{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0",
   "metadata": {},
   "source": [
    "# More climatology reductions using Cubed\n",
    "\n",
    "This is the Cubed equivalent of [More climatology reductions](climatology-hourly.ipynb).\n",
    "\n",
    "The task is to compute an hourly climatology from an hourly dataset with 744 hours in each chunk, using the \"map-reduce\" strategy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cubed\n",
    "import cubed.array_api as xp\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import xarray as xr\n",
    "\n",
    "import flox.xarray"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2",
   "metadata": {},
   "source": [
    "## Create data\n",
    "\n",
    "Note that we use fewer lat/long points so the computation can be run locally."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3",
   "metadata": {},
   "outputs": [],
   "source": [
    "spec = cubed.Spec(allowed_mem=\"2GB\")\n",
    "ds = xr.Dataset(\n",
    "    {\n",
    "        \"tp\": (\n",
    "            (\"time\", \"latitude\", \"longitude\"),\n",
    "            xp.ones((8760, 72, 144), chunks=(744, 5, 144), dtype=np.float32, spec=spec),\n",
    "        )\n",
    "    },\n",
    "    coords={\"time\": pd.date_range(\"2021-01-01\", \"2021-12-31 23:59\", freq=\"h\")},\n",
    ")\n",
    "ds"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4",
   "metadata": {},
   "source": [
    "## Computation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5",
   "metadata": {},
   "outputs": [],
   "source": [
    "hourly = flox.xarray.xarray_reduce(ds.tp, ds.time.dt.hour, func=\"mean\", reindex=True)\n",
    "hourly"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6",
   "metadata": {},
   "outputs": [],
   "source": [
    "hourly.compute()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7",
   "metadata": {},
   "source": [
    "## Other climatologies: resampling by month\n",
    "\n",
    "This uses the \"blockwise\" strategy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8",
   "metadata": {},
   "outputs": [],
   "source": [
    "monthly = ds.tp.resample(time=\"ME\").sum(method=\"blockwise\")\n",
    "monthly"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9",
   "metadata": {},
   "outputs": [],
   "source": [
    "monthly.compute()"
   ]
  }
 ],
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}