Argo-Awkward Array demo

argo-demo.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b7d1be8e-7d8a-4b9e-a4cb-37c25a9cf03e",
   "metadata": {},
   "source": [
    "Get Awkward Array from\n",
    "\n",
    "```bash\n",
    "pip install 'awkward>=1.9.0rc2'\n",
    "```\n",
    "\n",
    "We'll be using version 2.0, the development version, which is a submodule within 1.9.0rc2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2fe4578e-ee28-414c-8ba7-0bccff23fa5e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import awkward._v2 as ak"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d0c27481-c95f-4e20-9b90-d579b829e9d8",
   "metadata": {},
   "source": [
    "This file has all of the Argo data from 1997 through 2021 inclusive.\n",
    "\n",
    "   * It is 7.0 GB.\n",
    "   * The equivalent NetCDF is 135 GB.\n",
    "   * Uncompressed, the expert-level data is 118 GB, and the subset that is standard-level data is 42 GB.\n",
    "\n",
    "We'll be opening just the first row group (10000 sets of levels) to explore the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "447a4186-e863-47bc-8ff7-e6d611dc9b46",
   "metadata": {},
   "outputs": [],
   "source": [
    "expert_fields = ak.from_parquet(\"s3://pivarski-princeton/argo-floats-expert.parquet\", row_groups=[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5f05065-faf0-43c1-9845-f40002f6a53a",
   "metadata": {},
   "source": [
    "The data consist of nested record structures and variable-length lists.\n",
    "\n",
    "`show` prints as much as will fit into 20 lines and 80 characters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "65d4fb8f-dfa7-4dca-a7c7-b216df82fc73",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[{latitude: -0.126, longitude: -11.9, time: 1997-07-28T20:26:20.000000, ...},\n",
      " {latitude: 0.267, longitude: -16, time: 1997-07-29T20:03:00.000000, ...},\n",
      " {latitude: 0.236, longitude: -19.7, time: 1997-07-30T14:45:11.000000, ...},\n",
      " {latitude: -0.402, longitude: -27, time: 1997-08-01T07:59:00.000000, ...},\n",
      " {latitude: 0.429, longitude: -34.2, time: 1997-08-02T15:53:47.000000, ...},\n",
      " {latitude: 0.497, longitude: -37.1, time: 1997-08-03T17:23:13.000000, ...},\n",
      " {latitude: -0.511, longitude: -36.7, time: 1997-08-03T03:01:01.000000, ...},\n",
      " {latitude: 0.4, longitude: -40.2, time: 1997-08-04T06:10:56.000000, ...},\n",
      " {latitude: 0.072, longitude: -17.7, time: 1997-08-09T19:21:12.000000, ...},\n",
      " {latitude: -0.035, longitude: -13.8, time: 1997-08-09T01:52:41.000000, ...},\n",
      " ...,\n",
      " {latitude: 4.72, longitude: -41.3, time: 2002-07-06T21:56:43.000000, ...},\n",
      " {latitude: 43.5, longitude: -45.8, time: 2002-07-06T21:45:24.000000, ...},\n",
      " {latitude: 6.33, longitude: -27.4, time: 2002-07-06T20:25:41.000000, ...},\n",
      " {latitude: 2.68, longitude: -34.2, time: 2002-07-06T18:04:48.000000, ...},\n",
      " {latitude: 26.6, longitude: -68.8, time: 2002-07-06T15:53:00.000000, ...},\n",
      " {latitude: 3.81, longitude: -31.2, time: 2002-07-06T15:05:40.000000, ...},\n",
      " {latitude: 35, longitude: -32, time: 2002-07-06T14:09:28.000000, ...},\n",
      " {latitude: 24.1, longitude: -49.7, time: 2002-07-06T09:57:21.000000, ...},\n",
      " {latitude: 32.1, longitude: -19.6, time: 2002-07-06T08:14:23.000000, ...}]\n"
     ]
    }
   ],
   "source": [
    "expert_fields.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4278ae80-9b69-484e-9ebf-34e5393f19f5",
   "metadata": {},
   "source": [
    "The data's `type` also has a `show`, which reveals the structure without any values.\n",
    "\n",
    "The syntax is [datashape](https://datashape.readthedocs.io/en/latest/); the `10000 *` means an array of fixed length and the `var *` means lists of variable lengths."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "391ff890-1530-4da9-93c7-03d339ea53b2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000 * {\n",
      "    latitude: float64,\n",
      "    longitude: float64,\n",
      "    time: datetime64[us],\n",
      "    levels: var * {\n",
      "        pres: float32,\n",
      "        pres_adjusted: float32,\n",
      "        pres_adjusted_error: float32,\n",
      "        pres_adjusted_qc: string,\n",
      "        pres_qc: string,\n",
      "        psal: float32,\n",
      "        psal_adjusted: float32,\n",
      "        psal_adjusted_error: float32,\n",
      "        psal_adjusted_qc: string,\n",
      "        psal_qc: string,\n",
      "        temp: float32,\n",
      "        temp_adjusted: float32,\n",
      "        temp_adjusted_error: float32,\n",
      "        temp_adjusted_qc: string,\n",
      "        temp_qc: string\n",
      "    },\n",
      "    config_mission_number: int32,\n",
      "    cycle_number: int32,\n",
      "    data_centre: string,\n",
      "    data_mode: string,\n",
      "    data_state_indicator: string,\n",
      "    dc_reference: string,\n",
      "    direction: string,\n",
      "    firmware_version: string,\n",
      "    float_serial_no: string,\n",
      "    pi_name: string,\n",
      "    platform_number: int32,\n",
      "    platform_type: string,\n",
      "    positioning_system: string,\n",
      "    position_qc: string,\n",
      "    profile_pres_qc: string,\n",
      "    profile_psal_qc: string,\n",
      "    profile_temp_qc: string,\n",
      "    project_name: string,\n",
      "    time_location: datetime64[us],\n",
      "    time_qc: string,\n",
      "    vertical_sampling_scheme: string,\n",
      "    wmo_inst_type: int16\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "expert_fields.type.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b694b9e-092e-498b-b9ab-10e6cdd16997",
   "metadata": {},
   "source": [
    "Find a few that are small enough to print out."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a1722a9c-a58a-45b3-b584-7010884a8422",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([573, 632]),)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.nonzero(ak.num(expert_fields.levels) == 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "78714285-558c-48b0-9ade-fb545175cd12",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'latitude': 43.134,\n",
       " 'longitude': -32.227,\n",
       " 'time': datetime.datetime(1998, 8, 9, 2, 26, 20),\n",
       " 'levels': [{'pres': 1343.199951171875,\n",
       "   'pres_adjusted': 1343.199951171875,\n",
       "   'pres_adjusted_error': 2.4000000953674316,\n",
       "   'pres_adjusted_qc': '1',\n",
       "   'pres_qc': '1',\n",
       "   'psal': 34.94169998168945,\n",
       "   'psal_adjusted': nan,\n",
       "   'psal_adjusted_error': nan,\n",
       "   'psal_adjusted_qc': '4',\n",
       "   'psal_qc': '4',\n",
       "   'temp': 4.125,\n",
       "   'temp_adjusted': nan,\n",
       "   'temp_adjusted_error': nan,\n",
       "   'temp_adjusted_qc': '4',\n",
       "   'temp_qc': '4'},\n",
       "  {'pres': 1358.0999755859375,\n",
       "   'pres_adjusted': 1358.0999755859375,\n",
       "   'pres_adjusted_error': 2.4000000953674316,\n",
       "   'pres_adjusted_qc': '1',\n",
       "   'pres_qc': '1',\n",
       "   'psal': 35.01839828491211,\n",
       "   'psal_adjusted': nan,\n",
       "   'psal_adjusted_error': nan,\n",
       "   'psal_adjusted_qc': '4',\n",
       "   'psal_qc': '4',\n",
       "   'temp': 3.7119998931884766,\n",
       "   'temp_adjusted': nan,\n",
       "   'temp_adjusted_error': nan,\n",
       "   'temp_adjusted_qc': '4',\n",
       "   'temp_qc': '4'}],\n",
       " 'config_mission_number': 1,\n",
       " 'cycle_number': 6,\n",
       " 'data_centre': 'IF',\n",
       " 'data_mode': 'D',\n",
       " 'data_state_indicator': '2C  ',\n",
       " 'dc_reference': 'fl0173.006',\n",
       " 'direction': 'A',\n",
       " 'firmware_version': 'n/a',\n",
       " 'float_serial_no': '144',\n",
       " 'pi_name': 'Klaus-Peter KOLTERMANN',\n",
       " 'platform_number': 69018,\n",
       " 'platform_type': 'APEX',\n",
       " 'positioning_system': 'S   ',\n",
       " 'position_qc': '1',\n",
       " 'profile_pres_qc': 'A',\n",
       " 'profile_psal_qc': 'F',\n",
       " 'profile_temp_qc': 'F',\n",
       " 'project_name': 'Euro-Argo',\n",
       " 'time_location': datetime.datetime(1998, 8, 9, 5, 5, 59),\n",
       " 'time_qc': '1',\n",
       " 'vertical_sampling_scheme': 'Primary sampling: discrete []',\n",
       " 'wmo_inst_type': 846}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "expert_fields[573].tolist()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aa7eda88-f82c-4de5-bf92-5abf1092f144",
   "metadata": {},
   "source": [
    "Pull out a few fields and look at them individually."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "438bed35-fb99-45a7-b700-4d9aa8e54c4b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(<Array [-11.9, -16, -19.7, -27, ..., -32, -49.7, -19.6] type='10000 * float64'>,\n",
       " <Array [-0.126, 0.267, 0.236, ..., 35, 24.1, 32.1] type='10000 * float64'>,\n",
       " <Array [1997-07-28T20:26:20.000000, ...] type='10000 * datetime64[us]'>)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "expert_fields.longitude, expert_fields[\"latitude\"], expert_fields.time"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f65d61da-c758-4224-a0cf-aed3ffbe78e0",
   "metadata": {},
   "source": [
    "Look at just the temperature and its quality control.\n",
    "\n",
    "   * Numbers, slice objects, arrays, etc. slice rows.\n",
    "   * Strings slice columns, even nested columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "57b5e018-4c18-4b7e-8f4f-72fa22e50d0f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[{temp: 21.8, temp_qc: '1'}, {...}, ..., {...}, {temp: 4.45, temp_qc: '1'}],\n",
      " [{temp: 22.2, temp_qc: '1'}, {...}, ..., {...}, {temp: 4.45, temp_qc: '1'}],\n",
      " [{temp: 22.8, temp_qc: '1'}, {...}, ..., {...}, {temp: 5.17, temp_qc: '1'}],\n",
      " [{temp: 25.1, temp_qc: '1'}, {...}, ..., {...}, {temp: 4.66, temp_qc: '1'}],\n",
      " [{temp: 26.1, temp_qc: '1'}, {...}, ..., {...}, {temp: 4.54, temp_qc: '1'}],\n",
      " [{temp: 26.8, temp_qc: '4'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 26.5, temp_qc: '1'}, {...}, ..., {...}, {temp: 4.52, temp_qc: '1'}],\n",
      " [{temp: 26.4, temp_qc: '1'}, {...}, ..., {...}, {temp: 4.65, temp_qc: '1'}],\n",
      " [{temp: 23.3, temp_qc: '1'}, {...}, ..., {...}, {temp: 4.47, temp_qc: '1'}],\n",
      " [{temp: 22.6, temp_qc: '1'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " ...,\n",
      " [{temp: 28.3, temp_qc: '1'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 15, temp_qc: '1'}, {temp: 15, ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 28, temp_qc: '1'}, {temp: 28, ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 27.6, temp_qc: '1'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 28.1, temp_qc: '2'}, {temp: 28, ...}, ..., {temp: 5.5, temp_qc: '2'}],\n",
      " [{temp: 27.8, temp_qc: '1'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 22.3, temp_qc: '1'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 26.4, temp_qc: '1'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}],\n",
      " [{temp: 21.3, temp_qc: '1'}, {temp: ..., ...}, ..., {temp: nan, temp_qc: ' '}]]\n"
     ]
    }
   ],
   "source": [
    "expert_fields[\"levels\", [\"temp\", \"temp_qc\"]].show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eec50460-e734-4f4a-9fba-ec26f0ad5f74",
   "metadata": {},
   "source": [
    "How many values are in each level?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "50ea0ec1-a6ed-41df-91d7-1aac8673c8c5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Array [102, 112, 109, 105, 107, ..., 369, 369, 369, 369] type='10000 * int64'>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ak.num(expert_fields.levels)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c517391-b4e7-4ac2-996b-76752656d8b2",
   "metadata": {},
   "source": [
    "Which pressures are above 40?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e5413900-b355-4fbf-935b-89c1b5a777b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Array [[False, False, False, ..., True, True], ...] type='10000 * var * bool'>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "expert_fields.levels.pres > 40"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b551da2c-2421-4ec3-96ee-30dfd1929170",
   "metadata": {},
   "source": [
    "Which sets of levels have at least one over 40?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "df3eb8ca-b699-40f1-942c-5805a3b07dbb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Array [True, True, True, True, ..., True, True, True] type='10000 * bool'>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ak.any(expert_fields.levels.pres > 40, axis=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ab95885-a3c1-4847-968f-b44d5655e7f4",
   "metadata": {},
   "source": [
    "Which salinities pass quality control?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "13813f22-d7ff-483d-b7a3-f4ed6ddcd4b4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Array [[False, False, ..., False, False], ...] type='10000 * var * bool'>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "expert_fields.levels.psal_qc == \"1\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3b29af7-5a85-4dcc-ae8d-af99243bd452",
   "metadata": {},
   "source": [
    "For which sets of levels do all salinities pass quality control?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "d9c5e4b9-7c51-4b59-8d98-d943126b7d6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Array [True, True, True, True, ..., False, False, False] type='10000 * bool'>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ak.all(expert_fields.levels.pres_qc == \"1\", axis=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74872c1e-2fa8-496f-bbc8-35aee6ad029f",
   "metadata": {},
   "source": [
    "What's the mean pressure in each set?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "0f7c839f-00b8-45da-8428-de3658c1e575",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Array [492, 499, 501, 510, ..., nan, nan, nan, nan] type='10000 * ?float64'>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ak.mean(expert_fields.levels.pres, axis=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a09f24ff-21e2-464b-b2a4-b9202ec157c3",
   "metadata": {},
   "source": [
    "Disregarding not-a-number values?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "370ed2e9-6f7d-41f4-9030-71a682ab5c21",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Array [492, 499, 501, 510, ..., 347, 315, 348, 572] type='10000 * ?float64'>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ak.nanmean(expert_fields.levels.pres, axis=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55e81472-f2ed-4fe5-9817-93061f0b206e",
   "metadata": {},
   "source": [
    "All of the above should be reminiscent of exploring data with NumPy, except that the data have more complex structures."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54b473e1-0d2c-4b6a-87db-f340c1fb0e13",
   "metadata": {},
   "source": [
    "Just as Numba can iterate over NumPy arrays, it can iterate over Awkward Arrays.\n",
    "\n",
    "The following calculates means, disregarding not-a-number, just like `nanmean`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "87b854d6-2666-4aa6-8b72-0047f8c9d0a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numba as nb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6d262bc0-ea6d-4895-890c-b5a59e253ffe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([491.65784372, 499.49642835, 501.15871513, ..., 314.63414634,\n",
       "       347.6744186 , 571.81818182])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "@nb.njit\n",
    "def walk_over_array(array):\n",
    "    out = np.empty(len(array))\n",
    "    for i, levels in enumerate(array.levels):\n",
    "        numer = 0.0\n",
    "        denom = 0.0\n",
    "        for pres in levels.pres:\n",
    "            if not np.isnan(pres):\n",
    "                numer += pres\n",
    "                denom += 1.0\n",
    "\n",
    "        if denom != 0.0:\n",
    "            out[i] = numer / denom\n",
    "        else:\n",
    "            out[i] = np.nan\n",
    "\n",
    "    return out\n",
    "\n",
    "walk_over_array(expert_fields)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e63a98d-a74e-4c3d-93c9-03b2d4c92ff7",
   "metadata": {},
   "source": [
    "Now to use this feature to make a complex cut: suppose we want only data that are a certain distance from shore?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "de5592bc-8584-454b-acd1-e9497f954305",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "1a674121-edb7-49c7-ae2c-804a994cad91",
   "metadata": {},
   "outputs": [],
   "source": [
    "land = matplotlib.image.imread(\"land.png\")[:, :, 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "4482fd99-f065-4427-a268-8defee205272",
   "metadata": {},
   "outputs": [],
   "source": [
    "@nb.njit\n",
    "def delta_longitude_latitude(at_latitude):\n",
    "    a, e2 = 6378.137, 0.00669437999014    # WGS84\n",
    "    radians = np.deg2rad(at_latitude)\n",
    "    partial = (1.0 - e2*np.sin(radians)**2)\n",
    "    delta_longitude = 180.0 * partial**0.5 / (np.pi*a*np.cos(radians))\n",
    "    delta_latitude = 180.0 * partial**1.5 / (np.pi*a*(1.0 - e2))\n",
    "    return delta_longitude, delta_latitude"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ae99b53e-aa9d-4018-b2c2-540c57a46803",
   "metadata": {},
   "outputs": [],
   "source": [
    "@nb.njit\n",
    "def fraction_in_land(longitude, latitude, ball_kilometers):\n",
    "    delta_longitude, delta_latitude = delta_longitude_latitude(latitude)\n",
    "    inv_longitude, inv_latitude = 1.0 / delta_longitude, 1.0 / delta_latitude\n",
    "\n",
    "    pixels_per_degree = 1.0\n",
    "    degrees_per_pixel = 1.0 / pixels_per_degree\n",
    "    min_horizontal = max(0, int(np.floor(pixels_per_degree * (180 + longitude - ball_kilometers * delta_longitude))))\n",
    "    max_horizontal = min(3600, int(np.ceil(pixels_per_degree * (180 + longitude + ball_kilometers * delta_longitude))))\n",
    "    min_vertical = max(1, int(np.floor(pixels_per_degree * (90 + latitude - ball_kilometers * delta_latitude))))\n",
    "    max_vertical = min(1801, int(np.floor(pixels_per_degree * (90 + latitude + ball_kilometers * delta_latitude))))\n",
    "\n",
    "    num_land = 0.0\n",
    "    num_pixels = 0.0\n",
    "    for horizontal in range(min_horizontal, max_horizontal + 1):\n",
    "        for vertical in range(min_vertical, max_vertical + 1):\n",
    "            kilometers_east = inv_longitude * (degrees_per_pixel * (horizontal + 0.5) - 180.0 - longitude)\n",
    "            kilometers_north = inv_latitude * (degrees_per_pixel * (vertical - 0.5) - 90.0 - latitude)\n",
    "            if np.sqrt(kilometers_east**2 + kilometers_north**2) < ball_kilometers:\n",
    "                num_land += land[-vertical, horizontal]\n",
    "                num_pixels += 1.0\n",
    "\n",
    "    if num_pixels == 0.0:\n",
    "        return 0.0\n",
    "    else:\n",
    "        return num_land / num_pixels"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46e99555-cd68-42d5-a057-3566c02fdd4d",
   "metadata": {},
   "source": [
    "Make it a NumPy ufunc.\n",
    "\n",
    "Awkward Arrays can be used with `@nb.jit` functions and any ufunc, including new ones made by Numba."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "8c0139d3-0077-4e51-a4d0-677283022ee3",
   "metadata": {},
   "outputs": [],
   "source": [
    "@nb.vectorize([nb.float64(nb.float64, nb.float64, nb.float64)])\n",
    "def fraction_in_land_ufunc(longitude, latitude, ball_kilometers):\n",
    "    return fraction_in_land(longitude, latitude, ball_kilometers)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a1d6369-d45d-4d48-b892-f1e28dc6a455",
   "metadata": {},
   "source": [
    "Now get the full dataset, but only a few columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "6b62abff-64f8-44af-8458-a7f26fd97e1a",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_coordinates = ak.from_parquet(\n",
    "    \"s3://pivarski-princeton/argo-floats-expert.parquet\",\n",
    "    columns=[\"longitude\", \"latitude\", \"time\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "509b78f6-7143-4309-974a-85a1be8f7bd0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 2534567)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(expert_fields), len(all_coordinates)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1015ef4-05cd-43a2-9da0-869f12ff48ee",
   "metadata": {},
   "source": [
    "Create a selection as an array of booleans and apply it as a slice (like NumPy)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "9bbfe47d-f81e-4042-9849-9822a41f05d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "selection = (\n",
    "    (fraction_in_land_ufunc(all_coordinates.longitude, all_coordinates.latitude, 1100.0) > 0.1) &\n",
    "    (fraction_in_land_ufunc(all_coordinates.longitude, all_coordinates.latitude, 990.0) < 0.1)\n",
    ")\n",
    "\n",
    "selected = all_coordinates[selection]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "1dc70471-cba8-4de9-99e8-77fd4b9482a7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.02544339920783313"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(selected) / len(all_coordinates)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "989de32d-4f2a-40f8-888a-797d54bbb230",
   "metadata": {},
   "source": [
    "Only measurements taken close to 1000 km from shore are included in the 2.5% that have been selected."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "3a479380-b61e-451e-8c3c-ddc99ddc2fa0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7f51c3479a60>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "\n",
    "ax.imshow(land, extent=(-180, 180, -90, 90), cmap=\"gray\")\n",
    "ax.scatter(selected.longitude, selected.latitude, s=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf3eb975-8ece-4487-b7a9-05ca03caf9b5",
   "metadata": {},
   "source": [
    "Eventually, data analysts will want Pandas DataFrames and Xarrays.\n",
    "\n",
    "Awkward Arrays could be used just in the selection process, then project the data onto standard rectilinear formats."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "f3a7c948-f953-4bf3-9d58-b997eeacbf28",
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_pandas(row_groups, selection_function):\n",
    "    awkward_array = ak.from_parquet(\n",
    "        \"/home/jpivarski/storage/data/argo-floats/argo-floats-expert.parquet\",   # use my local file instead of S3\n",
    "        row_groups=row_groups,\n",
    "        columns=[\n",
    "            \"latitude\",\n",
    "            \"longitude\",\n",
    "            \"time\",\n",
    "            \"levels.pres\",\n",
    "            \"levels.pres_qc\",\n",
    "            \"levels.psal\",\n",
    "            \"levels.psal_qc\",\n",
    "            \"levels.temp\",\n",
    "            \"levels.temp_qc\",\n",
    "            \"config_mission_number\",\n",
    "            \"cycle_number\",\n",
    "            \"data_mode\",\n",
    "            \"direction\",\n",
    "            \"platform_number\",\n",
    "            \"position_qc\",\n",
    "            \"time_qc\",\n",
    "        ],\n",
    "    )\n",
    "    # Treat any missing lists as empty lists.\n",
    "    awkward_array[\"levels\"] = ak.fill_none(awkward_array[\"levels\"], [], axis=1)\n",
    "    \n",
    "    # Apply the selection function.\n",
    "    selected = awkward_array[selection_function(awkward_array)]\n",
    "    \n",
    "    # Turn the quantities in levels into non-nested columns.\n",
    "    for name in selected[\"levels\"].fields:\n",
    "        selected[name] = selected[\"levels\", name]\n",
    "    \n",
    "    # Remove the original, nested version.\n",
    "    del selected[\"levels\"]\n",
    "    \n",
    "    # Now this projects into a Pandas DataFrame.\n",
    "    return ak.to_pandas(selected)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b6236300-d36e-4502-977a-e5ac2bd34766",
   "metadata": {},
   "outputs": [],
   "source": [
    "def selection_function(array):\n",
    "    return (\n",
    "        (fraction_in_land_ufunc(array.longitude, array.latitude, 1100.0) > 0.1) &\n",
    "        (fraction_in_land_ufunc(array.longitude, array.latitude, 990.0) < 0.1)\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0d208d8-e12c-4390-825e-9c977a9ac415",
   "metadata": {},
   "source": [
    "Here's the first three row groups."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "6189829c-8347-4633-aa59-51bedae586f8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>time</th>\n",
       "      <th>config_mission_number</th>\n",
       "      <th>cycle_number</th>\n",
       "      <th>data_mode</th>\n",
       "      <th>direction</th>\n",
       "      <th>platform_number</th>\n",
       "      <th>position_qc</th>\n",
       "      <th>time_qc</th>\n",
       "      <th>pres</th>\n",
       "      <th>pres_qc</th>\n",
       "      <th>psal</th>\n",
       "      <th>psal_qc</th>\n",
       "      <th>temp</th>\n",
       "      <th>temp_qc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>entry</th>\n",
       "      <th>subentry</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">0</th>\n",
       "      <th>0</th>\n",
       "      <td>-0.126</td>\n",
       "      <td>-11.863</td>\n",
       "      <td>1997-07-28 20:26:20</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>13858</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>15.500000</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>21.804001</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.126</td>\n",
       "      <td>-11.863</td>\n",
       "      <td>1997-07-28 20:26:20</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>13858</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>21.100000</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>21.788000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.126</td>\n",
       "      <td>-11.863</td>\n",
       "      <td>1997-07-28 20:26:20</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>13858</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>26.600000</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>21.521000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.126</td>\n",
       "      <td>-11.863</td>\n",
       "      <td>1997-07-28 20:26:20</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>13858</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>32.200001</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>20.671000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.126</td>\n",
       "      <td>-11.863</td>\n",
       "      <td>1997-07-28 20:26:20</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>13858</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>37.799999</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>19.691000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">1162</th>\n",
       "      <th>369</th>\n",
       "      <td>7.386</td>\n",
       "      <td>-45.771</td>\n",
       "      <td>2004-06-25 01:12:30</td>\n",
       "      <td>1</td>\n",
       "      <td>151</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>39007</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>370</th>\n",
       "      <td>7.386</td>\n",
       "      <td>-45.771</td>\n",
       "      <td>2004-06-25 01:12:30</td>\n",
       "      <td>1</td>\n",
       "      <td>151</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>39007</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>371</th>\n",
       "      <td>7.386</td>\n",
       "      <td>-45.771</td>\n",
       "      <td>2004-06-25 01:12:30</td>\n",
       "      <td>1</td>\n",
       "      <td>151</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>39007</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>372</th>\n",
       "      <td>7.386</td>\n",
       "      <td>-45.771</td>\n",
       "      <td>2004-06-25 01:12:30</td>\n",
       "      <td>1</td>\n",
       "      <td>151</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>39007</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>373</th>\n",
       "      <td>7.386</td>\n",
       "      <td>-45.771</td>\n",
       "      <td>2004-06-25 01:12:30</td>\n",
       "      <td>1</td>\n",
       "      <td>151</td>\n",
       "      <td>R</td>\n",
       "      <td>A</td>\n",
       "      <td>39007</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>316608 rows × 16 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                latitude  longitude                time  \\\n",
       "entry subentry                                            \n",
       "0     0           -0.126    -11.863 1997-07-28 20:26:20   \n",
       "      1           -0.126    -11.863 1997-07-28 20:26:20   \n",
       "      2           -0.126    -11.863 1997-07-28 20:26:20   \n",
       "      3           -0.126    -11.863 1997-07-28 20:26:20   \n",
       "      4           -0.126    -11.863 1997-07-28 20:26:20   \n",
       "...                  ...        ...                 ...   \n",
       "1162  369          7.386    -45.771 2004-06-25 01:12:30   \n",
       "      370          7.386    -45.771 2004-06-25 01:12:30   \n",
       "      371          7.386    -45.771 2004-06-25 01:12:30   \n",
       "      372          7.386    -45.771 2004-06-25 01:12:30   \n",
       "      373          7.386    -45.771 2004-06-25 01:12:30   \n",
       "\n",
       "                config_mission_number  cycle_number data_mode direction  \\\n",
       "entry subentry                                                            \n",
       "0     0                             1             1         R         A   \n",
       "      1                             1             1         R         A   \n",
       "      2                             1             1         R         A   \n",
       "      3                             1             1         R         A   \n",
       "      4                             1             1         R         A   \n",
       "...                               ...           ...       ...       ...   \n",
       "1162  369                           1           151         R         A   \n",
       "      370                           1           151         R         A   \n",
       "      371                           1           151         R         A   \n",
       "      372                           1           151         R         A   \n",
       "      373                           1           151         R         A   \n",
       "\n",
       "                platform_number position_qc time_qc       pres pres_qc  psal  \\\n",
       "entry subentry                                                                 \n",
       "0     0                   13858           1       1  15.500000       1   NaN   \n",
       "      1                   13858           1       1  21.100000       1   NaN   \n",
       "      2                   13858           1       1  26.600000       1   NaN   \n",
       "      3                   13858           1       1  32.200001       1   NaN   \n",
       "      4                   13858           1       1  37.799999       1   NaN   \n",
       "...                         ...         ...     ...        ...     ...   ...   \n",
       "1162  369                 39007           1       1        NaN           NaN   \n",
       "      370                 39007           1       1        NaN           NaN   \n",
       "      371                 39007           1       1        NaN           NaN   \n",
       "      372                 39007           1       1        NaN           NaN   \n",
       "      373                 39007           1       1        NaN           NaN   \n",
       "\n",
       "               psal_qc       temp temp_qc  \n",
       "entry subentry                             \n",
       "0     0                 21.804001       1  \n",
       "      1                 21.788000       1  \n",
       "      2                 21.521000       1  \n",
       "      3                 20.671000       1  \n",
       "      4                 19.691000       1  \n",
       "...                ...        ...     ...  \n",
       "1162  369                     NaN          \n",
       "      370                     NaN          \n",
       "      371                     NaN          \n",
       "      372                     NaN          \n",
       "      373                     NaN          \n",
       "\n",
       "[316608 rows x 16 columns]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "make_pandas([0, 1, 2], selection_function)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "56afb02a-cb6d-42be-b6e3-22f8a31d7e04",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "254"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_row_groups = ak.metadata_from_parquet(\n",
    "    \"s3://pivarski-princeton/argo-floats-expert.parquet\"\n",
    ").metadata.num_row_groups\n",
    "num_row_groups"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18a47c45-fdad-413a-ac56-d7d1781c0bc6",
   "metadata": {},
   "source": [
    "There are 254 row groups in this dataset. How long does it take to turn them all into Xarray?\n",
    "\n",
    "It depends heavily on network. Downloading the 7.0 GB file and having tens of GB of RAM can make it pretty quick.\n",
    "\n",
    "Dask can make it take advantage of multiple threads or computers in a cluster, and [that's under development](https://github.com/ContinuumIO/dask-awkward/)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "5fa921a1-8dfd-4bda-ab0c-82d1e9ccf9df",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n",
      "/home/jpivarski/mambaforge/lib/python3.9/site-packages/awkward/_v2/_connect/numpy.py:193: RuntimeWarning: invalid value encountered in fraction_in_land_ufunc\n",
      "  result = getattr(ufunc, method)(*args, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6min 57s ± 0 ns per loop (mean ± std. dev. of 1 run, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "%%timeit -r 1 -n 1\n",
    "\n",
    "for start in range(0, num_row_groups, 5):\n",
    "    make_pandas(range(start, min(start + 5, num_row_groups)), selection_function).reset_index().to_xarray()"
   ]
  }
 ],
 "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}

land.png