expr_eval_cmp.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "69602765",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import numexpr as ne\n",
    "import iarray as ia\n",
    "\n",
    "p1 = np.random.random((100,100,1000)).astype('float32')\n",
    "p2 = np.random.random((100,100,1000)).astype('float32')\n",
    "p3 = np.random.random((100,100,1000)).astype('float32')\n",
    "\n",
    "p1_ia = ia.numpy2iarray(p1)\n",
    "p2_ia = ia.numpy2iarray(p2)\n",
    "p3_ia = ia.numpy2iarray(p3)\n",
    "\n",
    "expr1 = \"(p1 + p2 + p3) / 3\"\n",
    "expr2 = \"2.71828**p1 / p2 * log(p3)\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c9871f75",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 132 ms, sys: 68 ms, total: 200 ms\n",
      "Wall time: 51.4 ms\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.49992874"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "res = ne.evaluate(expr1, local_dict={'p1': p1, 'p2': p2, 'p3': p3})\n",
    "res.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3cf6663b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 229 ms, sys: 20.2 ms, total: 249 ms\n",
      "Wall time: 267 ms\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.4999297"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "res = ia.expr_from_string(expr1, {'p1': p1_ia, 'p2': p2_ia, 'p3': p3_ia}).eval()\n",
    "ia.mean(res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "07b58387",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.49 s, sys: 202 ms, total: 1.69 s\n",
      "Wall time: 330 ms\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "-48.931897281492255"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "res = ne.evaluate(expr2, local_dict={'p1': p1, 'p2': p2, 'p3': p3})\n",
    "res.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "591e5c63",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "\n",
    "res = ia.expr_from_string(expr2, {'p1': p1_ia, 'p2': p2_ia, 'p3': p3_ia}).eval()\n",
    "ia.mean(res)\n",
    "\n",
    "# Makes my kernel die"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "70a65d24",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}