1 - Data Analysis Speedup.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4adb8c11-7532-4628-8bb5-08e78efcfad8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "from pathlib import Path\n",
    "\n",
    "import numba\n",
    "import numpy as np\n",
    "\n",
    "from utils.measure_and_compare import measure_and_compare, reset_measure_and_compare\n",
    "from utils.preparation import get_segmentation, plot\n",
    "\n",
    "reset_measure_and_compare()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3693b372-9156-42f3-ba5b-ec47b211ec67",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data(path):\n",
    "    segments = []\n",
    "    counts = []\n",
    "    centers = []\n",
    "    for json_path in sorted(path.glob(\"*.json\"), key=lambda p: int(p.stem)):\n",
    "        with open(json_path) as f:\n",
    "            data = json.load(f)\n",
    "            segments.append(data[\"segments\"])\n",
    "            counts.append(data[\"counts\"])\n",
    "            centers.append(data[\"centers\"])\n",
    "    return segments, counts, centers\n",
    "\n",
    "\n",
    "def combine_stats(chunked_segments, chunked_counts, chunked_centers):\n",
    "    counts_per_segment = {}\n",
    "    centers_per_segement = {}\n",
    "    for segments_i, counts_i, centers_i in zip(\n",
    "        chunked_segments, chunked_counts, chunked_centers\n",
    "    ):\n",
    "        for segment, count, center in zip(segments_i, counts_i, centers_i):\n",
    "            counts_per_segment.setdefault(segment, []).append(count)\n",
    "            centers_per_segement.setdefault(segment, []).append(center)\n",
    "\n",
    "    result_segments = []\n",
    "    result_counts = []\n",
    "    result_centers = []\n",
    "    for segment in counts_per_segment.keys():\n",
    "        center_x = 0\n",
    "        center_y = 0\n",
    "        count_sum = 0\n",
    "        for count, center in zip(\n",
    "            counts_per_segment[segment], centers_per_segement[segment]\n",
    "        ):\n",
    "            center_x += center[0] * count\n",
    "            center_y += center[1] * count\n",
    "            count_sum += count\n",
    "\n",
    "        result_segments.append(segment)\n",
    "        result_counts.append(count_sum)\n",
    "        result_centers.append((center_x / count_sum, center_y / count_sum))\n",
    "    return result_segments, result_counts, result_centers\n",
    "\n",
    "\n",
    "def load_and_combine_stats(path):\n",
    "    chunked_stats = load_data(path)\n",
    "    return combine_stats(*chunked_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e82f306f-e521-49d0-a2c9-40e84bdab26b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats(Path(\"data/small\"))\n",
    "\n",
    "segmentation = get_segmentation()\n",
    "plot(segmentation, combined_stats)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e9e688ae-7102-44c5-8080-95a5b2722699",
   "metadata": {},
   "source": [
    "# Profiling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "aedaa01b-32a6-4cc5-a896-9f02c7b7f337",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15.4 s ± 593 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit\n",
    "load_and_combine_stats(Path(\"data/large\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c10d8d18-465f-443e-9492-8f5f6e2ebbcb",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Clock type: CPU\n",
      "Ordered by: totaltime, desc\n",
      "\n",
      "name                                  ncall  tsub      ttot      tavg      \n",
      "..:3483 ZMQInteractiveShell.run_code  1      0.000017  59.39702  59.39702\n",
      "..nel_15257/2801445275.py:6 <module>  1      0.413383  59.39698  59.39698\n",
      "..65336.py:44 load_and_combine_stats  1      0.053888  58.98350  58.98350\n",
      "..5257/339865336.py:14 combine_stats  1      21.66024  39.29477  39.29477\n",
      "..nel_15257/339865336.py:1 load_data  1      0.602757  19.63484  19.63484\n",
      "..ython3.7/json/__init__.py:274 load  10000  0.050002  18.25148  0.001825\n",
      "..thon3.7/json/__init__.py:299 loads  10000  0.043793  17.84126  0.001784\n",
      "../decoder.py:332 JSONDecoder.decode  10000  0.080255  17.78021  0.001778\n",
      "..oder.py:343 JSONDecoder.raw_decode  10000  17.63666  17.63667  0.001764\n",
      "..3.7/pathlib.py:1091 PosixPath.glob  10001  0.017548  0.261054  0.000026\n",
      "..516 _WildcardSelector._select_from  10001  0.106270  0.243425  0.000024\n",
      "..s.py:319 IncrementalDecoder.decode  10000  0.030140  0.124473  0.000012\n",
      "..athlib.py:703 PosixPath.__fspath__  10002  0.025801  0.104998  0.000010\n",
      "..1020 PosixPath._make_child_relpath  10000  0.026290  0.080040  0.000008\n",
      "..7/pathlib.py:693 PosixPath.__str__  10002  0.044364  0.079197  0.000008\n",
      "..rnel_15257/339865336.py:5 <lambda>  10000  0.015156  0.077930  0.000008\n",
      "..n3.7/pathlib.py:806 PosixPath.stem  10000  0.028821  0.062773  0.000006\n",
      "../pathlib.py:666 _from_parsed_parts  10000  0.032345  0.053750  0.000005\n",
      "..tlocale.py:33 getpreferredencoding  10000  0.026720  0.048639  0.000005\n",
      "..athlib.py:676 _format_parsed_parts  10001  0.021704  0.034834  0.000003\n",
      "..py:309 IncrementalDecoder.__init__  10000  0.020904  0.030830  0.000003\n",
      "..n3.7/pathlib.py:779 PosixPath.name  10000  0.014684  0.020969  0.000002\n",
      ".. _TerminatingSelector._select_from  20000  0.015553  0.015553  0.000001\n",
      "..hlib.py:305 _PosixFlavour.casefold  20001  0.015306  0.015306  0.000001\n",
      "...7/pathlib.py:1010 PosixPath._init  10001  0.010404  0.010404  0.000001\n",
      "..py:260 IncrementalDecoder.__init__  10000  0.009926  0.009926  0.000001\n",
      "..lets/traitlets.py:597 List.__set__  2      0.000013  0.000251  0.000126\n",
      "..aitlets/traitlets.py:2647 List.set  2      0.000026  0.000239  0.000119\n",
      "..raitlets/traitlets.py:579 List.set  2      0.000022  0.000208  0.000104\n",
      "..ncio/events.py:86 TimerHandle._run  1      0.000014  0.000118  0.000118\n",
      "..:1218 HistoryManager._notify_trait  1      0.000011  0.000103  0.000103\n",
      "..:1227 HistoryManager.notify_change  1      0.000008  0.000092  0.000092\n",
      "../python3.7/pathlib.py:1000 __new__  1      0.000013  0.000091  0.000091\n",
      "..1 HistoryManager._notify_observers  1      0.000051  0.000084  0.000084\n",
      "..ts/traitlets.py:608 List._validate  2      0.000019  0.000084  0.000042\n",
      "..p.py:132 XCachingCompiler.__call__  2      0.000013  0.000083  0.000042\n",
      "..thon3.7/pathlib.py:653 _from_parts  1      0.000014  0.000076  0.000076\n",
      "..p.py:733 AsyncIOLoop._run_callback  1      0.000022  0.000076  0.000076\n",
      "..:476 _WildcardSelector.select_from  1      0.000006  0.000067  0.000067\n",
      "..thon3.7/pathlib.py:633 _parse_args  1      0.000024  0.000061  0.000061\n",
      "..7/pathlib.py:1346 PosixPath.is_dir  1      0.000009  0.000060  0.000060\n",
      "..l/iostream.py:481 OutStream._flush  1      0.000013  0.000054  0.000054\n",
      "..3.7/pathlib.py:1156 PosixPath.stat  1      0.000005  0.000050  0.000050\n",
      "..ts/traitlets.py:2499 List.validate  2      0.000019  0.000049  0.000025\n",
      "..ib.py:62 _PosixFlavour.parse_parts  2      0.000031  0.000045  0.000022\n",
      "..eam.py:562 OutStream._flush_buffer  1      0.000017  0.000042  0.000042\n",
      ".._GeneratorContextManager.__enter__  2      0.000008  0.000024  0.000012\n",
      "..python3.7/contextlib.py:237 helper  2      0.000010  0.000023  0.000012\n",
      "..tils/bunch.py:11 Bunch.__getattr__  3      0.000020  0.000022  0.000007\n",
      "..ets.py:2640 List.validate_elements  2      0.000016  0.000021  0.000011\n",
      "..am.py:569 OutStream._rotate_buffer  1      0.000018  0.000018  0.000018\n",
      "..py:631 _UnixSelectorEventLoop.time  1      0.000013  0.000016  0.000016\n",
      ".. _GeneratorContextManager.__exit__  2      0.000007  0.000016  0.000008\n",
      "..itlets.py:617 List._cross_validate  2      0.000010  0.000014  0.000007\n",
      "..toryManager._writeout_output_cache  1      0.000008  0.000013  0.000013\n",
      ".. _GeneratorContextManager.__init__  2      0.000010  0.000013  0.000006\n",
      "..lets/traitlets.py:566 Bool.__get__  2      0.000008  0.000013  0.000006\n",
      "..03 CommandChainDispatcher.__call__  2      0.000009  0.000011  0.000006\n",
      "..eractiveShell._update_code_co_name  2      0.000007  0.000009  0.000005\n",
      "..y:166 XCachingCompiler.extra_flags  4      0.000009  0.000009  0.000002\n",
      "..lets/traitlets.py:566 List.__get__  2      0.000006  0.000008  0.000004\n",
      "..e/interactiveshell.py:124 <lambda>  2      0.000007  0.000007  0.000003\n",
      "..ts/traitlets.py:1847 List.validate  2      0.000005  0.000006  0.000003\n",
      "..reading.py:240 Condition.__enter__  1      0.000004  0.000004  0.000004\n",
      "..raitlets/traitlets.py:535 Bool.get  2      0.000004  0.000004  0.000002\n",
      "..reading.py:255 Condition._is_owned  1      0.000003  0.000004  0.000004\n",
      "..ages/cycler.py:239 Cycler.__iter__  2      0.000004  0.000004  0.000002\n",
      "..ipstruct.py:125 Struct.__getattr__  2      0.000004  0.000004  0.000002\n",
      "..ets.py:2510 List.validate_elements  2      0.000004  0.000004  0.000002\n",
      "..lib.py:290 _PosixFlavour.splitroot  2      0.000003  0.000003  0.000002\n",
      "..xSelectorEventLoop._process_events  1      0.000003  0.000003  0.000003\n",
      "..ZMQInteractiveShell.user_global_ns  2      0.000003  0.000003  0.000002\n",
      "..ing.py:249 Condition._release_save  1      0.000002  0.000003  0.000003\n",
      "..raitlets/traitlets.py:535 List.get  2      0.000002  0.000002  0.000001\n",
      "..e/interactiveshell.py:3424 compare  2      0.000002  0.000002  0.000001\n",
      "..InteractiveShell.pre_run_code_hook  2      0.000002  0.000002  0.000001\n",
      "..hon3.7/threading.py:534 Event.wait  1      0.000000  0.000000  0.000000\n",
      "...7/threading.py:264 Condition.wait  1      0.000000  0.000000  0.000000\n",
      "..nel_15257/2801445275.py:7 <module>  1      0.000000  0.000000  0.000000\n",
      "..9 _UnixSelectorEventLoop._run_once  1      0.000000  0.000000  0.000000\n",
      "..ectors.py:451 EpollSelector.select  1      0.000000  0.000000  0.000000\n"
     ]
    }
   ],
   "source": [
    "import yappi\n",
    "\n",
    "yappi.set_clock_type(\"cpu\")\n",
    "yappi.clear_stats()\n",
    "yappi.start()\n",
    "_ = load_and_combine_stats(Path(\"data/large\"))\n",
    "yappi.stop()\n",
    "yappi.get_func_stats().print_all()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "e01da51d-8129-4b6c-aadc-d9fb74891772",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='profiles/flamegraph.svg' target='_blank'>profiles/flamegraph.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/flamegraph.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from os import getpid\n",
    "from subprocess import DEVNULL, Popen\n",
    "\n",
    "from IPython.display import FileLink, display\n",
    "\n",
    "py_spy_process = Popen(\n",
    "    [\n",
    "        \"py-spy\",\n",
    "        \"record\",\n",
    "        \"--native\",\n",
    "        \"--subprocesses\",\n",
    "        \"--nolineno\",\n",
    "        \"-o\",\n",
    "        \"profiles/flamegraph.svg\",\n",
    "        \"--duration\",\n",
    "        \"35\",\n",
    "        \"--pid\",\n",
    "        str(getpid()),\n",
    "    ],\n",
    "    stdout=DEVNULL,\n",
    ")\n",
    "\n",
    "_ = load_and_combine_stats(Path(\"data/large\"))\n",
    "\n",
    "py_spy_process.wait()\n",
    "display(FileLink(\"profiles/flamegraph.svg\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "abca65ef-6ff0-457a-82f9-b92fb50f1842",
   "metadata": {},
   "outputs": [],
   "source": [
    "@measure_and_compare()\n",
    "def load_and_combine_stats(path):\n",
    "    chunked_stats = load_data(path)\n",
    "    return combine_stats(*chunked_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7c2147a7-4e24-4447-940e-c1c08ebc771f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "22.28s (28.92s wall)  load_and_combine_stats(data/large)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='profiles/load_and_combine_stats__data_large.svg' target='_blank'>profiles/load_and_combine_stats__data_large.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/load_and_combine_stats__data_large.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats(Path(\"data/large\"), max_time=35)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bf9b10dd-d5a5-4f81-aee3-87d391a97660",
   "metadata": {},
   "source": [
    "# Efficient IO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "52a1490d-8fa3-4bce-8235-f1a891fc333a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import h5py\n",
    "\n",
    "\n",
    "def load_data_hdf5(path):\n",
    "    segments = []\n",
    "    counts = []\n",
    "    centers = []\n",
    "    for hdf5_path in sorted(path.glob(\"*.hdf5\"), key=lambda p: int(p.stem)):\n",
    "        with h5py.File(hdf5_path, \"r\") as f:\n",
    "            segments.append(f[\"segments\"][:])\n",
    "            counts.append(f[\"counts\"][:])\n",
    "            centers.append(f[\"centers\"][:])\n",
    "    return segments, counts, centers\n",
    "\n",
    "\n",
    "@measure_and_compare()\n",
    "def load_and_combine_stats_v1(path):\n",
    "    chunked_stats = load_data_hdf5(path)\n",
    "    return combine_stats(*chunked_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "48228bfc-7b10-4829-9b08-ee5bb3eef20c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14.47s (27.11s wall) load_and_combine_stats_v1(data/large), 54.04% faster (load_and_combine_stats was 22.28s)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='profiles/load_and_combine_stats_v1__data_large.svg' target='_blank'>profiles/load_and_combine_stats_v1__data_large.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/load_and_combine_stats_v1__data_large.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats_v1(Path(\"data/small\"))\n",
    "\n",
    "plot(segmentation, combined_stats)\n",
    "\n",
    "combined_stats = load_and_combine_stats_v1(Path(\"data/large\"), max_time=30)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bfb733f0-2052-4aa4-8bef-024bfb0ee3ca",
   "metadata": {},
   "source": [
    "# Vectorization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "318aa3b1-cad8-4325-93f9-9c0a1fde26f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def combine_stats_numpy(chunked_stats):\n",
    "    chunked_segments, chunked_counts, chunked_centers = chunked_stats\n",
    "    counts_per_segment = {}\n",
    "    centers_per_segement = {}\n",
    "    for segments_i, counts_i, centers_i in zip(\n",
    "        chunked_segments, chunked_counts, chunked_centers\n",
    "    ):\n",
    "        for segment, count, center in zip(segments_i, counts_i, centers_i):\n",
    "            counts_per_segment.setdefault(segment, []).append(count)\n",
    "            centers_per_segement.setdefault(segment, []).append(center)\n",
    "\n",
    "    result_segments = []\n",
    "    result_counts = []\n",
    "    result_centers = []\n",
    "    for segment in counts_per_segment.keys():\n",
    "        segment_counts = np.array(counts_per_segment[segment])\n",
    "        segment_centers = np.array(centers_per_segement[segment])\n",
    "        count_sum = segment_counts.sum()\n",
    "        center = (segment_centers * segment_counts[:, np.newaxis]).sum(\n",
    "            axis=0\n",
    "        ) / count_sum\n",
    "        result_segments.append(segment)\n",
    "        result_counts.append(count_sum)\n",
    "        result_centers.append(center)\n",
    "    return result_segments, result_counts, result_centers\n",
    "\n",
    "\n",
    "@measure_and_compare()\n",
    "def load_and_combine_stats_v2(path):\n",
    "    chunked_stats = load_data_hdf5(path)\n",
    "    return combine_stats_numpy(chunked_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c7bf0b89-cbd7-414c-b1ad-fb949a7d3e99",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12.79s (23.71s wall) load_and_combine_stats_v2(data/large), 74.22% faster (load_and_combine_stats was 22.28s)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='profiles/load_and_combine_stats_v2__data_large.svg' target='_blank'>profiles/load_and_combine_stats_v2__data_large.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/load_and_combine_stats_v2__data_large.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats_v2(Path(\"data/small\"))\n",
    "\n",
    "plot(segmentation, combined_stats)\n",
    "\n",
    "combined_stats = load_and_combine_stats_v2(Path(\"data/large\"), max_time=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9d39d034-8c27-43bf-9590-8a2f0928dafc",
   "metadata": {},
   "outputs": [],
   "source": [
    "def combine_stats_numpy_fast_groupby(chunked_stats):\n",
    "    all_segments = np.concatenate([i for i in chunked_stats[0] if len(i) > 0])\n",
    "    all_counts = np.concatenate([i for i in chunked_stats[1] if len(i) > 0])\n",
    "    all_centers = np.concatenate([i for i in chunked_stats[2] if len(i) > 0])\n",
    "\n",
    "    # sort all arrays by all_segments\n",
    "    sort_idx = np.argsort(all_segments)\n",
    "    all_segments = all_segments[sort_idx]\n",
    "    all_counts = all_counts[sort_idx]\n",
    "    all_centers = all_centers[sort_idx]\n",
    "\n",
    "    # group by segments (g_idx is where the groups start)\n",
    "    g_segments, g_idx = np.unique(all_segments, return_index=True)\n",
    "    g_counts = np.split(all_counts, g_idx[1:])\n",
    "    g_centers = np.split(all_centers, g_idx[1:])\n",
    "\n",
    "    # empty arrays for results\n",
    "    segments = np.ndarray(len(g_segments), dtype=all_segments.dtype)\n",
    "    counts = np.ndarray(len(g_segments), dtype=all_counts.dtype)\n",
    "    centers = np.ndarray((len(g_segments), 2), dtype=all_centers.dtype)\n",
    "\n",
    "    # combine stats for each segment\n",
    "    for i, (segment, segment_counts, segment_centers) in enumerate(\n",
    "        zip(g_segments, g_counts, g_centers)\n",
    "    ):\n",
    "        count_sum = segment_counts.sum()\n",
    "        center = (segment_centers * segment_counts[:, np.newaxis]).sum(\n",
    "            axis=0\n",
    "        ) / count_sum\n",
    "        segments[i] = segment\n",
    "        counts[i] = count_sum\n",
    "        centers[i] = center\n",
    "    return segments, counts, centers\n",
    "\n",
    "\n",
    "@measure_and_compare()\n",
    "def load_and_combine_stats_v3(path):\n",
    "    chunked_stats = load_data_hdf5(path)\n",
    "    return combine_stats_numpy_fast_groupby(chunked_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "2d6f7125-5832-4fcf-8075-33b687917f6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 8.04s (16.74s wall) load_and_combine_stats_v3(data/large), 177.14% faster (load_and_combine_stats was 22.28s)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='profiles/load_and_combine_stats_v3__data_large.svg' target='_blank'>profiles/load_and_combine_stats_v3__data_large.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/load_and_combine_stats_v3__data_large.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats_v3(Path(\"data/small\"))\n",
    "\n",
    "plot(segmentation, combined_stats)\n",
    "\n",
    "combined_stats = load_and_combine_stats_v3(Path(\"data/large\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07f52ff8-c589-4114-985b-8b8d21bbf129",
   "metadata": {},
   "source": [
    "# Memory & Precision Tradeoffs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "3b730a26-a731-4b4d-91af-89de04760687",
   "metadata": {},
   "outputs": [],
   "source": [
    "def combine_stats_numpy_fast_groupby(chunked_stats):\n",
    "    all_segments = np.concatenate(\n",
    "        [i.astype(np.uint32) for i in chunked_stats[0] if len(i) > 0]\n",
    "    )\n",
    "    all_counts = np.concatenate(\n",
    "        [i.astype(np.uint32) for i in chunked_stats[1] if len(i) > 0]\n",
    "    )\n",
    "    all_centers = np.concatenate(\n",
    "        [i.astype(np.float32) for i in chunked_stats[2] if len(i) > 0]\n",
    "    )\n",
    "\n",
    "    # sort all arrays by all_segments\n",
    "    sort_idx = np.argsort(all_segments)\n",
    "    all_segments = all_segments[sort_idx]\n",
    "    all_counts = all_counts[sort_idx]\n",
    "    all_centers = all_centers[sort_idx]\n",
    "\n",
    "    # group by segments (g_idx is where the groups start)\n",
    "    g_segments, g_idx = np.unique(all_segments, return_index=True)\n",
    "    g_counts = np.split(all_counts, g_idx[1:])\n",
    "    g_centers = np.split(all_centers, g_idx[1:])\n",
    "\n",
    "    # empty arrays for results\n",
    "    segments = np.ndarray(len(g_segments), dtype=all_segments.dtype)\n",
    "    counts = np.ndarray(len(g_segments), dtype=all_counts.dtype)\n",
    "    centers = np.ndarray((len(g_segments), 2), dtype=all_centers.dtype)\n",
    "\n",
    "    # combine stats for each segment\n",
    "    for i, (segment, segment_counts, segment_centers) in enumerate(\n",
    "        zip(g_segments, g_counts, g_centers)\n",
    "    ):\n",
    "        count_sum = segment_counts.sum()\n",
    "        center = (segment_centers * segment_counts[:, np.newaxis]).sum(\n",
    "            axis=0\n",
    "        ) / count_sum\n",
    "        segments[i] = segment\n",
    "        counts[i] = count_sum\n",
    "        centers[i] = center\n",
    "    return segments, counts, centers\n",
    "\n",
    "\n",
    "@measure_and_compare()\n",
    "def load_and_combine_stats_v4(path):\n",
    "    chunked_stats = load_data_hdf5(path)\n",
    "    return combine_stats_numpy_fast_groupby(chunked_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "540906b5-9bf7-49ba-a9cf-820369b02429",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 7.56s (15.86s wall) load_and_combine_stats_v4(data/large), 194.68% faster (load_and_combine_stats was 22.28s)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='profiles/load_and_combine_stats_v4__data_large.svg' target='_blank'>profiles/load_and_combine_stats_v4__data_large.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/load_and_combine_stats_v4__data_large.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats_v4(Path(\"data/small\"))\n",
    "\n",
    "plot(segmentation, combined_stats)\n",
    "\n",
    "combined_stats = load_and_combine_stats_v4(Path(\"data/large\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed4003e4-a6bb-4ded-9f97-ddb9fe3b8305",
   "metadata": {},
   "source": [
    "# Jit-ting with numba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "e8441682-1cca-4804-9130-92b8ad57df36",
   "metadata": {},
   "outputs": [],
   "source": [
    "@numba.njit\n",
    "def combine_stats_per_segment(segment_counts, segment_centers):\n",
    "    count_sum = segment_counts.sum()\n",
    "    center = (segment_centers * segment_counts[:, np.newaxis]).sum(axis=0) / count_sum\n",
    "    return count_sum, center\n",
    "\n",
    "\n",
    "def combine_stats_numpy_fast_groupby_numba(chunked_stats):\n",
    "    all_segments = np.concatenate(\n",
    "        [i.astype(np.uint32) for i in chunked_stats[0] if len(i) > 0]\n",
    "    )\n",
    "    all_counts = np.concatenate(\n",
    "        [i.astype(np.uint32) for i in chunked_stats[1] if len(i) > 0]\n",
    "    )\n",
    "    all_centers = np.concatenate(\n",
    "        [i.astype(np.float32) for i in chunked_stats[2] if len(i) > 0]\n",
    "    )\n",
    "    sort_idx = np.argsort(all_segments)\n",
    "    all_segments = all_segments[sort_idx]\n",
    "    all_counts = all_counts[sort_idx]\n",
    "    all_centers = all_centers[sort_idx]\n",
    "    g_segments, g_idx = np.unique(all_segments, return_index=True)\n",
    "    g_counts = np.split(all_counts, g_idx[1:])\n",
    "    g_centers = np.split(all_centers, g_idx[1:])\n",
    "\n",
    "    def empty_array(other):\n",
    "        return np.ndarray((len(other),) + other[0].shape[1:], dtype=other[0].dtype)\n",
    "\n",
    "    segments = empty_array(g_segments)\n",
    "    counts = empty_array(g_counts)\n",
    "    centers = empty_array(g_centers)\n",
    "    for i, (segment, segment_counts, segment_centers) in enumerate(\n",
    "        zip(g_segments, g_counts, g_centers)\n",
    "    ):\n",
    "        count_sum, center = combine_stats_per_segment(segment_counts, segment_centers)\n",
    "        segments[i] = segment\n",
    "        counts[i] = count_sum\n",
    "        centers[i] = center\n",
    "    return segments, counts, centers\n",
    "\n",
    "\n",
    "@measure_and_compare()\n",
    "def load_and_combine_stats_v5(path):\n",
    "    chunked_stats = load_data_hdf5(path)\n",
    "    return combine_stats_numpy_fast_groupby_numba(chunked_stats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "03405ac2-4f8a-427f-965a-afe9e28b5f25",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "ename": "TypingError",
     "evalue": "Failed in nopython mode pipeline (step: nopython frontend)\nNo implementation of function Function(<built-in function getitem>) found for signature:\n \n >>> getitem(array(uint32, 1d, C), Tuple(slice<a:b>, none))\n \nThere are 22 candidate implementations:\n  - Of which 20 did not match due to:\n  Overload of function 'getitem': File: <numerous>: Line N/A.\n    With argument(s): '(array(uint32, 1d, C), Tuple(slice<a:b>, none))':\n   No match.\n  - Of which 2 did not match due to:\n  Overload in function 'GetItemBuffer.generic': File: numba/core/typing/arraydecl.py: Line 166.\n    With argument(s): '(array(uint32, 1d, C), Tuple(slice<a:b>, none))':\n   Rejected as the implementation raised a specific error:\n     NumbaTypeError: unsupported array index type none in Tuple(slice<a:b>, none)\n  raised from /home/jonathan/.cache/pypoetry/virtualenvs/data-analysis-speedup-Yl14xPLQ-py3.7/lib/python3.7/site-packages/numba/core/typing/arraydecl.py:73\n\nDuring: typing of intrinsic-call at /tmp/ipykernel_15257/1199427906.py (4)\nDuring: typing of static-get-item at /tmp/ipykernel_15257/1199427906.py (4)\n\nFile \"../../../../tmp/ipykernel_15257/1199427906.py\", line 4:\n<source missing, REPL/exec in use?>\n",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypingError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[0;32m/tmp/ipykernel_15257/2708378825.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mcombined_stats\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mload_and_combine_stats_v5\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mPath\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"data/large\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/code/data-analysis-speedup/utils/measure_and_compare.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     50\u001b[0m             \u001b[0mwall_start\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mperf_counter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     51\u001b[0m             \u001b[0mstart\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mprocess_time\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 52\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     53\u001b[0m             \u001b[0mend\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mprocess_time\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     54\u001b[0m             \u001b[0mwall_end\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mperf_counter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/tmp/ipykernel_15257/1199427906.py\u001b[0m in \u001b[0;36mload_and_combine_stats_v5\u001b[0;34m(path)\u001b[0m\n\u001b[1;32m     43\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mload_and_combine_stats_v5\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     44\u001b[0m     \u001b[0mchunked_stats\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mload_data_hdf5\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 45\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mcombine_stats_numpy_fast_groupby_numba\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mchunked_stats\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/tmp/ipykernel_15257/1199427906.py\u001b[0m in \u001b[0;36mcombine_stats_numpy_fast_groupby_numba\u001b[0;34m(chunked_stats)\u001b[0m\n\u001b[1;32m     33\u001b[0m         \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_segments\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mg_counts\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mg_centers\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     34\u001b[0m     ):\n\u001b[0;32m---> 35\u001b[0;31m         \u001b[0mcount_sum\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcenter\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcombine_stats_per_segment\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msegment_counts\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msegment_centers\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     36\u001b[0m         \u001b[0msegments\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msegment\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     37\u001b[0m         \u001b[0mcounts\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcount_sum\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.cache/pypoetry/virtualenvs/data-analysis-speedup-Yl14xPLQ-py3.7/lib/python3.7/site-packages/numba/core/dispatcher.py\u001b[0m in \u001b[0;36m_compile_for_args\u001b[0;34m(self, *args, **kws)\u001b[0m\n\u001b[1;32m    466\u001b[0m                 \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpatch_message\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmsg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    467\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 468\u001b[0;31m             \u001b[0merror_rewrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'typing'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    469\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0merrors\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mUnsupportedError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    470\u001b[0m             \u001b[0;31m# Something unsupported is present in the user code, add help info\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.cache/pypoetry/virtualenvs/data-analysis-speedup-Yl14xPLQ-py3.7/lib/python3.7/site-packages/numba/core/dispatcher.py\u001b[0m in \u001b[0;36merror_rewrite\u001b[0;34m(e, issue_type)\u001b[0m\n\u001b[1;32m    407\u001b[0m                 \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    408\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 409\u001b[0;31m                 \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwith_traceback\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    410\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    411\u001b[0m         \u001b[0margtypes\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mTypingError\u001b[0m: Failed in nopython mode pipeline (step: nopython frontend)\nNo implementation of function Function(<built-in function getitem>) found for signature:\n \n >>> getitem(array(uint32, 1d, C), Tuple(slice<a:b>, none))\n \nThere are 22 candidate implementations:\n  - Of which 20 did not match due to:\n  Overload of function 'getitem': File: <numerous>: Line N/A.\n    With argument(s): '(array(uint32, 1d, C), Tuple(slice<a:b>, none))':\n   No match.\n  - Of which 2 did not match due to:\n  Overload in function 'GetItemBuffer.generic': File: numba/core/typing/arraydecl.py: Line 166.\n    With argument(s): '(array(uint32, 1d, C), Tuple(slice<a:b>, none))':\n   Rejected as the implementation raised a specific error:\n     NumbaTypeError: unsupported array index type none in Tuple(slice<a:b>, none)\n  raised from /home/jonathan/.cache/pypoetry/virtualenvs/data-analysis-speedup-Yl14xPLQ-py3.7/lib/python3.7/site-packages/numba/core/typing/arraydecl.py:73\n\nDuring: typing of intrinsic-call at /tmp/ipykernel_15257/1199427906.py (4)\nDuring: typing of static-get-item at /tmp/ipykernel_15257/1199427906.py (4)\n\nFile \"../../../../tmp/ipykernel_15257/1199427906.py\", line 4:\n<source missing, REPL/exec in use?>\n"
     ]
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats_v5(Path(\"data/large\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "b8ac7813-b61b-422e-9c71-bc1aa6b06bc9",
   "metadata": {},
   "outputs": [],
   "source": [
    "@numba.njit()\n",
    "def combine_stats_per_segment(segment_counts, segment_centers):\n",
    "    count_sum = segment_counts.sum()\n",
    "    center = (segment_centers * segment_counts.reshape((-1, 1))).sum(axis=0) / count_sum\n",
    "    return count_sum, center"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "0a95c966-5ed9-4159-91e6-b1dfadeb9d12",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 8.34s (16.35s wall) load_and_combine_stats_v5(data/large), 167.15% faster (load_and_combine_stats was 22.28s)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='profiles/load_and_combine_stats_v5__data_large.svg' target='_blank'>profiles/load_and_combine_stats_v5__data_large.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/load_and_combine_stats_v5__data_large.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 7.61s (16.17s wall) load_and_combine_stats_v5(data/large), 192.88% faster (load_and_combine_stats was 22.28s)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<a href='profiles/load_and_combine_stats_v5__data_large.svg' target='_blank'>profiles/load_and_combine_stats_v5__data_large.svg</a><br>"
      ],
      "text/plain": [
       "/home/jonathan/code/data-analysis-speedup/profiles/load_and_combine_stats_v5__data_large.svg"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_stats = load_and_combine_stats_v5(Path(\"data/large\"))\n",
    "combined_stats = load_and_combine_stats_v5(Path(\"data/large\"))"
   ]
  }
 ],
 "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": 5
}