{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import mitsuba as mi \n",
    "mi.set_variant(\"cuda_ad_rgb\")\n",
    "from PDE3D import PATH \n",
    "from PDE3D.BoundaryShape import * \n",
    "import os \n",
    "from matplotlib import gridspec\n",
    "import matplotlib\n",
    "import matplotlib.ticker as ticker\n",
    "from PDE3D.utils import *\n",
    "\n",
    "matplotlib.rcParams.update(MPL_STYLE)\n",
    "frame_color = \"white\"\n",
    "col_width = COLUMN_WIDTH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "jit_flush_malloc_cache(): Dr.Jit exhausted the available memory and had to flush its allocation cache to free up additional memory. This is an expensive operation and will have a negative effect on performance. You may want to change your computation so that it uses less memory. This warning will only be displayed once.\n"
     ]
    }
   ],
   "source": [
    "name = \"motorbike-engine\"\n",
    "folder_name = os.path.join(PATH, \"scenes\", name)\n",
    "xml_name = os.path.join(folder_name, \"scene.xml\")\n",
    "sdf_data = np.load(os.path.join(folder_name, \"sdf.npy\")) - 0.002\n",
    "boundary = SDF(sdf_data, scale = 12, mesh_scene_xml=xml_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "slice = Slice(offset =0.0, scale = 7, axis = \"z\")\n",
    "cam_origin = mi.ScalarPoint3f([7.8,7,20])\n",
    "scale_cam = 0.28\n",
    "cam_target = mi.ScalarPoint3f([0,-0.1,-0.2])\n",
    "cam_up = mi.ScalarPoint3f([0,1,0])\n",
    "cam_res = [1024, 1024]\n",
    "spp = 512\n",
    "\n",
    "cmap = \"coolwarm\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "res_primal = 4\n",
    "fd = 0.001\n",
    "res_tensor = 5\n",
    "path = os.path.join(PATH, \"output3D\", \"finite_differences\", \"variable\")\n",
    "source_prb =np.load(os.path.join(path, f\"source-resprimal{res_primal}\", f\"source{res_tensor}-prb.npy\"))\n",
    "source_fd =np.load(os.path.join(path, f\"source-resprimal{res_primal}\", f\"source{res_tensor}-fd{fd}.npy\"))\n",
    "screening_prb = np.load(os.path.join(path, f\"screening-resprimal{res_primal}\", f\"screening{res_tensor}-prb.npy\"))\n",
    "screening_fd = np.load(os.path.join(path, f\"screening-resprimal{res_primal}\", f\"screening{res_tensor}-fd{fd}.npy\"))\n",
    "diffusion_prb = np.load(os.path.join(path, f\"diffusion-resprimal{res_primal}\", f\"diffusion{res_tensor}-prb.npy\"))\n",
    "diffusion_fd = np.load(os.path.join(path, f\"diffusion-resprimal{res_primal}\", f\"diffusion{res_tensor}-fd{fd}.npy\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "bbox = boundary.bbox\n",
    "bbox_pad = (bbox.max - bbox.min) / 10\n",
    "bbox_coeff = mi.ScalarBoundingBox3f(bbox.min - bbox_pad, bbox.max + bbox_pad) \n",
    "\n",
    "f_prb = TextureCoefficient(\"prb\", bbox_coeff, source_prb, interpolation=\"nearest\")\n",
    "f_fd  = TextureCoefficient( \"fd\", bbox_coeff, source_fd, interpolation = \"nearest\")\n",
    "σ_prb = TextureCoefficient(\"prb\", bbox_coeff, screening_prb, interpolation = \"nearest\")\n",
    "σ_fd = TextureCoefficient(\"fd\", bbox_coeff, screening_fd, interpolation = \"nearest\")\n",
    "α_prb = TextureCoefficient(\"prb\", bbox_coeff, diffusion_prb, interpolation = \"nearest\")\n",
    "α_fd = TextureCoefficient(\"fd\", bbox_coeff, diffusion_fd, interpolation = \"nearest\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "f_prb_3D, f_prb_norm = boundary.visualize(colormap = cmap, cam_origin= cam_origin, spp = spp, image_res = cam_res, \n",
    "                                          scale_cam=scale_cam, cam_up = cam_up, slice = slice, cam_target = cam_target, \n",
    "                                          coeff= f_prb, sym_colorbar = True)\n",
    "f_fd_3D, f_fd_norm = boundary.visualize(colormap = cmap, cam_origin= cam_origin, spp = spp, image_res = cam_res, \n",
    "                                          scale_cam=scale_cam, cam_up = cam_up, slice = slice, cam_target = cam_target, \n",
    "                                          coeff= f_fd, sym_colorbar = True)\n",
    "\n",
    "σ_prb_3D, σ_prb_norm = boundary.visualize(colormap = cmap, cam_origin= cam_origin, spp = spp, image_res = cam_res, \n",
    "                                          scale_cam=scale_cam, cam_up = cam_up, slice = slice, cam_target = cam_target, \n",
    "                                          coeff= σ_prb, sym_colorbar = True)\n",
    "σ_fd_3D, σ_fd_norm = boundary.visualize(colormap = cmap, cam_origin= cam_origin, spp = spp, image_res = cam_res, \n",
    "                                          scale_cam=scale_cam, cam_up = cam_up, slice = slice, cam_target = cam_target, \n",
    "                                          coeff= σ_fd, sym_colorbar = True)\n",
    "\n",
    "α_prb_3D, α_prb_norm = boundary.visualize(colormap = cmap, cam_origin= cam_origin, spp = spp, image_res = cam_res, \n",
    "                                          scale_cam=scale_cam, cam_up = cam_up, slice = slice, cam_target = cam_target, \n",
    "                                          coeff= α_prb, sym_colorbar = True)\n",
    "α_fd_3D, α_fd_norm = boundary.visualize(colormap = cmap, cam_origin= cam_origin, spp = spp, image_res = cam_res, \n",
    "                                          scale_cam=scale_cam, cam_up = cam_up, slice = slice, cam_target = cam_target, \n",
    "                                          coeff= α_fd, sym_colorbar = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.7355973].\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.7355975].\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.7355976].\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.7355975].\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.7355975].\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.7355974].\n",
      "GPL Ghostscript 9.55.0: Can't embed the complete font LinBiolinumT as it is too large, embedding a subset.\n",
      "GPL Ghostscript 9.55.0: Can't embed the complete font LinLibertineT as it is too large, embedding a subset.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'/home/ekrem/repos/MC-PDE3D/figures/fd_3D/fd_3D.pdf'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 337.705x186.32 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "imsize = 30\n",
    "imsize_h = 24\n",
    "imsize_v = 27\n",
    "pre_offset = 1\n",
    "offset = 2\n",
    "post_offset = 10\n",
    "cbar_begin_offset = 2\n",
    "cbar_end_offset = 6\n",
    "o1 = 10\n",
    "o2 = 5\n",
    "total_offset = pre_offset + offset + post_offset\n",
    "\n",
    "lx = 2 * imsize_v + o1\n",
    "ly = (3 * imsize_h + 3 * total_offset + o2)\n",
    "scale = col_width / ly\n",
    "fig = plt.figure(figsize = (ly * scale, lx * scale))\n",
    "g = gridspec.GridSpec(lx, ly, figure = fig, wspace =0, hspace=0)\n",
    "\n",
    "ax = fig.add_subplot(g[:,:])\n",
    "disable_ticks(ax)\n",
    "plt.setp(ax.spines.values(), color=frame_color)\n",
    "f_prb_ax = fig.add_subplot(g[0 + o1 : imsize_v+ o1 , 0 + o2 : imsize_h+ o2 ])\n",
    "f_fd_ax = fig.add_subplot(g[imsize_v+ o1  : 2 * imsize_v+ o1 , 0+ o2  : imsize_h+ o2 ])\n",
    "\n",
    "σ_prb_ax = fig.add_subplot(g[0 + o1 : imsize_v+ o1 , imsize_h + total_offset + o2 : 2 * imsize_h + total_offset+ o2 ])\n",
    "σ_fd_ax = fig.add_subplot(g[imsize_v + o1 : 2 * imsize_v+ o1 , imsize_h + total_offset+ o2  : 2 * imsize_h + total_offset+ o2 ])\n",
    "\n",
    "α_prb_ax = fig.add_subplot(g[0 + o1 : imsize_v+ o1 , 2 * imsize_h + 2 * total_offset+ o2  : 3 * imsize_h + 2 * total_offset+ o2 ])\n",
    "α_fd_ax = fig.add_subplot(g[imsize_v + o1 : 2 * imsize_v+ o1 , 2 * imsize_h + 2 * total_offset+ o2  : 3 * imsize_h + 2 * total_offset+ o2 ])\n",
    "\n",
    "f_cbar_ax = fig.add_subplot(g[cbar_begin_offset + o1 :2 * imsize_v+ o1  - cbar_begin_offset, imsize_h + pre_offset+ o2  : imsize_h + pre_offset + offset+ o2 ])\n",
    "σ_cbar_ax = fig.add_subplot(g[cbar_begin_offset+ o1  :2 * imsize_v - cbar_begin_offset+ o1 , \n",
    "                              2 * imsize_h + total_offset + pre_offset+ o2  : 2 * imsize_h + total_offset + pre_offset + offset+ o2 ])\n",
    "α_cbar_ax = fig.add_subplot(g[cbar_begin_offset+ o1  :2 * imsize_v - cbar_begin_offset+ o1, \n",
    "                              3 * imsize_h + 2 * total_offset + pre_offset+ o2  : 3 * imsize_h + 2 * total_offset + pre_offset + offset+ o2 ])\n",
    "\n",
    "disable_border(f_prb_ax)\n",
    "disable_ticks(f_prb_ax)\n",
    "disable_border(σ_prb_ax)\n",
    "disable_ticks(σ_prb_ax)\n",
    "disable_border(α_prb_ax)\n",
    "disable_ticks(α_prb_ax)\n",
    "\n",
    "disable_border(f_fd_ax)\n",
    "disable_ticks(f_fd_ax)\n",
    "disable_border(σ_fd_ax)\n",
    "disable_ticks(σ_fd_ax)\n",
    "disable_border(α_fd_ax)\n",
    "disable_ticks(α_fd_ax)\n",
    "\n",
    "\n",
    "\n",
    "plot_image_3D(f_prb_3D[:int(imsize_v / imsize * cam_res[0]), -int(imsize_h / imsize * cam_res[1]):], f_prb_ax, cmap = cmap)\n",
    "plot_image_3D(f_fd_3D[:int(imsize_v / imsize * cam_res[0]), -int(imsize_h / imsize * cam_res[1]):], f_fd_ax, cmap = cmap)\n",
    "\n",
    "plot_image_3D(σ_prb_3D[:int(imsize_v / imsize * cam_res[0]), -int(imsize_h / imsize * cam_res[1]):],  σ_prb_ax, cmap = cmap)\n",
    "plot_image_3D(σ_fd_3D[:int(imsize_v / imsize * cam_res[0]), -int(imsize_h / imsize * cam_res[1]):], σ_fd_ax, cmap = cmap)\n",
    "\n",
    "plot_image_3D(α_prb_3D[:int(imsize_v / imsize * cam_res[0]), -int(imsize_h / imsize * cam_res[1]):], α_prb_ax, cmap = cmap)\n",
    "plot_image_3D(α_fd_3D[:int(imsize_v / imsize * cam_res[0]), -int(imsize_h / imsize * cam_res[1]):], α_fd_ax, cmap = cmap)\n",
    "\n",
    "num_ticks = 3\n",
    "f_cbar = plt.colorbar(mpl.cm.ScalarMappable(norm=f_fd_norm, cmap=cmap), cax=f_cbar_ax, orientation='vertical')\n",
    "tick_locator = ticker.MaxNLocator(nbins=num_ticks)\n",
    "f_cbar.locator = tick_locator\n",
    "f_cbar.formatter.set_powerlimits((0, 0))\n",
    "f_cbar.ax.yaxis.set_offset_position('left') \n",
    "f_cbar.update_ticks()\n",
    "\n",
    "σ_cbar = plt.colorbar(mpl.cm.ScalarMappable(norm=σ_fd_norm, cmap=cmap), cax=σ_cbar_ax, orientation='vertical')\n",
    "tick_locator = ticker.MaxNLocator(nbins=num_ticks)\n",
    "σ_cbar.locator = tick_locator\n",
    "σ_cbar.formatter.set_powerlimits((0, 0))\n",
    "σ_cbar.ax.yaxis.set_offset_position('left') \n",
    "σ_cbar.update_ticks()\n",
    "\n",
    "α_cbar = plt.colorbar(mpl.cm.ScalarMappable(norm=α_fd_norm, cmap=cmap), cax=α_cbar_ax, orientation='vertical')\n",
    "tick_locator = ticker.MaxNLocator(nbins=num_ticks)\n",
    "α_cbar.locator = tick_locator\n",
    "α_cbar.formatter.set_powerlimits((0, 0))\n",
    "α_cbar.ax.yaxis.set_offset_position('left') \n",
    "α_cbar.update_ticks()\n",
    "\n",
    "pad = 10\n",
    "f_prb_ax.set_title(r\"Source, $f$\", pad = pad, fontsize = DEFAULT_FONTSIZE_SMALL)\n",
    "σ_prb_ax.set_title(r\"Screening, $\\sigma$\", pad = pad, fontsize = DEFAULT_FONTSIZE_SMALL)\n",
    "α_prb_ax.set_title(r\"Diffusion, $\\alpha$\", pad = pad, fontsize = DEFAULT_FONTSIZE_SMALL)\n",
    "\n",
    "f_prb_ax.set_ylabel(\"Ours\")\n",
    "f_fd_ax.set_ylabel(\"FD\")\n",
    "\n",
    "plt.savefig(os.path.join(PATH, \"figures\", \"png\", \"fd_3D.png\"), bbox_inches='tight', pad_inches=0.04, dpi=1000)\n",
    "save_fig(\"fd_3D\", os.path.join(PATH, \"figures\"))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "invpde-3D",
   "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}