floZ_SampleGeneration.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "8e535945-5b71-405d-a19e-eac4021b7e08",
   "metadata": {},
   "source": [
    "# $floZ$: Bayesian evidence estimation from posterior samples with normalizing flows\n",
    "## Run this notebook if you wish to reproduce the $floZ$ paper results.\n",
    "\n",
    "### Note our convention: we use the natural logarithm (base e), instead of the common log (base 10).\n",
    "### If you have used this code for your research, please cite https://arxiv.org/abs/2404.12294"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "30bf8f81-c208-46d2-82a4-edadeeda8c41",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import os\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import multivariate_normal\n",
    "import corner\n",
    "plt.rcParams.update({'font.size': 20})\n",
    "plt.rc('font',**{'family':'serif'})\n",
    "plt.rc('text', usetex=True)\n",
    "import pandas as pd\n",
    "\n",
    "# Import floZ training and pre-processing functions\n",
    "from floZ_LossCycle import Trainer\n",
    "from floZ_LossCycle.train import _pre_whitening\n",
    "\n",
    "# Import functions to generate and sample from likelihood, and calculate evidence.\n",
    "from floZ_LossCycle.utils_SampleGeneration import *\n",
    "from floZ_LossCycle.utils_Zestimation import symmetrize\n",
    "\n",
    "# Setting seed, if needed reproducability. Default = None for random seeds.\n",
    "MCMC_seed = None\n",
    "# MCMC_seed = 1\n",
    "np.random.seed(MCMC_seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8cd75491-c424-4773-ae71-452212f67902",
   "metadata": {},
   "source": [
    "## Initialize likelihood label, sample dimension and number of samples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "984a4f77-f27e-4214-ae9c-b5545b18915b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Un-comment the required likelihood label\n",
    "\n",
    "## Single Gaussian\n",
    "N = 1; likelihood_label = 'Likelihood-'+str(N)+'Gaussian'; ndims = 2; ninput = 10000\n",
    "\n",
    "## Gaussian Mixture\n",
    "# N = 5; likelihood_label = 'Likelihood-'+str(N)+'Gaussian'; ndims = 2; ninput = 10000\n",
    "\n",
    "## Rosenbrock\n",
    "# likelihood_label = 'Likelihood-Rosenbrock'; ndims = 2; ninput = 10000\n",
    "\n",
    "## Exponential\n",
    "# likelihood_label = 'Likelihood-Exponential'; ndims = 2; ninput = 10000"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6b52a15-aa30-444a-a2c1-3bc948df76f1",
   "metadata": {},
   "source": [
    "## Generating reference samples - distribution parameters saved for reproducability"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "89e467bb-2787-4f56-a268-f826dcebc223",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "./Likelihood-1Gaussian_2d_ninput[10000]\n",
      "Mean  1\n",
      "[23, 35]\n",
      "Covariance Matrix  1\n",
      "[[299, 31], [31, 284]]\n",
      "Shapes of means (1, 2), and covariances (1, 2, 2)\n",
      "\n",
      "\n",
      "lnPRIOR =  [-9.82652066 -9.82652066 -9.82652066 ... -9.82652066 -9.82652066\n",
      " -9.82652066]\n",
      "ln Evidence  -9.826553052179312\n",
      "\n",
      "\n",
      "Likelihood =  [1.37438541e-04 1.86588551e-04 2.71807722e-04 ... 1.88211741e-05\n",
      " 2.35217815e-04 6.16923777e-05]\n",
      "ln Z =  -9.826553052179312\n",
      "Limits  [array([-38.80465543, -33.10972861]), array([ 96.20158917, 104.05733573])]\n"
     ]
    }
   ],
   "source": [
    "if likelihood_label != 'custom':\n",
    "    floZpath_ref = f'./'+likelihood_label+'_{}d_ninput[10000]'.format(ndims)\n",
    "    \n",
    "    get_new_distribution = True\n",
    "    if os.path.exists(floZpath_ref):\n",
    "        get_new_distribution = False\n",
    "    else:\n",
    "        os.makedirs(floZpath_ref)\n",
    "\n",
    "    print(floZpath_ref)\n",
    "\n",
    "    if get_new_distribution:\n",
    "        if(likelihood_label == 'Likelihood-Rosenbrock'):\n",
    "            init_Rosenbrock(ndims = ndims, ninput=10000, floZpath_ref=floZpath_ref)\n",
    "        elif(likelihood_label == 'Likelihood-Exponential'):\n",
    "            if ndims ==2:\n",
    "                init_Exponential(ndims = ndims, ninput=10000, floZpath_ref=floZpath_ref, lambdas = [0.009057, 0.005257])\n",
    "            else:\n",
    "                init_Exponential(ndims = ndims, ninput=10000, floZpath_ref=floZpath_ref)\n",
    "        elif(len(likelihood_label.split('Gaussian'))>1):\n",
    "            if(ndims ==2 and N==1):\n",
    "                init_NGaussian(N, ndims = ndims, ninput=10000, floZpath_ref=floZpath_ref,\n",
    "                          means = [[23, 35]], covs = [[[299, 31],[31, 284]]])\n",
    "            elif(ndims ==2 and N==5):\n",
    "                init_NGaussian(N, ndims = ndims, ninput=10000, floZpath_ref=floZpath_ref,\n",
    "                              means = [[39, 19], [30, 38], [18, 12], [46, 44], [28, 28]], \n",
    "                               covs = [[[29, 8], [8,118]], [[250,15],[15,171]], [[152,4],[4, 32]], [[173, 12],[12, 107]], [[198, 17],[17, 468]]])  \n",
    "            else:\n",
    "                init_NGaussian(N, ndims = ndims, ninput=10000, floZpath_ref=floZpath_ref)\n",
    "                "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8fc2d4c1-3461-47c5-8d04-83d90b77d3a5",
   "metadata": {},
   "source": [
    "## Loading pre-existing samples, if any"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7674a50f-3b0a-4448-8329-20120832e7e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "./Likelihood-1Gaussian_2d_ninput[10000]\n"
     ]
    }
   ],
   "source": [
    "floZpath = f'./'+likelihood_label+'_{}d_ninput[{}]'.format(ndims, ninput)\n",
    "print(floZpath)\n",
    "\n",
    "pre_existing = os.path.exists(os.path.join(floZpath, 'loglikelihood_logprior_samples.txt'))      \n",
    "\n",
    "if pre_existing:\n",
    "    lnZ_GT = np.loadtxt(os.path.join(floZpath, 'lnZ_GT.txt'))\n",
    "    s = np.loadtxt(os.path.join(floZpath, 'loglikelihood_logprior_samples.txt'))\n",
    "    log_likelihood = s[:,0]\n",
    "    log_prior = s[:,1]\n",
    "    log_p = log_likelihood + log_prior\n",
    "    samples = s[:,2:]\n",
    "\n",
    "    log_prior = log_prior[log_p!=-np.inf]\n",
    "    samples = samples[log_p!=-np.inf]\n",
    "    log_likelihood = log_likelihood[log_p!=-np.inf]\n",
    "    log_p = log_p[log_p!=-np.inf]\n",
    "    meanmean = np.mean(samples, axis = 0)\n",
    "\n",
    "    if ndims < 10:\n",
    "        # Plot the corner plot\n",
    "        figure = corner.corner(np.transpose(samples.T), labels=[\"Var\"+str(index) for index in range(len(samples.T))], show_titles=True, truths=meanmean)\n",
    "        figure.suptitle(\"Current Samples\", y=1.02, fontsize=16)\n",
    "        plt.show()\n",
    "    \n",
    "    \n",
    "    plt.figure()\n",
    "    plt.title('log_p histogram')\n",
    "    print(np.min(log_p), np.max(log_p))\n",
    "    bins = np.linspace(np.min(log_p), np.max(log_p), 100)\n",
    "    plt.hist(log_p[log_p != -np.inf], bins = bins, histtype = 'step', label = 'Final samples: {}'.format(len(log_p)))\n",
    "    plt.yscale('log')\n",
    "    plt.xlabel('log_p')\n",
    "    plt.ylabel('Frequency')\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5207be73-9549-4fa3-aed6-5925b0341a43",
   "metadata": {},
   "source": [
    "## If it doesnt already exist, generate samples from the reference, recalculating log(likelihood x prior)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "287fd7af-ffed-4e2a-8042-2fb63796e310",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "Importing samples, means, covs from  ./Likelihood-1Gaussian_2d_ninput[10000]\n",
      "Imported ln Z =  -9.826553052179312\n",
      "Shape of imported means (1, 2) and covs (1, 2, 2)\n",
      "Imported Limits [[-38.80465543 -33.10972861]\n",
      " [ 96.20158917 104.05733573]]\n",
      "Oversampled sample size  (50000, 2)\n",
      "Boxed sample size  (49984, 2)\n",
      "Downsampled sample size (10000, 2)\n",
      "\n",
      "\n",
      "lnPRIOR =  [-9.82652066 -9.82652066 -9.82652066 ... -9.82652066 -9.82652066\n",
      " -9.82652066]\n",
      "ln Evidence  -9.826553052179312\n",
      "\n",
      "\n",
      "Likelihood =  [4.44517010e-04 8.85173264e-05 1.90787631e-04 ... 5.03955079e-04\n",
      " 4.54460212e-04 3.40145698e-04]\n",
      "Recalculated ln Z =  -9.826553052179312\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 550x550 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 550x550 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-26.12503399917459 -17.333460370863442\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "if not pre_existing:\n",
    "    ###################### Rosenbrock ###########################\n",
    "    \n",
    "    if(likelihood_label == 'Likelihood-Rosenbrock'):\n",
    "        print('\\n\\nImporting samples from ', floZpath_ref)\n",
    "        s = np.loadtxt(os.path.join(floZpath_ref, 'samples_logp.txt'))\n",
    "        samples_ref = s[:, :-1]\n",
    "        log_p = s[:, -1]\n",
    "        Z = np.loadtxt(os.path.join(floZpath_ref, 'Z.txt'))\n",
    "        limits = np.loadtxt(os.path.join(floZpath_ref, 'limits.txt'))\n",
    "        min_theta = limits[0,:]\n",
    "        max_theta = limits[1,:]\n",
    "        print('Imported ln Z = ', np.log(Z))\n",
    "        \n",
    "        if ndims <= 16:\n",
    "            figure = corner.corner(np.transpose(samples_ref.T), labels=[\"Var\"+str(index) for index in range(len(samples_ref.T))], show_titles=True)\n",
    "            figure.suptitle(\"Samples from imported distribution\", y=1.02, fontsize=16)\n",
    "        \n",
    "        # Generate MCMC samples with loaded prior bounds\n",
    "        initial_state = np.zeros(ndims)\n",
    "        samples = metropolis_hastings(6*ninput, initial_state)\n",
    "        print('Oversampled sample size ', samples.shape)\n",
    "    \n",
    "        # Discard samples outside prior limits.\n",
    "        for dim in range(ndims):\n",
    "            samples[np.where((samples[:,dim]>max_theta[dim]) | (samples[:,dim]<min_theta[dim])),:]*=np.nan\n",
    "        samples=samples[~np.isnan(samples[:,0]), :]\n",
    "        samples = np.array(samples)\n",
    "        print('Boxed sample size ', samples.shape)\n",
    "    \n",
    "        # Downsampling to ninputs samples.\n",
    "        samples=samples[random.sample(range(len(samples)), ninput),:]\n",
    "        samples = np.array(samples)\n",
    "        print('Downsampled sample size', samples.shape)\n",
    "    \n",
    "        log_p, Z = calc_Rosenbrock_log_p_Z(samples, min_theta, max_theta, ndims)\n",
    "        print('Recalculated log Z ', np.log(Z))\n",
    "        lnZ_GT = np.log(Z)\n",
    "        \n",
    "        save_Rosenbrock(floZpath, samples, log_p, limits, Z)\n",
    "        meanmean = np.mean(samples, axis = 0)\n",
    "    \n",
    "    ###################### Exponential ###########################\n",
    "    \n",
    "    elif(likelihood_label == 'Likelihood-Exponential'):\n",
    "        print('\\n\\nImporting samples and lambda from ', floZpath_ref)\n",
    "        s = np.loadtxt(os.path.join(floZpath_ref, 'samples_logp.txt'))\n",
    "        samples_ref = s[:, :-1]\n",
    "        log_p = s[:,-1]\n",
    "        Z = np.loadtxt(os.path.join(floZpath_ref, 'Z.txt'))\n",
    "        limits = np.loadtxt(os.path.join(floZpath_ref, 'limits.txt'))\n",
    "        lambdas = np.loadtxt(os.path.join(floZpath_ref, 'lambdas.txt'))\n",
    "    \n",
    "        print('\\nLambdas: \\n', lambdas)\n",
    "        min_theta = limits[0,:]\n",
    "        max_theta = limits[1,:]\n",
    "        print('Imported ln Z = ', np.log(Z))\n",
    "        \n",
    "        if ndims <= 16:\n",
    "            figure = corner.corner(np.transpose(samples_ref.T), labels=[\"Var\"+str(index) for index in range(len(samples_ref.T))], show_titles=True)\n",
    "            figure.suptitle(\"Samples from imported distribution\", y=1.02, fontsize=16)\n",
    "        \n",
    "        # Generate MCMC samples with loaded likelihood and prior bounds\n",
    "        samples = generate_n_dimensional_exponential(N=ndims, lambdas=lambdas, size=5*ninput)\n",
    "        print('Oversampled sample size ', samples.shape)\n",
    "        \n",
    "        # Discard samples outside prior limits.\n",
    "        for dim in range(ndims):\n",
    "            samples[np.where((samples[:,dim]>max_theta[dim]) | (samples[:,dim]<min_theta[dim])),:]*=np.nan\n",
    "        samples=samples[~np.isnan(samples[:,0]), :]\n",
    "        samples = np.array(samples)\n",
    "        print('Boxed sample size ', samples.shape)\n",
    "    \n",
    "        # Downsampling to ninputs samples.\n",
    "        if(len(samples)>ninput):\n",
    "            samples=samples[:ninput,:]\n",
    "            samples = np.array(samples)\n",
    "        print('Downsampled sample size', samples.shape)\n",
    "    \n",
    "        log_p, Z = calc_Exponential_log_p_Z(samples, min_theta, max_theta, lambdas)\n",
    "        print('Recalculated log Z ', np.log(Z))\n",
    "        lnZ_GT = np.log(Z)\n",
    "    \n",
    "        save_Exponential(floZpath, samples, log_p, limits, Z, lambdas)\n",
    "        \n",
    "        meanmean = np.mean(samples, axis = 0)\n",
    "\n",
    "    ####################### NGaussian ###########################\n",
    "    \n",
    "    elif(len(likelihood_label.split('Gaussian'))>1):\n",
    "        print('\\n\\nImporting samples, means, covs from ', floZpath_ref)\n",
    "        s = np.loadtxt(os.path.join(floZpath_ref, 'samples_logp.txt'))\n",
    "        samples_ref = s[:, :-1]\n",
    "        lnZ_GT = np.loadtxt(os.path.join(floZpath_ref, 'ln_Z.txt'))\n",
    "        print('Imported ln Z = ', lnZ_GT)\n",
    "        means = np.load(os.path.join(floZpath_ref, 'means.npy'))\n",
    "        covs = np.load(os.path.join(floZpath_ref, 'covs.npy'))\n",
    "        print('Shape of imported means {} and covs {}'.format(means.shape, covs.shape))\n",
    "        limits = np.loadtxt(os.path.join(floZpath_ref, 'limits.txt'))\n",
    "        print('Imported Limits', limits)\n",
    "        min_theta = limits[0,:]\n",
    "        max_theta = limits[1,:]\n",
    "    \n",
    "        figure = corner.corner(np.transpose(samples_ref.T), labels=[\"Var\"+str(index) for index in range(len(samples_ref.T))], show_titles=True)\n",
    "        figure.suptitle(\"Samples from imported distribution\", y=1.02, fontsize=16)\n",
    "    \n",
    "        samples = []\n",
    "        vars = []\n",
    "        meanmean = np.zeros(ndims)\n",
    "        # Sample from loaded likelihood.\n",
    "        for idx in range(N):\n",
    "            meanmean += means[idx]\n",
    "            samples += list(np.random.multivariate_normal(means[idx], covs[idx], 5*ninput//N))\n",
    "            vars += [multivariate_normal(mean=means[idx], cov=covs[idx])]\n",
    "        meanmean /= N\n",
    "        samples = np.array(samples)\n",
    "        print('Oversampled sample size ', samples.shape)\n",
    "    \n",
    "        # Discard samples outside prior limits.\n",
    "        for dim in range(ndims):\n",
    "            samples[np.where((samples[:,dim]>max_theta[dim]) | (samples[:,dim]<min_theta[dim])),:]*=np.nan\n",
    "        samples=samples[~np.isnan(samples[:,0]), :]\n",
    "        samples = np.array(samples)\n",
    "        print('Boxed sample size ', samples.shape)\n",
    "    \n",
    "        # Downsampling to ninputs samples.\n",
    "        samples=samples[random.sample(range(len(samples)), ninput),:]\n",
    "        samples = np.array(samples)\n",
    "        print('Downsampled sample size', samples.shape)\n",
    "    \n",
    "        log_p, lnZ_GT = calc_NGaussian_log_p_lnZ(N, samples, min_theta, max_theta, vars)\n",
    "        print('Recalculated ln Z = ', lnZ_GT)\n",
    "        \n",
    "        save_NGaussian(floZpath, samples, log_p, limits, lnZ_GT, means, covs)\n",
    "        \n",
    "    \n",
    "    if ndims < 10:\n",
    "        # Plot the corner plot\n",
    "        figure = corner.corner(np.transpose(samples.T), labels=[\"Var\"+str(index) for index in range(len(samples.T))], show_titles=True, truths = meanmean)\n",
    "        figure.suptitle(\"Current Samples\", y=1.02, fontsize=16)\n",
    "        plt.show()\n",
    "    \n",
    "    #sc = plt.scatter(samples.T[0],samples.T[1], c=log_p, cmap='viridis', s=10, alpha=0.7)\n",
    "    #cbar = plt.colorbar(sc, label='log (likelihood x prior)')\n",
    "    \n",
    "    plt.figure()\n",
    "    plt.title('log_p histogram')\n",
    "    print(np.min(log_p[log_p!=-np.inf]), np.max(log_p))\n",
    "    bins = np.linspace(np.min(log_p[log_p!=-np.inf]), np.max(log_p), 100)\n",
    "    plt.hist(log_p, bins = bins, histtype = 'step', label = 'Final samples: {}'.format(len(log_p)))\n",
    "    plt.yscale('log')\n",
    "    plt.xlabel('log_p')\n",
    "    plt.ylabel('Frequency')\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "\n",
    "else:\n",
    "    print('Distribution samples already exist!')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cdcdb3fe-9796-441f-8609-49f497aee233",
   "metadata": {},
   "source": [
    "## If not pre-existing, save Z, log_prior, log_likelihood, samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "59773edc-1d05-47cf-a25c-5bafcb14b4eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "if not pre_existing:\n",
    "    # Log likelihood for kNN calculation\n",
    "    if(len(likelihood_label.split('pyring'))==1):\n",
    "        prior = get_priors(samples, min_theta, max_theta)\n",
    "        if prior[0] != 0.:\n",
    "            log_prior = np.log(prior)\n",
    "        else:\n",
    "            log_prior = get_log_priors(samples, min_theta, max_theta)\n",
    "        log_prior = log_prior\n",
    "        log_likelihood = log_p-log_prior\n",
    "    [log_p!=-np.inf]\n",
    "    np.savetxt(os.path.join(floZpath, 'loglikelihood_logprior_samples.txt'), np.hstack((log_likelihood[:, np.newaxis], log_prior[:, np.newaxis], samples)))\n",
    "    np.savetxt(os.path.join(floZpath, 'lnZ_GT.txt'), [lnZ_GT])"
   ]
  }
 ],
 "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": "3.9.19"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}