notebooks/data_discretization.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d8f878dc-e6aa-42e0-a29c-b042db7fef83",
   "metadata": {},
   "source": [
    "In this notebook I produce a discretized version of the quark gluon dataset.\n",
    "\n",
    "I scale it down to NxN and discretize each pixel to M values."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d2a0fd18-998a-49a9-a6e0-0dfcc52396b6",
   "metadata": {},
   "source": [
    "Parameters:\n",
    "    \n",
    "size: N (scale down to NxN)\n",
    "Nvals: M (M values per pixel)\n",
    "Nsamples: number of samples to write to the nwe dataset)\n",
    "file_name: filename of the output dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "63a8f96f-7fb8-40ff-9ffb-bd4433e7965d",
   "metadata": {},
   "outputs": [],
   "source": [
    "size = 4\n",
    "Nvals = 4\n",
    "Nsamples = 100000\n",
    "file_name = \"QG_train_normalized_N4_M4_bilinear_xl\"\n",
    "\n",
    "# which channel to use\n",
    "channel = 1\n",
    "# batch size when reading and writing\n",
    "batch_size = 200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "231cb03b-260a-42f8-aa6a-d8c1fadd1180",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import h5py\n",
    "import matplotlib\n",
    "from itertools import product"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "abd5b2a4-e2d6-42ca-b993-5cd7f93184e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "f = h5py.File(\"../data/quark-gluon_train-set_n793900.hdf5\",\"r\")\n",
    "x_train = f.get('X_jets')\n",
    "y_train = f.get('y')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eff0bd61-19d7-46ce-8ff1-45b182488235",
   "metadata": {},
   "source": [
    "Method to discretize an image to Nvals values per pixel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3646422e-e991-43c9-9ada-9ecf1cc0a873",
   "metadata": {},
   "outputs": [],
   "source": [
    "def discretize_img(img):\n",
    "    bins = np.linspace(0,1,Nvals+1)\n",
    "    vals = np.linspace(0,1,Nvals)\n",
    "    inds = np.digitize(img, bins)\n",
    "    return vals[inds-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "737b72b6-9bbc-4e56-864c-621e657e99b3",
   "metadata": {},
   "source": [
    "crap and resize an image with the given rescale method.\n",
    "crop fraction gives the percentage of the image to retain after the crop "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "edbd9983-0019-4756-9c8c-76fc48757453",
   "metadata": {},
   "outputs": [],
   "source": [
    "def crop_and_resize(x, channel, scale, crop_fraction=0.8, meth=\"bilinear\"):\n",
    "    cropped = tf.image.central_crop(x[:,:,:,channel].reshape(x.shape[0],125,125,1), crop_fraction)\n",
    "    return tf.image.resize(cropped, (scale,scale), method=meth).numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "89d1cb2f-4bdc-4a3b-8ec1-db72543c9cc8",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-10-04 12:50:25.609694: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory\n",
      "2023-10-04 12:50:25.609728: W tensorflow/compiler/xla/stream_executor/cuda/cuda_driver.cc:265] failed call to cuInit: UNKNOWN ERROR (303)\n",
      "2023-10-04 12:50:25.609755: I tensorflow/compiler/xla/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (tomskopfbahnhof): /proc/driver/nvidia/version does not exist\n",
      "2023-10-04 12:50:25.610113: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "batch  499 / 500\r"
     ]
    }
   ],
   "source": [
    "scale = size\n",
    "num_batches = Nsamples//batch_size\n",
    "\n",
    "events = num_batches*batch_size\n",
    "\n",
    "fnew = h5py.File(file_name, \"w\")\n",
    "\n",
    "dsetx = fnew.create_dataset(\"X\", (events,scale,scale), dtype='f')\n",
    "dsety = fnew.create_dataset(\"y\", (events,), dtype='i')\n",
    " \n",
    "for i in range(int(num_batches)):\n",
    "    y = y_train[i * batch_size: (i + 1) * batch_size]\n",
    "    x = x_train[i * batch_size: (i + 1) * batch_size]\n",
    "    \n",
    "    #div1 = np.max(x, axis=(1,2)).reshape((batch_size,1,1,3))\n",
    "    #div1[div1 == 0] = 1\n",
    "    #x = x / (div1+0.00001)\n",
    "    \n",
    "    x_train_small = crop_and_resize(x, channel, scale, crop_fraction=0.7)#, meth=\"lanczos5\")\n",
    "    \n",
    "    div1 = np.max(x_train_small, axis=(1,2)).reshape((batch_size,1,1,1))\n",
    "    div1[div1 == 0] = 1\n",
    "    x_train_small = x_train_small / (div1+0.00001)\n",
    "    \n",
    "    x_discr = discretize_img(x_train_small)\n",
    "    \n",
    "    dsety[i * batch_size: (i + 1) * batch_size] = y\n",
    "    dsetx[i * batch_size: (i + 1) * batch_size] = x_discr.reshape((x_discr.shape[0], scale, scale))\n",
    "    print(\"batch \",i,\"/\",num_batches, end=\"\\r\")\n",
    "    \n",
    "fnew.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "673d0c72-74b7-4280-b249-1acb85628a02",
   "metadata": {},
   "outputs": [],
   "source": [
    "fnew.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0592e063-3ee3-4fd0-9ec2-c4838173d025",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7fc92ae87f10>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ftest = h5py.File(file_name,\"r\")\n",
    "plt.imshow(np.mean(ftest['X'],axis=0),norm=matplotlib.colors.LogNorm())\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "32f2b2f4-a13d-4750-bc0e-081092e0fb58",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2, 3)\n",
    "for i,j in product(range(2),range(3)):\n",
    "    num = np.random.randint(0,ftest['X'].shape[0])\n",
    "    ax[i,j].imshow(ftest['X'][num])\n",
    "fig.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "13cab869-a973-40ff-9fdb-fc6c78b18df5",
   "metadata": {},
   "outputs": [],
   "source": [
    "ftest.close()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tfq",
   "language": "python",
   "name": "tfq"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}