sbio_eval_heldout_final.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "spectacular-prevention",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "\n",
    "sns.set_style('white')\n",
    "sns.set_context('paper')\n",
    "sns.set()\n",
    "import random\n",
    "from sklearn.manifold import TSNE\n",
    "\n",
    "legend_size = 14"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "yellow-rough",
   "metadata": {},
   "outputs": [],
   "source": [
    "s = '22'\n",
    "has_samp= True\n",
    "models = [f'results_Mechanistic_h{s}']\n",
    "selected = models[0]\n",
    "data_path = f\"data/proc/{selected}/\"\n",
    "treatment = np.load(data_path + 'treatments.npy')\n",
    "devices = np.load(data_path + 'devices.npy')\n",
    "y = np.load(data_path + 'observations.npy')\n",
    "times = np.load(data_path + 'times.npy')\n",
    "\n",
    "\n",
    "mu_50_post = np.load(data_path + 'mu_50_post.npy')\n",
    "mu_25_post = np.load(data_path + 'mu_25_post.npy')\n",
    "mu_75_post = np.load(data_path + 'mu_75_post.npy')\n",
    "x_post = np.load(data_path + 'solution_xt_post.npy')\n",
    "x_post = np.swapaxes(x_post,1,2)\n",
    "\n",
    "mu_50_prior = np.load(data_path + 'mu_50_prior.npy')\n",
    "mu_25_prior = np.load(data_path + 'mu_25_prior.npy')\n",
    "mu_75_prior = np.load(data_path + 'mu_75_prior.npy')\n",
    "x_prior = np.load(data_path + 'solution_xt_prior.npy')\n",
    "x_prior = np.swapaxes(x_prior,1,2)\n",
    "\n",
    "if has_samp:\n",
    "    mu_50_post = np.load(data_path + 'mu_50_post_sample.npy')\n",
    "    mu_25_post = np.load(data_path + 'mu_25_post_sample.npy')\n",
    "    mu_75_post = np.load(data_path + 'mu_75_post_sample.npy')\n",
    "    \n",
    "    mu_50_prior = np.load(data_path + 'mu_50_prior_sample.npy')\n",
    "    mu_25_prior = np.load(data_path + 'mu_25_prior_sample.npy')\n",
    "    mu_75_prior = np.load(data_path + 'mu_75_prior_sample.npy')\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "accepting-outreach",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(48, 2)\n",
      "(48, 7)\n",
      "(48, 4, 100)\n",
      "(100,)\n",
      "(48, 4, 100, 200)\n",
      "(48, 8, 100)\n"
     ]
    }
   ],
   "source": [
    "print(treatment.shape)\n",
    "print(devices.shape)\n",
    "print(y.shape)\n",
    "print(times.shape)\n",
    "print(mu_75_prior.shape)\n",
    "print(x_post.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "spread-export",
   "metadata": {},
   "outputs": [],
   "source": [
    "# print(np.sum(y==0))\n",
    "# print(np.min(y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "metropolitan-arctic",
   "metadata": {},
   "outputs": [],
   "source": [
    "# print(treatment)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "miniature-stuff",
   "metadata": {},
   "source": [
    "## Plot scatter plots at end time point"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "former-belize",
   "metadata": {},
   "outputs": [],
   "source": [
    "is_post = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "prospective-designation",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "is_post:  True\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/65/0xvnnn591ql2mz158__j26tw0000gn/T/ipykernel_10526/1297760176.py:80: UserWarning: color is redundantly defined by the 'color' keyword argument and the fmt string \"kx\" (-> color='k'). The keyword argument will take precedence.\n",
      "  ax.semilogx(exp_treat, y_av_c[c, -1], 'k'+obs_mk, ms=ms, lw=1, color=edges[ci], zorder=ci+20)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "results_Mechanistic_h22 l1_error_av:  5.873522117024376 is_post:  True\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.clf()\n",
    "\n",
    "fs = 14\n",
    "ms = 5\n",
    "print(\"is_post: \", is_post)\n",
    "if is_post:\n",
    "    mu_50 = mu_50_post\n",
    "    mu_75 = mu_75_post\n",
    "    mu_25 = mu_25_post\n",
    "else:\n",
    "    mu_50 = mu_50_prior\n",
    "    mu_75 = mu_75_prior\n",
    "    mu_25 = mu_25_prior\n",
    "    \n",
    "    \n",
    "mu_50[mu_50 < 0] =0\n",
    "mu_25[mu_25 <0] = 0\n",
    "mu_75[mu_75 <0] = 0\n",
    "    \n",
    "#colors = sns.color_palette(\"Set2\", 8)\n",
    "obs_mk = 'x'\n",
    "pred_mk = 'o'\n",
    "colors = ['g','b', 'gray']\n",
    "edges = ['darkgreen','darkblue', 'darkgray']\n",
    "device_id = devices[0]\n",
    "\n",
    "l1_error = 0\n",
    "unique_treat = np.unique(treatment)\n",
    "unique_devices = np.unique(devices, axis=0)\n",
    "#print(\"unique_devices: \", unique_devices, len(unique_devices))\n",
    "cols = ['OD', 'RFP', 'YFP', 'CFP']\n",
    "conditions = ['C6', 'C12']\n",
    "fig, axs = plt.subplots(2, len(cols), sharex=True, sharey=True, figsize=(8, 4))\n",
    "\n",
    "\n",
    "for t_idx, treat in enumerate(unique_treat):\n",
    "    comp = np.equal(devices, device_id)\n",
    "    sel_device = np.sum(comp, 1) == devices.shape[1]\n",
    "    for  ci, cv in enumerate(conditions):\n",
    "        idx_C = np.logical_and(np.logical_and(treatment[:, ci] ==  treat, sel_device), treatment[:, ci] >0)\n",
    "    \n",
    "        for c in range(len(cols)):\n",
    "            if has_samp:\n",
    "                mu_50_av_c = np.mean(np.mean(mu_50[idx_C], 0), 2)\n",
    "                #print(\"mu_50_av_c: \", mu_50_av_c.shape)\n",
    "#                 mu_75_av_c = np.mean(np.mean(mu_75[idx_C], 0),2)\n",
    "#                 mu_25_av_c = np.mean(np.mean(mu_25[idx_C], 0), 2)\n",
    "                std = np.std(np.std(mu_50[idx_C], 0), 2)\n",
    "                mu_75_av_c = np.mean(np.mean(mu_50[idx_C], 0), 2) + 2* std\n",
    "                mu_25_av_c = np.mean(np.mean(mu_50[idx_C], 0), 2) - 2* std\n",
    "            else:\n",
    "                mu_50_av_c = np.mean(mu_50[idx_C], 0)\n",
    "                mu_75_av_c = np.mean(mu_75[idx_C], 0)\n",
    "                mu_25_av_c = np.mean(mu_25[idx_C], 0)\n",
    "            y_av_c = np.mean(y[idx_C], 0)\n",
    "           \n",
    "            \n",
    "            if np.sum(idx_C) > 0:\n",
    "                l1_error += np.sum(np.abs(y_av_c - mu_50_av_c))\n",
    "         \n",
    "            ax = axs[ci, c]\n",
    "            \n",
    "            exp_treat = np.exp(treat)-1\n",
    "        \n",
    "#             if c == 3 and t_idx == len(unique_treat)-1:\n",
    "#                 ax.errorbar(exp_treat, mu_50_av_c[c, -1], fmt=pred_mk, ms=ms, lw=1, mec=edges[0], \n",
    "#                     color=colors[ci], zorder=ci,  label='{} (model)'.format(conditions[ci]))\n",
    "#                 ax.semilogx(exp_treat, y_av_c[c, -1], 'k'+obs_mk, ms=ms, lw=1, color=edges[0], zorder=ci+20, \n",
    "#                          label='{} (data)'.format(conditions[ci]))\n",
    "            \n",
    "           \n",
    "            lower = [(mu_50_av_c[c, -1] - mu_25_av_c[c, -1])]\n",
    "            upper = [(mu_75_av_c[c, -1] - mu_50_av_c[c, -1])]\n",
    "            yerr = [lower, upper]\n",
    "            ax.errorbar(exp_treat, mu_50_av_c[c, -1], \n",
    "                        yerr=yerr, \n",
    "                        fmt=pred_mk, ms=ms, lw=1, mec=edges[ci], \n",
    "                   color=colors[ci], zorder=ci, solid_capstyle='projecting', capsize=2)\n",
    "            \n",
    "            ax.semilogx(exp_treat, y_av_c[c, -1], 'k'+obs_mk, ms=ms, lw=1, color=edges[ci], zorder=ci+20)\n",
    "            \n",
    "    \n",
    "            ax.set_ylim(-0.1,1.1)\n",
    "            ax.tick_params(axis='both', which='major', labelsize=fs)\n",
    "            ax.set_xticks(np.logspace(0,4,3))\n",
    "            ax.set_yticks([0, 0.5, 1])\n",
    "    \n",
    "\n",
    "for ax, col in zip(axs[0], cols):\n",
    "    ax.set_title(col, fontsize=fs)\n",
    "\n",
    "for ax, row in zip(axs[:, 0], conditions):\n",
    "    ax.set_ylabel(row, fontsize=fs)\n",
    "\n",
    "fig.text(0, 0.5, \"Normalized output\", ha=\"center\", va=\"center\", rotation=90, fontsize=fs,  fontweight='bold')\n",
    "fig.text(0.5, -0.01, \"C6/C12 (nM)\", ha=\"center\", va=\"bottom\", fontsize=fs,  fontweight='bold')\n",
    "#fig.legend(bbox_to_anchor=(0.95,0.89), fontsize=fs-2)\n",
    "fig.tight_layout()\n",
    "fig.tight_layout()\n",
    "\n",
    "l1_error_av = l1_error/(len(unique_treat)* len(cols) * devices.shape[1] )\n",
    "print(selected, \"l1_error_av: \", l1_error_av, \"is_post: \", is_post)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "downtown-boston",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "is_post:  True\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/65/0xvnnn591ql2mz158__j26tw0000gn/T/ipykernel_10526/318419159.py:83: UserWarning: color is redundantly defined by the 'color' keyword argument and the fmt string \"kx\" (-> color='k'). The keyword argument will take precedence.\n",
      "  ax.semilogx(exp_treat, y_av_c[r, -1], 'k'+obs_mk, ms=ms, lw=1, color=edges[ci], zorder=ci+20)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "results_Mechanistic_h22 l1_error_av:  11.747044234048753 is_post:  True\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x800 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.clf()\n",
    "\n",
    "fs = 14\n",
    "ms = 5\n",
    "print(\"is_post: \", is_post)\n",
    "if is_post:\n",
    "    mu_50 = mu_50_post\n",
    "    mu_75 = mu_75_post\n",
    "    mu_25 = mu_25_post\n",
    "else:\n",
    "    mu_50 = mu_50_prior\n",
    "    mu_75 = mu_75_prior\n",
    "    mu_25 = mu_25_prior\n",
    "    \n",
    "    \n",
    "mu_50[mu_50 < 0] =0\n",
    "mu_25[mu_25 <0] = 0\n",
    "mu_75[mu_75 <0] = 0\n",
    "    \n",
    "#colors = sns.color_palette(\"Set2\", 8)\n",
    "obs_mk = 'x'\n",
    "pred_mk = 'o'\n",
    "colors = ['g','b', 'gray']\n",
    "edges = ['darkgreen','darkblue', 'darkgray']\n",
    "device_id = devices[0]\n",
    "\n",
    "l1_error = 0\n",
    "unique_treat = np.unique(treatment)\n",
    "unique_devices = np.unique(devices, axis=0)\n",
    "#print(\"unique_devices: \", unique_devices, len(unique_devices))\n",
    "rows = ['OD', 'RFP', 'YFP', 'CFP']\n",
    "conditions = ['C6', 'C12']\n",
    "cols = conditions\n",
    "fig, axs = plt.subplots(len(rows), 2, sharex=True, sharey=True, figsize=(4, 8))\n",
    "\n",
    "\n",
    "for t_idx, treat in enumerate(unique_treat):\n",
    "    comp = np.equal(devices, device_id)\n",
    "    sel_device = np.sum(comp, 1) == devices.shape[1]\n",
    "    for  ci, cv in enumerate(conditions):\n",
    "        idx_C = np.logical_and(np.logical_and(treatment[:, ci] ==  treat, sel_device), treatment[:, ci] >0)\n",
    "        c = ci\n",
    "    \n",
    "        for r in range(len(rows)):\n",
    "            if has_samp:\n",
    "                mu_50_av_c = np.mean(np.mean(mu_50[idx_C], 0), 2)\n",
    "                #print(\"mu_50_av_c: \", mu_50_av_c.shape)\n",
    "#                 mu_75_av_c = np.mean(np.mean(mu_75[idx_C], 0),2)\n",
    "#                 mu_25_av_c = np.mean(np.mean(mu_25[idx_C], 0), 2)\n",
    "                std = np.std(np.std(mu_50[idx_C], 0), 2)\n",
    "                mu_75_av_c = np.mean(np.mean(mu_50[idx_C], 0), 2) + 2* std\n",
    "                mu_25_av_c = np.mean(np.mean(mu_50[idx_C], 0), 2) - 2* std\n",
    "            else:\n",
    "                mu_50_av_c = np.mean(mu_50[idx_C], 0)\n",
    "                mu_75_av_c = np.mean(mu_75[idx_C], 0)\n",
    "                mu_25_av_c = np.mean(mu_25[idx_C], 0)\n",
    "            y_av_c = np.mean(y[idx_C], 0)\n",
    "           \n",
    "            \n",
    "            if np.sum(idx_C) > 0:\n",
    "                l1_error += np.sum(np.abs(y_av_c - mu_50_av_c))\n",
    "         \n",
    "            #ax = axs[ci, c]\n",
    "            ax = axs[r, c]\n",
    "            \n",
    "            exp_treat = np.exp(treat)-1\n",
    "        \n",
    "#             if c == 3 and t_idx == len(unique_treat)-1:\n",
    "#                 ax.errorbar(exp_treat, mu_50_av_c[c, -1], fmt=pred_mk, ms=ms, lw=1, mec=edges[0], \n",
    "#                     color=colors[ci], zorder=ci,  label='{} (model)'.format(conditions[ci]))\n",
    "#                 ax.semilogx(exp_treat, y_av_c[c, -1], 'k'+obs_mk, ms=ms, lw=1, color=edges[0], zorder=ci+20, \n",
    "#                          label='{} (data)'.format(conditions[ci]))\n",
    "            \n",
    "           \n",
    "            lower = [(mu_50_av_c[r, -1] - mu_25_av_c[r, -1])]\n",
    "            upper = [(mu_75_av_c[r, -1] - mu_50_av_c[r, -1])]\n",
    "            yerr = [lower, upper]\n",
    "            ax.errorbar(exp_treat, mu_50_av_c[r, -1], \n",
    "                        yerr=yerr, \n",
    "                        fmt=pred_mk, ms=ms, lw=1, mec=edges[ci], \n",
    "                   color=colors[ci], zorder=ci, solid_capstyle='projecting', capsize=2)\n",
    "            \n",
    "            ax.semilogx(exp_treat, y_av_c[r, -1], 'k'+obs_mk, ms=ms, lw=1, color=edges[ci], zorder=ci+20)\n",
    "            \n",
    "    \n",
    "            ax.set_ylim(-0.1,1.1)\n",
    "            ax.tick_params(axis='both', which='major', labelsize=fs)\n",
    "            ax.set_xticks(np.logspace(0,4,3))\n",
    "            ax.set_yticks([0, 0.5, 1])\n",
    "    \n",
    "\n",
    "for ax, col in zip(axs[0], cols):\n",
    "    ax.set_title(col, fontsize=fs)\n",
    "\n",
    "for ax, row in zip(axs[:, 0], rows):\n",
    "    ax.set_ylabel(row, fontsize=fs)\n",
    "\n",
    "fig.text(0, 0.5, \"Normalized output\", ha=\"center\", va=\"center\", rotation=90, fontsize=fs,  fontweight='bold')\n",
    "fig.text(0.5, -0.01, \"C6/C12 (nM)\", ha=\"center\", va=\"bottom\", fontsize=fs,  fontweight='bold')\n",
    "#fig.legend(bbox_to_anchor=(0.95,0.89), fontsize=fs-2)\n",
    "fig.tight_layout()\n",
    "fig.tight_layout()\n",
    "\n",
    "l1_error_av = l1_error/(len(unique_treat)* len(cols) * devices.shape[1] )\n",
    "print(selected, \"l1_error_av: \", l1_error_av, \"is_post: \", is_post)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "prescribed-interstate",
   "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": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}