Parquet direct dictionary decoding

benchmark.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pyarrow as pa\n",
    "import pyarrow.parquet as pq\n",
    "import pandas as pd\n",
    "from pandas.util.testing import rands\n",
    "        \n",
    "NUNIQUE = 1000\n",
    "STRING_SIZE = 50\n",
    "LENGTH = 10_000_000\n",
    "REPEATS = LENGTH // NUNIQUE\n",
    "\n",
    "data = [rands(STRING_SIZE) for i in range(NUNIQUE)] * REPEATS\n",
    "table = pa.table([data], names=['f0'])\n",
    "\n",
    "out_stream = pa.BufferOutputStream()\n",
    "pq.write_table(table, out_stream)\n",
    "contents = out_stream.getvalue()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1129939"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(contents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n"
     ]
    }
   ],
   "source": [
    "import gc\n",
    "class memory_use:\n",
    "    \n",
    "    def __init__(self):\n",
    "        self.start_use = pa.total_allocated_bytes()\n",
    "    \n",
    "    def __enter__(self):\n",
    "        return\n",
    "    \n",
    "    def __exit__(self, type, value, traceback):\n",
    "        gc.collect()\n",
    "        print(pa.total_allocated_bytes() - self.start_use)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "541250112\n"
     ]
    }
   ],
   "source": [
    "with memory_use():\n",
    "    memory_use_no_dict = pq.read_table(pa.BufferReader(contents))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "41304128\n"
     ]
    }
   ],
   "source": [
    "with memory_use():\n",
    "    memory_use_dict = pq.read_table(pa.BufferReader(contents), read_dictionary=['f0'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.79 s ± 7.86 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "%timeit memory_use_no_dict = pq.read_table(pa.BufferReader(contents))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "106 ms ± 11.2 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit memory_use_dict = pq.read_table(pa.BufferReader(contents), read_dictionary=['f0'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}