benchmarking_v2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import torch\n",
    "# torch.multiprocessing.set_start_method(\"spawn\")\n",
    "import numpy as np\n",
    "import cv2\n",
    "\n",
    "from torch.utils import data\n",
    "from torchvision import transforms\n",
    "\n",
    "import time\n",
    "import matplotlib.pyplot as plt\n",
    "import random\n",
    "import numbers\n",
    "import math\n",
    "from PIL import Image\n",
    "from tqdm import tqdm_notebook as tqdm\n",
    "import h5py\n",
    "import importlib\n",
    "import opencv_transforms as cv_transforms\n",
    "import warnings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_times_and_normalized(cv_times, pil_times, img_sizes,transform_name, should_save=False, savedir=None):\n",
    "    normalized_cv = cv_times /cv_times.mean(axis=0)\n",
    "    normalized_pil = pil_times / cv_times.mean(axis=0)\n",
    "    \n",
    "    sizes = [str(i) for i in img_sizes]\n",
    "    \n",
    "    fig, ax = plt.subplots(1,2,figsize=(12,6))\n",
    "    x = np.arange(0, cv_times.shape[1])\n",
    "    \n",
    "    ax[0].errorbar(x, cv_times.mean(axis=0), yerr=cv_times.std(axis=0), fmt='o-', label='OpenCV')\n",
    "    ax[0].errorbar(x, pil_times.mean(axis=0), yerr=pil_times.std(axis=0), fmt='o-', label='PIL')\n",
    "    ax[0].set_xticks(x)\n",
    "    ax[0].set_xticklabels(sizes, rotation=45)\n",
    "    ax[0].set_xlabel('Image sizes')\n",
    "    ax[0].set_ylabel('Execution time (ms)')\n",
    "    ax[0].legend()\n",
    "    \n",
    "    ax[1].errorbar(x, normalized_cv.mean(axis=0), yerr=normalized_cv.std(axis=0), \n",
    "                   fmt='o-', label='OpenCV')\n",
    "    ax[1].errorbar(x, normalized_pil.mean(axis=0), yerr=normalized_pil.std(axis=0), \n",
    "                   fmt='o-', label='PIL')\n",
    "    \n",
    "    y = normalized_pil.mean(axis=0)\n",
    "    [ax[1].text(x[i]+0.05, y[i]+0.05, '%0.2f'%(t)) for i,t in enumerate(y)]\n",
    "    ax[1].set_xticks(x)\n",
    "    ax[1].set_xticklabels(sizes, rotation=45)\n",
    "    ax[1].set_xlabel('Image sizes')\n",
    "    ax[1].set_ylabel('Execution time normalized to OpenCV')\n",
    "    \n",
    "    fig.suptitle('Transform: %s' %(transform_name), fontsize=16)\n",
    "    plt.tight_layout()\n",
    "    plt.subplots_adjust(top=0.92)\n",
    "    plt.show()\n",
    "    \n",
    "    if should_save and savedir is not None:\n",
    "        figname = 'benchmarking_%s.png' %(transform_name)\n",
    "        savename = os.path.join(savedir, figname)\n",
    "        fig.savefig(savename)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TESTING!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'opencv_transforms' from 'C:\\\\Users\\\\jpb\\\\Documents\\\\python\\\\dataloader_benchmarks\\\\opencv_transforms.py'>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importlib.reload(cv_transforms)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.style.use('ggplot')\n",
    "\n",
    "cityscapes_dir = r'E:\\DATA\\semantic_segmentation\\Cityscapes'\n",
    "fname = os.path.join(cityscapes_dir, 'train.h5')\n",
    "\n",
    "img_sizes = [(2048, 1024), (1024, 512), (512,256), (256, 128), (128, 64)]\n",
    "n_repeats = 100 # must be less than dataset size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "transform_pil = transforms.Compose([transforms.Resize(size=(512,256)),\n",
    "                                   transforms.RandomHorizontalFlip(p=0.5), \n",
    "                                   transforms.ColorJitter(brightness=0.5,\n",
    "                                                         contrast=0.25),\n",
    "                                   transforms.RandomRotation(degrees=10)])\n",
    "transform_cv = cv_transforms.Compose([cv_transforms.Resize(size=(512,256)),\n",
    "                                   cv_transforms.RandomHorizontalFlip(p=0.5), \n",
    "                                   cv_transforms.ColorJitter(brightness=0.5,\n",
    "                                                         contrast=0.25),\n",
    "                                   cv_transforms.RandomRotation(degrees=10)])\n",
    "\n",
    "transform_name = 'Resize flip brightness contrast rotate'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a6b61348ba024230bc5c59b89988d173",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=5), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "95a29776dad44e10bf035a5b4aee0bcf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d91e9d78fe2744b9ad6d91f15e849fcb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "04bc76c26c334e1fae683c90f908316f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "044dc8885bab4e6eacd0d6b599e18a30",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0e645f40881344d090dc70f9159849b2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "pil_times = np.zeros((n_repeats, len(img_sizes)), dtype=np.float64)\n",
    "cv_times = np.zeros((n_repeats, len(img_sizes)), dtype=np.float64)\n",
    "\n",
    "with h5py.File(fname, 'r') as f:\n",
    "    for i in tqdm(range(len(img_sizes))):\n",
    "        img_size = img_sizes[i]\n",
    "        resize_pil = transforms.Resize((img_size[0], img_size[1]))\n",
    "        resize_cv = cv_transforms.Resize((img_size[0], img_size[1]))\n",
    "        \n",
    "        # args = {'size': (img_size[0]//2, img_size[1]//2)}\n",
    "        # this_transform_pil = transform_pil(**args)\n",
    "        # this_transform_cv = transform_cv(**args)\n",
    "        this_transform_pil = transform_pil\n",
    "        this_transform_cv = transform_cv\n",
    "        for j in tqdm(range(n_repeats)):\n",
    "            \n",
    "            img = f['images'][j]\n",
    "            img_pil = Image.fromarray(img)\n",
    "            img = resize_cv(img)\n",
    "            img_pil = resize_pil(img_pil)\n",
    "            \n",
    "            t0 = time.clock()\n",
    "            out_pil = this_transform_pil(img_pil)\n",
    "            t2 = (time.clock()-t0)*1000 # milliseconds\n",
    "            pil_times[j,i]= t2\n",
    "            \n",
    "            t0 = time.clock()\n",
    "            out_cv = this_transform_cv(img)\n",
    "            t2 = (time.clock()-t0)*1000 # milliseconds\n",
    "            cv_times[j,i]= t2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "savedir = r'D:\\Analysis\\cv_transforms'\n",
    "plot_times_and_normalized(cv_times, pil_times, img_sizes, transform_name, \n",
    "                         should_save=True, savedir=savedir)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}