</td>\n",
       "      <td>-175.135508</td>\n",
       "      <td>-215.435956</td>\n",
       "      <td>-215.927515</td>\n",
       "      <td>-215.980354</td>\n",
       "      <td>-924.829921</td>\n",
       "      <td>-925.860393</td>\n",
       "      <td>-926.163106</td>\n",
       "      <td>-160.327846</td>\n",
       "      <td>-160.544750</td>\n",
       "      <td>-160.345647</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>35 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          S  rep  JA_for_y00  JB_for_y00  JC_for_y00   JA_for_y01  JB_for_y01  \\\n",
       "0        10    1 -504.599944 -544.264204 -540.585473  -824.147746 -879.313596   \n",
       "1        10    2 -633.396705 -407.102428 -763.572720 -1023.030494 -702.024901   \n",
       "2        10    3 -603.768243 -487.469855 -410.639103  -965.081168 -816.863299   \n",
       "3        10    4 -482.904526 -583.419262 -432.683418  -805.221630 -934.070839   \n",
       "4        10    5 -709.752998 -554.744192 -682.566045 -1129.075044 -904.294981   \n",
       "5       100    1 -566.219846 -599.499154 -583.822087  -914.516777 -964.468152   \n",
       "6       100    2 -653.867152 -563.618621 -574.426610 -1020.510590 -906.375594   \n",
       "7       100    3 -527.393471 -606.609245 -566.645229  -862.765895 -967.301501   \n",
       "8       100    4 -552.963947 -616.276417 -578.325265  -893.022381 -980.645857   \n",
       "9       100    5 -569.297038 -564.259051 -528.128732  -916.814200 -914.834462   \n",
       "10     1000    1 -562.016913 -581.866984 -604.004425  -909.908136 -935.859304   \n",
       "11     1000    2 -606.706797 -571.064737 -590.301581  -967.279768 -920.530537   \n",
       "12     1000    3 -585.709580 -564.689133 -567.664480  -941.429815 -912.562477   \n",
       "13     1000    4 -581.655949 -574.379631 -589.358752  -934.163602 -927.494195   \n",
       "14     1000    5 -557.680345 -573.923961 -571.223055  -906.388499 -924.579181   \n",
       "15    10000    1 -576.524419 -576.504283 -586.295566  -927.958920 -928.088192   \n",
       "16    10000    2 -578.788448 -575.481293 -579.969955  -930.757935 -926.505687   \n",
       "17    10000    3 -581.903733 -569.686028 -576.070964  -936.404169 -919.268266   \n",
       "18    10000    4 -569.865408 -573.794274 -575.102009  -919.818881 -925.034754   \n",
       "19    10000    5 -573.544258 -578.902726 -577.510750  -924.271802 -931.013185   \n",
       "20   100000    1 -574.297784 -577.051516 -575.518226  -925.254959 -928.860162   \n",
       "21   100000    2 -576.667510 -573.992584 -571.956151  -928.599385 -924.766107   \n",
       "22   100000    3 -576.171304 -574.729337 -573.782895  -927.881100 -925.910395   \n",
       "23   100000    4 -577.726847 -573.166664 -574.244118  -929.945356 -923.631285   \n",
       "24   100000    5 -575.809622 -576.379276 -576.034436  -927.458318 -927.895551   \n",
       "25   333333    1 -575.251485 -576.239894 -575.507605  -926.560265 -927.738229   \n",
       "26   333333    2 -576.087727 -575.306313 -574.381483  -927.535516 -926.546086   \n",
       "27   333333    3 -573.684754 -576.485423 -575.293270  -924.347570 -928.205933   \n",
       "28   333333    4 -576.488753 -574.526585 -573.624745  -928.097821 -925.414481   \n",
       "29   333333    5 -577.541562 -576.092478 -576.306150  -929.532601 -927.461851   \n",
       "30  1000000    1 -575.804492 -575.933666 -576.422700  -927.222742 -927.289643   \n",
       "31  1000000    2 -576.345343 -575.622218 -575.679025  -927.809252 -926.866017   \n",
       "32  1000000    3 -574.795840 -576.124152 -575.475049  -925.893284 -927.625433   \n",
       "33  1000000    4 -576.257743 -575.724558 -575.478194  -927.699038 -927.004579   \n",
       "34  1000000    5 -575.625466 -576.482836 -576.626751  -927.002951 -928.066221   \n",
       "\n",
       "     JC_for_y01  JA_for_y02  JB_for_y02  ...  JC_for_y06  JA_for_y07  \\\n",
       "0   -836.685851 -259.039507 -273.744322  ... -203.522563 -197.383789   \n",
       "1  -1191.571636 -305.437541 -198.918735  ... -304.240798 -206.984719   \n",
       "2   -692.788195 -309.709773 -233.215351  ... -155.763340 -226.136786   \n",
       "3   -762.216208 -234.226755 -289.321111  ...  -87.338233 -171.028031   \n",
       "4  -1051.320059 -348.191704 -275.186319  ... -317.761440 -233.947298   \n",
       "5   -946.682944 -288.387351 -301.385293  ... -136.290110 -211.662654   \n",
       "6   -942.854252 -353.980681 -293.780362  ... -144.550750 -267.594518   \n",
       "7   -919.197835 -264.767375 -316.253357  ... -153.320607 -194.913134   \n",
       "8   -941.544693 -284.595018 -316.230209  ... -173.230261 -212.182040   \n",
       "9   -875.515050 -291.854327 -281.897050  ... -126.246040 -215.522220   \n",
       "10  -965.948466 -284.347752 -297.064635  ... -183.186599 -207.842773   \n",
       "11  -946.468805 -314.049703 -291.912708  ... -189.501974 -230.849224   \n",
       "12  -913.941178 -298.619050 -286.309300  ... -178.872108 -217.965238   \n",
       "13  -943.102934 -298.441650 -289.053238  ... -178.105494 -219.473433   \n",
       "14  -925.423794 -278.928720 -292.729738  ... -172.165369 -201.986951   \n",
       "15  -941.266684 -294.604317 -294.211545  ... -176.640137 -216.222099   \n",
       "16  -932.437494 -296.274060 -294.592438  ... -176.488624 -217.620719   \n",
       "17  -927.816766 -296.319629 -289.641746  ... -175.444753 -216.315772   \n",
       "18  -926.199753 -289.688754 -292.164721  ... -174.231817 -212.093011   \n",
       "19  -929.653767 -292.482619 -296.186915  ... -176.771049 -214.476997   \n",
       "20  -927.012102 -292.960297 -294.731272  ... -173.994741 -214.833960   \n",
       "21  -921.767141 -294.144316 -292.729540  ... -173.685564 -215.500443   \n",
       "22  -924.195729 -293.930867 -293.131589  ... -174.390515 -215.406601   \n",
       "23  -924.809421 -294.864300 -292.360224  ... -174.989405 -216.070411   \n",
       "24  -927.697352 -293.649321 -294.359473  ... -175.045348 -215.156932   \n",
       "25  -926.755714 -293.489031 -294.238732  ... -175.067989 -215.176192   \n",
       "26  -925.291259 -294.159986 -293.693804  ... -174.062404 -215.774996   \n",
       "27  -926.428182 -292.688411 -294.249862  ... -174.494161 -214.717292   \n",
       "28  -924.356793 -294.357882 -293.281373  ... -174.022539 -215.886933   \n",
       "29  -927.667825 -294.969300 -294.156462  ... -175.327471 -216.294648   \n",
       "30  -927.925308 -293.909065 -294.096814  ... -175.067821 -215.538330   \n",
       "31  -927.047273 -294.395957 -293.962440  ... -174.852378 -216.001790   \n",
       "32  -926.650895 -293.281813 -294.127444  ... -174.922671 -215.080754   \n",
       "33  -926.705098 -294.337719 -294.047789  ... -175.094499 -215.955899   \n",
       "34  -928.324998 -293.785086 -294.381684  ... -175.135508 -215.435956   \n",
       "\n",
       "    JB_for_y07  JC_for_y07   JA_for_y08  JB_for_y08   JC_for_y08  JA_for_y09  \\\n",
       "0  -201.740845 -228.025428  -823.205262 -895.201292  -936.007681 -192.159264   \n",
       "1  -150.761926 -330.839373 -1019.580342 -697.788793 -1135.197254  -94.234316   \n",
       "2  -166.354461 -155.581827  -962.587568 -815.990242  -695.239242 -156.287394   \n",
       "3  -208.950520 -117.412501  -804.899889 -950.563449  -744.999605 -160.775222   \n",
       "4  -199.635835 -338.572527 -1126.260826 -887.455564 -1015.689541  -72.842793   \n",
       "5  -216.112790 -190.054496  -912.415363 -960.092263  -969.243499 -160.932174   \n",
       "6  -220.392153 -193.276546 -1017.962878 -901.113374  -924.798213 -189.009752   \n",
       "7  -233.096642 -198.138268  -860.448629 -964.243889  -933.449665 -164.605414   \n",
       "8  -230.712288 -212.434045  -891.315362 -980.657077  -915.997349 -174.222637   \n",
       "9  -203.666385 -171.150431  -914.850892 -915.046792  -867.352899 -166.412423   \n",
       "10 -217.324658 -227.922659  -907.752921 -933.742012  -963.278154 -158.106786   \n",
       "11 -214.826884 -227.885652  -964.965416 -916.366417  -931.064139 -161.471651   \n",
       "12 -209.754116 -216.617971  -939.291905 -910.851347  -915.146265 -156.167795   \n",
       "13 -209.691271 -222.129912  -932.154064 -925.999704  -943.683674 -162.494299   \n",
       "14 -214.644501 -210.141622  -904.261501 -923.346429  -915.005224 -151.067394   \n",
       "15 -215.735577 -219.811130  -925.808509 -926.045927  -937.554263 -161.017099   \n",
       "16 -216.501513 -217.649576  -928.511815 -923.786772  -931.099879 -161.628855   \n",
       "17 -212.239326 -215.750634  -934.236934 -917.111570  -925.184229 -156.383104   \n",
       "18 -213.851336 -215.152917  -917.650316 -923.158839  -925.984416 -159.221467   \n",
       "19 -217.451654 -217.409681  -922.104504 -928.823691  -925.992972 -160.367160   \n",
       "20 -216.158328 -214.984632  -923.080399 -926.652467  -924.942234 -160.376802   \n",
       "21 -214.682201 -213.882756  -926.434496 -922.731147  -921.030253 -159.547601   \n",
       "22 -214.891622 -214.875478  -925.701048 -923.696591  -922.726926 -159.793604   \n",
       "23 -214.486692 -215.361236  -927.766293 -921.518576  -922.797429 -159.683653   \n",
       "24 -215.946649 -215.742074  -925.282463 -925.643291  -925.460761 -159.627647   \n",
       "25 -215.825248 -215.662971  -924.387800 -925.567247  -924.678354 -160.178004   \n",
       "26 -215.423810 -214.745888  -925.351327 -924.326289  -923.344278 -160.571405   \n",
       "27 -215.710653 -215.295989  -922.177929 -926.100969  -924.597686 -160.721634   \n",
       "28 -215.186336 -214.517232  -925.917515 -923.283548  -922.222678 -160.440703   \n",
       "29 -215.815823 -216.074305  -927.364280 -925.250315  -925.753047 -160.231887   \n",
       "30 -215.757228 -215.955805  -925.049539 -925.130374  -925.705481 -160.373591   \n",
       "31 -215.685728 -215.562747  -925.630072 -924.680993  -924.863904 -160.771157   \n",
       "32 -215.707634 -215.590657  -923.721794 -925.493003  -924.515728 -160.412710   \n",
       "33 -215.755421 -215.678559  -925.516991 -924.792517  -924.275483 -160.760485   \n",
       "34 -215.927515 -215.980354  -924.829921 -925.860393  -926.163106 -160.327846   \n",
       "\n",
       "    JB_for_y09  JC_for_y09  \n",
       "0  -163.901870 -191.383041  \n",
       "1  -178.685754 -155.668188  \n",
       "2  -142.790655 -212.520458  \n",
       "3  -151.329368 -106.433889  \n",
       "4  -160.340910 -267.303353  \n",
       "5  -141.976006  -99.873156  \n",
       "6  -179.689207 -124.477771  \n",
       "7  -161.773831 -134.993448  \n",
       "8  -156.215860 -150.429715  \n",
       "9  -149.138767 -132.801446  \n",
       "10 -157.960330 -155.297833  \n",
       "11 -164.052031 -172.331390  \n",
       "12 -160.375636 -167.461381  \n",
       "13 -152.689273 -161.430565  \n",
       "14 -160.029925 -153.395357  \n",
       "15 -160.395470 -160.367822  \n",
       "16 -161.901414 -159.494635  \n",
       "17 -160.240203 -160.231947  \n",
       "18 -158.643343 -159.240305  \n",
       "19 -161.223882 -162.316482  \n",
       "20 -160.379071 -159.853806  \n",
       "21 -160.528426 -160.416527  \n",
       "22 -160.115921 -161.004783  \n",
       "23 -160.793394 -161.592596  \n",
       "24 -160.707985 -159.997986  \n",
       "25 -160.544591 -160.593056  \n",
       "26 -160.507672 -160.333651  \n",
       "27 -159.996550 -160.618385  \n",
       "28 -160.788010 -160.442088  \n",
       "29 -160.859092 -160.816860  \n",
       "30 -160.692106 -160.790978  \n",
       "31 -160.774451 -160.420239  \n",
       "32 -160.380114 -160.829327  \n",
       "33 -160.786151 -160.966874  \n",
       "34 -160.544750 -160.345647  \n",
       "\n",
       "[35 rows x 32 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bdee1518",
   "metadata": {},
   "source": [
    "### Make a pretty plot comparing the estimators\n",
    "\n",
    "Each row shows the plot for a specific y-vector (denoted by index into the y_MD array)\n",
    "\n",
    "Each column corresponds to a different estimator (A, B, or C)\n",
    "\n",
    "Each panel shows, for a particular y vector and estimator, how the estimate of J changes as the number of MC samples increases.\n",
    "\n",
    "We should expect that each row's lines converge to the same J value at high number of samples (because of the UNBIASED properties we've proven)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "2ca29db4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_, grid = plt.subplots(\n",
    "    nrows=3, ncols=3,\n",
    "    figsize=(12, 8),\n",
    "    sharex=True, sharey=False)\n",
    "\n",
    "for row in [0,1,2]:\n",
    "    for col, prefix in [\n",
    "            (0, 'JA_for_y'),\n",
    "            (1, 'JB_for_y'),\n",
    "            (2, 'JC_for_y'),\n",
    "            ]:\n",
    "        ykey = prefix + \"%02d\" % row\n",
    "        ax = grid[row,col]\n",
    "        for rep in [1, 2, 3, 4, 5]:\n",
    "            r_df = res_df.query(\"rep == %d\" % rep)\n",
    "            ax.plot(r_df['S'], r_df[ykey], '.-', markersize=10)\n",
    "            ax.set_xscale('log')\n",
    "\n",
    "            if col == 0:\n",
    "                ax.set_ylabel('log p(y[%d])' % row)\n",
    "            if row == 0:\n",
    "                if col == 0:\n",
    "                    ax.set_title('Estimator A')\n",
    "                if col == 1:\n",
    "                    ax.set_title('Estimator B')\n",
    "                if col == 2:\n",
    "                    ax.set_title('Estimator C')\n",
    "            if row == 2:\n",
    "                ax.set_xlabel('num MC samples')\n",
    "\n",
    "plt.tight_layout();\n",
    "for row in range(grid.shape[0]):\n",
    "    # Standardize y lims in each row\n",
    "    l0,h0 = grid[row, 0].get_ylim()\n",
    "    l1,h1 = grid[row, 1].get_ylim()\n",
    "    l2,h2 = grid[row, 2].get_ylim()\n",
    "\n",
    "    aa = np.min([l0, l1, l2])\n",
    "    bb = np.max([h0, h1, h2])\n",
    "    for cc in [0,1,2]:\n",
    "        grid[row, cc].set_ylim(aa,bb)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "587b9c99",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}