notebooks/Paper-Figure-5-posterior-DDM-with-collapsing-bounds.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "835b1b13",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pickle\n",
    "import scipy\n",
    "import time\n",
    "import torch\n",
    "\n",
    "from scipy.stats import binom, binom_test, wilcoxon\n",
    "from tensorflow import keras\n",
    "\n",
    "import sbibm\n",
    "\n",
    "from sbibm.tasks.ddm.utils import LinearCollapseDDM\n",
    "from sbibm.utils.io import get_tensor_from_csv, get_ndarray_from_csv\n",
    "from sbi.analysis import pairplot\n",
    "from utils import compile_df, huber_loss, mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8048d6d4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seed 223174\n"
     ]
    }
   ],
   "source": [
    "# Plotting settings\n",
    "plt.style.use(\"plotting_settings.mplstyle\")\n",
    "# Colorblind color palette\n",
    "colors = [\n",
    "    \"#377eb8\",\n",
    "    \"#ff7f00\",\n",
    "    \"#4daf4a\",\n",
    "    \"#f781bf\",\n",
    "    \"#a65628\",\n",
    "    \"#984ea3\",\n",
    "    \"#999999\",\n",
    "    \"#e41a1c\",\n",
    "    \"#dede00\",\n",
    "]\n",
    "\n",
    "# Fix seed for figures.\n",
    "seed = 223174  # torch.randint(1000000, (1,))\n",
    "torch.manual_seed(seed)\n",
    "print(f\"seed {seed}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2c8a5f4b",
   "metadata": {},
   "outputs": [],
   "source": [
    "task = sbibm.get_task(\"ddm\")\n",
    "prior = task.get_prior_dist()\n",
    "simulator = task.get_simulator(seed=seed)  # Passing the seed to Julia."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18dc195a",
   "metadata": {},
   "source": [
    "## Load inference results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "200aeea0",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"../data/results/linearcollapseDDM_100runs.p\", \"rb\") as fh:\n",
    "    # thos contains gt parameters\n",
    "    # xos the corresponding observed data\n",
    "    # samples the inferred posterior samples\n",
    "    thos, xos, samples, mm, prior, transforms = pickle.load(fh).values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "db2d6e29",
   "metadata": {},
   "outputs": [],
   "source": [
    "limits = [\n",
    "    [prior.base_dist.low[i], prior.base_dist.high[i]]\n",
    "    for i in range(len(prior.base_dist.low))\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6d60c758",
   "metadata": {},
   "outputs": [],
   "source": [
    "# limits= [[-2.5, 2.5], [0.5, 2.0], [.25, .75], [.05, .95], [-2, 0]]\n",
    "labels = [r\"$v$\", r\"$a$\", r\"$w$\", r\"$\\tau$\", r\"$\\gamma$\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f583e65",
   "metadata": {},
   "source": [
    "### Calculate SBC ranks under inferred posteriors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "29543259",
   "metadata": {},
   "outputs": [],
   "source": [
    "N, num_params = thos.shape\n",
    "L = 1000  # number of SBC samples from posterior\n",
    "\n",
    "ranks = np.zeros((N, num_params))\n",
    "\n",
    "for obsi in range(N):\n",
    "    # rank tho under samples\n",
    "    ranks[obsi, :] = (\n",
    "        samples[obsi][:L, :]\n",
    "        < thos[\n",
    "            obsi,\n",
    "        ]\n",
    "    ).sum(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03db8212",
   "metadata": {},
   "source": [
    "### Define simulator for predictive checks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "93bd5ad0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/janfb/.local/lib/python3.7/site-packages/julia/core.py:691: FutureWarning: Accessing `Julia().<name>` to obtain Julia objects is deprecated.  Use `from julia import Main; Main.<name>` or `jl = Julia(); jl.eval('<name>')`.\n",
      "  FutureWarning,\n"
     ]
    }
   ],
   "source": [
    "ddm = LinearCollapseDDM()\n",
    "\n",
    "# Generate data full ddm\n",
    "def simulator(theta):\n",
    "    v, a, w, ndt, gamma = theta.numpy().T\n",
    "    # using boundary separation a and offset w\n",
    "    # pass negative lower bound as required by DiffModels.\n",
    "    bl = -w * a\n",
    "    bu = (1 - w) * a\n",
    "\n",
    "    rts, choices = ddm.simulate(v, bu, -bl, gamma)\n",
    "    rts += ndt.reshape(-1, 1)\n",
    "    # encode zero choices as negative RTs.\n",
    "    rts[choices == 0] *= -1\n",
    "\n",
    "    return torch.tensor(rts, dtype=torch.float32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "c7fb832c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Select one index from the 100 posteriors for good visualization\n",
    "pii = 8\n",
    "\n",
    "posterior_samples = samples[pii]\n",
    "prior_samples = prior.sample((10000,))\n",
    "\n",
    "prior_predictive_samples = simulator(prior_samples)\n",
    "posterior_predictive_samples = simulator(posterior_samples)\n",
    "true_params_samples = simulator(thos[pii].repeat(posterior_samples.shape[0], 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e0f847cc",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/qode/sbi/sbi/analysis/plot.py:425: UserWarning: No contour levels were found within the data range.\n",
      "  levels=opts[\"contour_offdiag\"][\"levels\"],\n",
      "/home/janfb/anaconda3/envs/mnle/lib/python3.7/site-packages/ipykernel_launcher.py:121: UserWarning: This figure includes Axes that are not compatible with tight_layout, so results might be incorrect.\n",
      "findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans.\n",
      "findfont: Generic family 'sans-serif' not found because none of the following families were found: open-sans\n",
      "findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans.\n",
      "findfont: Generic family 'sans-serif' not found because none of the following families were found: open-sans\n",
      "findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans.\n",
      "findfont: Generic family 'sans-serif' not found because none of the following families were found: open-sans\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x648 with 27 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(18, 9))\n",
    "outer_grid = fig.add_gridspec(1, 2, wspace=0.25, hspace=0, width_ratios=[0.6, 0.4])\n",
    "# Plotting settings\n",
    "notch = True\n",
    "density = True\n",
    "nbins = 100\n",
    "histtype = \"step\"\n",
    "repeats = 50\n",
    "linewidth = 2\n",
    "\n",
    "# posterior samples\n",
    "num_plots = 5\n",
    "inner_grid = outer_grid[0, 0].subgridspec(num_plots, num_plots, wspace=0.25, hspace=0.2)\n",
    "ax1 = inner_grid.subplots()  # Create all subplots for the inner grid.\n",
    "pairplot(\n",
    "    samples[pii],\n",
    "    points=thos[pii],\n",
    "    limits=limits,\n",
    "    ticks=limits,\n",
    "    samples_colors=colors[2:3],\n",
    "    diag=\"kde\",\n",
    "    upper=\"contour\",\n",
    "    kde_offdiag=dict(bw_method=\"scott\", bins=50),\n",
    "    contour_offdiag=dict(levels=[0.95], percentile=True),\n",
    "    points_offdiag=dict(marker=\"+\", markersize=10),\n",
    "    points_colors=[\"k\"],\n",
    "    fig=fig,\n",
    "    axes=ax1,\n",
    "    labels=labels,\n",
    ")\n",
    "plt.sca(ax1[0, 0])\n",
    "plt.legend(\n",
    "    [\n",
    "        \"MNLE posterior\",\n",
    "        r\"ground-truth $\\theta$\",\n",
    "    ],\n",
    "    bbox_to_anchor=(-0.4, -4.2),\n",
    "    handlelength=0.6,\n",
    "    loc=2,\n",
    ")\n",
    "\n",
    "# posterior metrics\n",
    "inner_grid = outer_grid[0, 1].subgridspec(\n",
    "    2,\n",
    "    1,\n",
    "    wspace=0,\n",
    "    hspace=0.4,\n",
    ")\n",
    "ax2 = inner_grid.subplots()  # Create all subplots for the inner grid.\n",
    "\n",
    "## PREDICTIVES\n",
    "plt.sca(ax2[0])\n",
    "_, bins = np.histogram(prior_predictive_samples.numpy(), bins=nbins)\n",
    "plt.hist(\n",
    "    true_params_samples.numpy(),\n",
    "    bins=bins,\n",
    "    density=density,\n",
    "    histtype=histtype,\n",
    "    color=colors[0],\n",
    "    linewidth=linewidth,\n",
    ")\n",
    "plt.hist(\n",
    "    prior_predictive_samples.numpy(),\n",
    "    bins=nbins,\n",
    "    alpha=1,\n",
    "    density=density,\n",
    "    histtype=histtype,\n",
    "    color=colors[3],\n",
    "    linewidth=linewidth,\n",
    ")\n",
    "plt.hist(\n",
    "    posterior_predictive_samples.numpy(),\n",
    "    bins=bins,\n",
    "    alpha=1,\n",
    "    density=density,\n",
    "    histtype=histtype,\n",
    "    color=colors[2],\n",
    "    linewidth=linewidth,\n",
    ")\n",
    "ax2[0].spines[\"left\"].set_visible(False)\n",
    "plt.yticks([])\n",
    "plt.xlabel(\"$x$: reaction time [s]\")\n",
    "ticks = np.array([-2, -1, 0, 1, 2])\n",
    "plt.xticks(ticks, abs(ticks))\n",
    "plt.xlim([ticks.min(), ticks.max()])\n",
    "plt.legend(\n",
    "    [\"observations\", \"prior \\npredictive\", \"posterior \\npredictive\"],\n",
    "    loc=1,\n",
    "    handlelength=0.9,\n",
    "    bbox_to_anchor=(1.1, 1.0),\n",
    ")\n",
    "\n",
    "# Arrows\n",
    "plt.axvline(0, color=\"k\", lw=1)\n",
    "y = 4.4\n",
    "length = 0.15\n",
    "axis_setoff = -0.02\n",
    "plt.arrow(\n",
    "    0, y, length, 0.0, width=0.03, head_width=0.2, head_length=0.1, color=\"k\", alpha=0.5\n",
    ")\n",
    "plt.arrow(\n",
    "    0,\n",
    "    y,\n",
    "    -length,\n",
    "    0.0,\n",
    "    width=0.03,\n",
    "    head_width=0.2,\n",
    "    head_length=0.1,\n",
    "    color=\"k\",\n",
    "    alpha=0.5,\n",
    ")\n",
    "\n",
    "plt.text(0.11, 1.05 * y, s=\"up\")\n",
    "plt.text(-0.66, 1.05 * y, s=\"down\")\n",
    "plt.ylim([0, 1.05 * y])\n",
    "\n",
    "## SBC\n",
    "plt.sca(ax2[1])\n",
    "color_offset = 4\n",
    "for jj in range(num_params):\n",
    "    hist, *_ = np.histogram(ranks[:, jj], bins=nbins, density=False)\n",
    "    histcs = hist.cumsum()\n",
    "    plt.plot(\n",
    "        np.linspace(0, nbins, repeats * nbins),\n",
    "        np.repeat(histcs / histcs.max(), repeats),\n",
    "        label=labels[jj],\n",
    "        color=colors[jj + color_offset],\n",
    "        alpha=0.8,\n",
    "    )\n",
    "    plt.yticks(np.linspace(0, 1, 3))\n",
    "    plt.ylabel(\"empirical CDF\")\n",
    "    plt.legend(loc=4, handlelength=0.9)\n",
    "    plt.ylim(0, 1)\n",
    "    plt.xlim(0, nbins)\n",
    "    plt.xticks(np.linspace(0, nbins, 3))\n",
    "    plt.gca().spines[\"left\"].set_position((\"axes\", axis_setoff))\n",
    "    plt.gca().spines[\"bottom\"].set_position((\"axes\", axis_setoff))\n",
    "    plt.xlabel(\"posterior rank\")\n",
    "    # Construct uniform histogram.\n",
    "    if jj == 0:\n",
    "        hb = binom(N, p=1 / nbins).ppf(0.5) * np.ones(nbins)\n",
    "        hbb = hb.cumsum() / hb.sum()\n",
    "        # avoid last value being exactly 1\n",
    "        hbb[-1] -= 1e-9\n",
    "\n",
    "        lower = [binom(N, p=p).ppf(0.005) for p in hbb]\n",
    "        upper = [binom(N, p=p).ppf(0.995) for p in hbb]\n",
    "\n",
    "        # Plot grey area with expected ECDF.\n",
    "        plt.fill_between(\n",
    "            x=np.linspace(0, nbins, repeats * nbins),\n",
    "            y1=np.repeat(lower / np.max(lower), repeats),\n",
    "            y2=np.repeat(upper / np.max(lower), repeats),\n",
    "            color=\"grey\",\n",
    "            alpha=0.2,\n",
    "        )\n",
    "\n",
    "\n",
    "weight = \"bold\"\n",
    "fontsize = 20\n",
    "y1 = 0.9\n",
    "x1 = 0.1\n",
    "dx = 0.47\n",
    "fig.text(x1, y1, \"a\", fontsize=fontsize, fontweight=weight)\n",
    "fig.text(x1 + dx, y1, \"b\", fontsize=fontsize, fontweight=weight)\n",
    "fig.text(x1 + dx, y1 - 0.43, \"c\", fontsize=fontsize, fontweight=weight)\n",
    "\n",
    "plt.tight_layout()\n",
    "fig.savefig(\n",
    "    \"figures/Figure5-MNLE-LinearCollapseDDM-Posterior.png\", dpi=400, bbox_inches=\"tight\"\n",
    ")\n",
    "fig.savefig(\"figures/Figure 5.pdf\", dpi=400, bbox_inches=\"tight\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2afdd6af",
   "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": "3.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}