Figure2_and_3_extreme_precip_connections.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Figures 2, 3\n",
    "\n",
    "## Figure 2\n",
    "Summary statistics from the experiments. Lines connect pairs of simulations to highlight the changes over oceans. Panels a and b separately show ocean gridpoints (filled markers) and all gridpoints (open markers)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import namedtuple\n",
    "import xarray as xr\n",
    "import numpy as np\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "from pathlib import Path\n",
    "# from sklearn.neighbors import KernelDensity\n",
    "import colorcet as cc\n",
    "from pathlib import Path\n",
    "import esmlab\n",
    "import scipy.stats as stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_colors = {}\n",
    "# for i, m in enumerate([\"cesm\", \"cnrm\", \"ipsl\", \"mri\", \"hadgem\"]):\n",
    "#     model_colors[m] = cc.cm.glasbey_hv(i)\n",
    "for i, m in [(0, \"cesm\"), (1, \"cnrm\"), (2, \"ipsl\"), (3, \"mri\"), (6, \"hadgem\")]:\n",
    "    model_colors[m] = cc.cm.glasbey_hv(i) \n",
    "\n",
    "# hard code locking to other blues, assumes CESM is first in list.\n",
    "model_colors['f-lock'] = cc.cm.glasbey_hv(5)\n",
    "model_colors['f-cntl'] = model_colors['f-lock']\n",
    "model_colors['c-lock'] = cc.cm.glasbey_hv(8)\n",
    "model_colors['c-cntl'] = model_colors['c-lock']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def quick_load_var(loc, var, table, model, experiment, member):\n",
    "    fils = sorted(\n",
    "        list(Path(loc).glob(\"_\".join([var, table, model, experiment, member, \"*.nc\"])))\n",
    "    )\n",
    "    if len(fils) == 0:\n",
    "        raise IOError(\"Can not find the files.\")\n",
    "    elif len(fils) == 1:\n",
    "        ds = xr.open_dataset(fils[0])\n",
    "    else:\n",
    "        ds = xr.open_mfdataset(fils, combine=\"by_coords\")\n",
    "    return ds[var].compute()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_loc = Path(\"/Volumes/Jedha/CMIP6\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "## do these:\n",
    "\n",
    "sims = {\"cesm-amip\": (\"CESM2\", \"amip\", \"r1i1p1f1\"),\n",
    "        \"cesm-amip-off\": (\"CESM2\",\"amip-lwoff\",\"r1i1p1f1\"),\n",
    "        \"cesm-aqua\": (\"CESM2\", \"aqua-control\", \"r1i1p1f1\"),\n",
    "        \"cesm-aqua-off\": (\"CESM2\", \"aqua-control-lwoff\", \"r1i1p1f1\"),\n",
    "        \"ipsl-amip\": (\"IPSL-CM6A-LR\", \"amip\", \"r1i1p1f1\"),\n",
    "        \"ipsl-amip-off\": (\"IPSL-CM6A-LR\", \"amip-lwoff\", \"r1i1p1f1\"),\n",
    "        \"ipsl-aqua\": (\"IPSL-CM6A-LR\", \"aqua-control\", \"r1i1p1f1\"),\n",
    "        \"ipsl-aqua-off\": (\"IPSL-CM6A-LR\", \"aqua-control-lwoff\", \"r1i1p1f1\"),\n",
    "        \"cnrm-amip\": (\"CNRM-CM6-1\", \"amip\", \"r1i1p1f2\"),\n",
    "        \"cnrm-amip-off\": (\"CNRM-CM6-1\", \"amip-lwoff\", \"r1i1p1f2\"),\n",
    "        \"cnrm-aqua\": (\"CNRM-CM6-1\", \"aqua-control\", \"r1i1p1f2\"),\n",
    "        \"cnrm-aqua-off\": (\"CNRM-CM6-1\", \"aqua-control-lwoff\", \"r1i1p1f2\"),\n",
    "        \"mri-amip\" : (\"MRI-ESM2-0\", \"amip\", \"r1i1p1f1\"),\n",
    "        \"mri-amip-off\" : (\"MRI-ESM2-0\", \"amip-lwoff\", \"r1i1p1f1\"),\n",
    "        \"hadgem-amip\" : (\"HadGEM3-GC31-LL\", \"amip\", \"r1i1p1f3\"),\n",
    "        \"hadgem-amip-off\" : (\"HadGEM3-GC31-LL\", \"amip-lwoff\", \"r5i1p1f2\"),\n",
    "        \"hadgem-aqua\" : (\"HadGEM3-GC31-LL\", \"aqua-control\", \"r1i1p1f3\"),\n",
    "        \"hadgem-aqua-off\" : (\"HadGEM3-GC31-LL\", \"aqua-control-lwoff\", \"r1i1p1f3\")\n",
    "       }\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "working on cesm-amip\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brianpm/miniconda3/envs/py39/lib/python3.9/site-packages/xarray/conventions.py:492: SerializationWarning: variable 'pr' has multiple fill values {1e+20, 1e+20}, decoding all values to NaN.\n",
      "  new_vars[k] = decode_cf_variable(\n",
      "/Users/brianpm/miniconda3/envs/py39/lib/python3.9/site-packages/xarray/conventions.py:492: SerializationWarning: variable 'sftlf' has multiple fill values {1e+20, 1e+20}, decoding all values to NaN.\n",
      "  new_vars[k] = decode_cf_variable(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "working on cesm-amip-off\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brianpm/miniconda3/envs/py39/lib/python3.9/site-packages/xarray/conventions.py:492: SerializationWarning: variable 'pr' has multiple fill values {1e+20, 1e+20}, decoding all values to NaN.\n",
      "  new_vars[k] = decode_cf_variable(\n",
      "/Users/brianpm/miniconda3/envs/py39/lib/python3.9/site-packages/xarray/conventions.py:492: SerializationWarning: variable 'sftlf' has multiple fill values {1e+20, 1e+20}, decoding all values to NaN.\n",
      "  new_vars[k] = decode_cf_variable(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "working on cesm-aqua\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brianpm/miniconda3/envs/py39/lib/python3.9/site-packages/xarray/conventions.py:492: SerializationWarning: variable 'pr' has multiple fill values {1e+20, 1e+20}, decoding all values to NaN.\n",
      "  new_vars[k] = decode_cf_variable(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "working on cesm-aqua-off\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brianpm/miniconda3/envs/py39/lib/python3.9/site-packages/xarray/conventions.py:492: SerializationWarning: variable 'pr' has multiple fill values {1e+20, 1e+20}, decoding all values to NaN.\n",
      "  new_vars[k] = decode_cf_variable(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "working on ipsl-amip\n",
      "working on ipsl-amip-off\n",
      "working on ipsl-aqua\n",
      "working on ipsl-aqua-off\n",
      "working on cnrm-amip\n",
      "working on cnrm-amip-off\n",
      "working on cnrm-aqua\n",
      "working on cnrm-aqua-off\n",
      "working on mri-amip\n",
      "working on mri-amip-off\n",
      "working on hadgem-amip\n",
      "working on hadgem-amip-off\n",
      "working on hadgem-aqua\n",
      "working on hadgem-aqua-off\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'cesm-amip': Result(case='CESM2_amip_r1i1p1f1', gavg=3.335214581376049e-05, goavg=3.510287202056346e-05, tavg=3.972882151858157e-05, toavg=3.810103323664371e-05, toavg_uw=3.799505066126585e-05, p95=0.00016276335591101053, tsum=11771.402, xsum=4660.9253),\n",
       " 'cesm-amip-off': Result(case='CESM2_amip-lwoff_r1i1p1f1', gavg=3.1898189656810654e-05, goavg=3.6691127216183814e-05, tavg=4.0337348532039076e-05, toavg=4.451005603576896e-05, toavg_uw=4.439959229785018e-05, p95=0.00017272413679165783, tsum=7618.756, xsum=2492.2173),\n",
       " 'cesm-aqua': Result(case='CESM2_aqua-control_r1i1p1f1', gavg=3.8199234236945356e-05, goavg=3.8199234236945356e-05, tavg=4.669582324942678e-05, toavg=4.669582324942678e-05, toavg_uw=4.54123510280624e-05, p95=0.00019911886047339054, tsum=3056.0347, xsum=1136.4103),\n",
       " 'cesm-aqua-off': Result(case='CESM2_aqua-control-lwoff_r1i1p1f1', gavg=3.29666995469341e-05, goavg=3.29666995469341e-05, tavg=4.349748845682736e-05, toavg=4.349748845682736e-05, toavg_uw=4.274130697012879e-05, p95=0.0001470438000978902, tsum=2876.2861, xsum=747.7706),\n",
       " 'ipsl-amip': Result(case='IPSL-CM6A-LR_amip_r1i1p1f1', gavg=3.524060593917966e-05, goavg=3.637862391769886e-05, tavg=4.1667939512990415e-05, toavg=3.897442002198659e-05, toavg_uw=3.8684163882862777e-05, p95=0.0001171608862932771, tsum=2565.1653, xsum=1002.51984),\n",
       " 'ipsl-amip-off': Result(case='IPSL-CM6A-LR_amip-lwoff_r1i1p1f1', gavg=3.3550342777743936e-05, goavg=3.5845834645442665e-05, tavg=4.2562394810374826e-05, toavg=4.228220132063143e-05, toavg_uw=4.203513890388422e-05, p95=0.00011516892263898617, tsum=2572.9119, xsum=617.3226),\n",
       " 'ipsl-aqua': Result(case='IPSL-CM6A-LR_aqua-control_r1i1p1f1', gavg=3.74919873138424e-05, goavg=3.74919873138424e-05, tavg=4.3372743675718084e-05, toavg=4.3372743675718084e-05, toavg_uw=4.2379600927233696e-05, p95=0.00012036662592436183, tsum=1135.8275, xsum=410.48297),\n",
       " 'ipsl-aqua-off': Result(case='IPSL-CM6A-LR_aqua-control-lwoff_r1i1p1f1', gavg=3.4100739867426455e-05, goavg=3.4100739867426455e-05, tavg=4.408869062899612e-05, toavg=4.408869062899612e-05, toavg_uw=4.3223772081546485e-05, p95=0.0001085830233932938, tsum=1158.4524, xsum=227.93706),\n",
       " 'cnrm-amip': Result(case='CNRM-CM6-1_amip_r1i1p1f2', gavg=3.4340141726356325e-05, goavg=3.878288835097966e-05, tavg=4.120439857927485e-05, toavg=4.4454349775372726e-05, toavg_uw=4.4503580284072086e-05, p95=0.00024274253737530658, tsum=4226.1523, xsum=2056.1082),\n",
       " 'cnrm-amip-off': Result(case='CNRM-CM6-1_amip-lwoff_r1i1p1f2', gavg=3.259602454775244e-05, goavg=4.015651122194292e-05, tavg=4.108919110385553e-05, toavg=5.005037512698122e-05, toavg_uw=4.999546581530012e-05, p95=0.00025542627554386854, tsum=4747.6733, xsum=1996.4293),\n",
       " 'cnrm-aqua': Result(case='CNRM-CM6-1_aqua-control_r1i1p1f2', gavg=3.6850641451165745e-05, goavg=3.6850641451165745e-05, tavg=4.517692816840705e-05, toavg=4.517692816840705e-05, toavg_uw=4.40364601672627e-05, p95=0.00022888192688697018, tsum=1729.6226, xsum=756.9929),\n",
       " 'cnrm-aqua-off': Result(case='CNRM-CM6-1_aqua-control-lwoff_r1i1p1f2', gavg=3.407762387203342e-05, goavg=3.407762387203342e-05, tavg=4.5323695443411205e-05, toavg=4.5323695443411205e-05, toavg_uw=4.438329779077321e-05, p95=0.00021453136650961824, tsum=1743.2452, xsum=737.55414),\n",
       " 'mri-amip': Result(case='MRI-ESM2-0_amip_r1i1p1f1', gavg=3.467894547813345e-05, goavg=3.666483117623455e-05, tavg=4.065503786469254e-05, toavg=3.982882838669824e-05, toavg_uw=3.966895383200608e-05, p95=0.000198718205501791, tsum=6338.0474, xsum=2632.849),\n",
       " 'mri-amip-off': Result(case='MRI-ESM2-0_amip-lwoff_r1i1p1f1', gavg=3.332971709487176e-05, goavg=3.689531949552828e-05, tavg=4.1905609217658335e-05, toavg=4.346558858779948e-05, toavg_uw=4.329719740781002e-05, p95=0.0001950505262357182, tsum=6917.7446, xsum=2278.846),\n",
       " 'hadgem-amip': Result(case='HadGEM3-GC31-LL_amip_r1i1p1f3', gavg=3.6221579488211625e-05, goavg=4.109380211574865e-05, tavg=4.3641184962569774e-05, toavg=4.723158428959843e-05, toavg_uw=4.7029840061441064e-05, p95=0.00024728564312681556, tsum=3898.405, xsum=1872.7053),\n",
       " 'hadgem-amip-off': Result(case='HadGEM3-GC31-LL_amip-lwoff_r5i1p1f2', gavg=3.4926754170444035e-05, goavg=4.117807240912427e-05, tavg=4.423388130348245e-05, toavg=5.031238203693338e-05, toavg_uw=5.012338806409389e-05, p95=0.00023647405541851155, tsum=4154.836, xsum=1617.2174),\n",
       " 'hadgem-aqua': Result(case='HadGEM3-GC31-LL_aqua-control_r1i1p1f3', gavg=3.8657482254848346e-05, goavg=3.8657482254848346e-05, tavg=4.5891523348801204e-05, toavg=4.5891523348801204e-05, toavg_uw=4.4626394810620695e-05, p95=0.00022948442056076575, tsum=1480.5967, xsum=629.43207),\n",
       " 'hadgem-aqua-off': Result(case='HadGEM3-GC31-LL_aqua-control-lwoff_r1i1p1f3', gavg=3.613022467591852e-05, goavg=3.613022467591852e-05, tavg=4.6079605154802934e-05, toavg=4.6079605154802934e-05, toavg_uw=4.498550697462633e-05, p95=0.00019454662324278602, tsum=1492.5111, xsum=496.61697)}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for each simulation, I want to get:\n",
    "# - global mean precip\n",
    "# - global mean precip, ocean only\n",
    "# - tropical mean precip\n",
    "# - tropical mean precip, ocean only\n",
    "# - tropical UNWEIGHTED mean precip, ocean only\n",
    "# - 95th percentile of tropical ocean precip\n",
    "# - sum of tropical ocean precip with p >= p95\n",
    "# - sum of tropical ocean precip\n",
    "# Define a namedtuple:\n",
    "Result = namedtuple('Result', ['case', 'gavg', 'goavg', 'tavg', 'toavg', 'toavg_uw', 'p95', 'tsum', 'xsum'])\n",
    "\n",
    "\n",
    "def get_pr_info(dloc, model, experiment, member):\n",
    "    tropics = slice(-30,30)\n",
    "    pr = quick_load_var(data_loc, \"pr\", \"day\", model, experiment, member)\n",
    "    lat = pr['lat']\n",
    "    wgt = np.cos(np.radians(lat))\n",
    "    gavg = esmlab.weighted_mean(pr.mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt).item()\n",
    "    tavg = esmlab.weighted_mean(pr.sel(lat=tropics).mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt.sel(lat=tropics)).item()\n",
    "    if 'amip' in c:\n",
    "        land = quick_load_var(data_loc, \"sftlf\", \"fx\", model, experiment, member)\n",
    "        land_trop = land.sel(lat=tropics)\n",
    "        msk, _ = xr.broadcast(land, pr)\n",
    "        pr = pr.where(msk <= 0)\n",
    "        goavg = esmlab.weighted_mean(pr.mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt).item()\n",
    "        toavg = esmlab.weighted_mean(pr.sel(lat=tropics).mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt.sel(lat=tropics)).item()\n",
    "    else:\n",
    "        goavg = gavg\n",
    "        toavg = tavg\n",
    "    ptrop = pr.sel(lat=tropics)\n",
    "    toavg_uw = ptrop.mean().item()\n",
    "    p95 = np.nanquantile(ptrop.values.flatten(), 0.95)\n",
    "    tsum = np.nansum(ptrop.values.flatten())\n",
    "    xsum = np.nansum(ptrop.where(ptrop>=p95).values.flatten())\n",
    "    return Result(\"_\".join([model,experiment,member]), gavg, goavg, tavg, toavg, toavg_uw, p95, tsum, xsum)\n",
    "\n",
    "\n",
    "R = {}\n",
    "for c in sims:\n",
    "    print(f\"working on {c}\")\n",
    "    R[c] = get_pr_info(data_loc, *sims[c])\n",
    "\n",
    "R"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "f-lock\n",
      "f-cntl\n",
      "c-lock\n",
      "c-cntl\n"
     ]
    }
   ],
   "source": [
    "# Add the locking results:\n",
    "variable = \"PRECT\"\n",
    "cases = {\n",
    "    'f-lock': '/Volumes/Samsung_T5/F1850JJB_c201_CLOCK.cam.h2.ncrcat.PRECT.nc',\n",
    "    'f-cntl': '/Volumes/Samsung_T5/F1850JJB_c201_CTL.cam.h2.ncrcat.PRECT.nc',\n",
    "    'c-lock': '/Volumes/Samsung_T5/B1850_c201_CLOCK/daily/B1850_c201_CLOCK.cam.h2.ncrcat.PRECT.nc',\n",
    "    'c-cntl': '/Volumes/Samsung_T5/B1850_c201_CTL/daily/B1850_c201_CTL.cam.h2.ncrcat.PRECT.nc'}\n",
    "\n",
    "case_names = {'f-lock': 'F1850JJB_c201_CLOCK',\n",
    "    'f-cntl': 'F1850JJB_c201_CTL',\n",
    "    'c-lock': 'B1850_c201_CLOCK',\n",
    "    'c-cntl': 'B1850_c201_CTL'}\n",
    "\n",
    "# f-cntl ntime = 9126\n",
    "# f-lock ntime = 9126\n",
    "# c-cntl ntime = 10982\n",
    "# c-lock ntime = 9491\n",
    "\n",
    "\n",
    "# LAND\n",
    "land_file = xr.open_dataset(\"/Users/brianpm/Dropbox/Data/cesm2_f09_land.nc\")\n",
    "LAND = land_file['LANDFRAC'].squeeze()\n",
    "\n",
    "def get_cesm_pr_info(fil, experiment, land=None):\n",
    "    tropics = slice(-30,30)\n",
    "    pr_ds = xr.open_dataset(fil)\n",
    "    pr = pr_ds['PRECT']*1000. # converts to same units as CMIP6 pr\n",
    "    lat = pr['lat']\n",
    "    wgt = np.cos(np.radians(lat))\n",
    "    gavg = esmlab.weighted_mean(pr.mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt).item()\n",
    "    tavg = esmlab.weighted_mean(pr.sel(lat=tropics).mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt.sel(lat=tropics)).item()\n",
    "    if land is not None:\n",
    "        land_trop = land.sel(lat=tropics)\n",
    "        msk, _ = xr.broadcast(land, pr)\n",
    "        pr = pr.where(msk <= 0)\n",
    "        goavg = esmlab.weighted_mean(pr.mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt).item()\n",
    "        toavg = esmlab.weighted_mean(pr.sel(lat=tropics).mean(dim=('time','lon')), dim=[\"lat\"], weights=wgt.sel(lat=tropics)).item()\n",
    "    else:\n",
    "        goavg = gavg\n",
    "        toavg = tavg\n",
    "    ptrop = pr.sel(lat=tropics)\n",
    "    toavg_uw = ptrop.mean().item()\n",
    "    p95 = np.nanquantile(ptrop.values.flatten(), 0.95)\n",
    "    tsum = np.nansum(ptrop.values.flatten())\n",
    "    xsum = np.nansum(ptrop.where(ptrop>=p95).values.flatten())\n",
    "    return Result(experiment, gavg, goavg, tavg, toavg, toavg_uw, p95, tsum, xsum)\n",
    "\n",
    "for c in cases:\n",
    "    print(c)\n",
    "    R[c] = get_cesm_pr_info(cases[c], c, land=LAND)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# ensure correct order:\n",
    "order = ['f-cntl', 'f-lock', 'c-cntl', 'c-lock',\n",
    "'cesm-amip',\n",
    "'cesm-amip-off',\n",
    "'cesm-aqua',\n",
    "'cesm-aqua-off',\n",
    "'cnrm-amip',\n",
    "'cnrm-amip-off',\n",
    "'cnrm-aqua',\n",
    "'cnrm-aqua-off',\n",
    "'hadgem-amip', 'hadgem-amip-off', 'hadgem-aqua', 'hadgem-aqua-off',\n",
    "'ipsl-amip',\n",
    "'ipsl-amip-off',\n",
    "'ipsl-aqua',\n",
    "'ipsl-aqua-off',\n",
    "'mri-amip',\n",
    "'mri-amip-off',\n",
    "]\n",
    "RR = {r:R[r] for r in order}\n",
    "# RR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x648 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Now we have R which can be used to make a summary plot:\n",
    "# R is a dict of namedtuples ... so to access: R[case].field\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(5,9), nrows=4, sharex=True, constrained_layout=True)\n",
    "xval = 1\n",
    "for i, c in enumerate(RR):\n",
    "    pieces = c.split(\"-\")\n",
    "    if c not in ['f-cntl', 'f-lock', 'c-cntl', 'c-lock']:\n",
    "        model = pieces[0]\n",
    "    else:\n",
    "        model = c\n",
    "    if ('amip' in c):\n",
    "        m = '^'\n",
    "    elif (c in ['f-cntl', 'f-lock', 'c-cntl', 'c-lock']):\n",
    "        m = \"s\"\n",
    "    else:\n",
    "        m = 'o'\n",
    "#     print(c, model)\n",
    "    mfill = 'full'\n",
    "    mafill = 'none'\n",
    "\n",
    "    if ('amip' in c) or (c in ['f-cntl', 'f-lock', 'c-cntl', 'c-lock']):\n",
    "        ax[0].plot(xval, RR[c].gavg*86400.0, marker=m, color=model_colors[model], fillstyle=mafill)\n",
    "        ax[1].plot(xval, RR[c].tavg*86400.0, marker=m, color=model_colors[model], fillstyle=mafill)\n",
    "    ax[0].plot(xval, RR[c].goavg*86400.0, marker=m, color=model_colors[model], fillstyle=mfill)\n",
    "    ax[1].plot(xval, RR[c].toavg*86400.0, marker=m, color=model_colors[model], fillstyle=mfill)\n",
    "    #ax[1].plot(xval, RR[c].toavg_uw*86400.0, marker='*', color=model_colors[model], fillstyle=mfill)\n",
    "    ax[2].plot(xval, RR[c].p95*86400.0, marker=m, fillstyle=mfill, color=model_colors[model])\n",
    "    ax[3].plot(xval, RR[c].xsum/R[c].tsum, marker=m, fillstyle=mfill, color=model_colors[model])\n",
    "    # ADD A LINE CONNECTING PAIRS\n",
    "    if i % 2 == 1:\n",
    "        ax[0].plot( [xval-1, xval], [RR[order[i-1]].goavg*86400.0, RR[order[i]].goavg*86400.0], color=model_colors[model], linewidth=1)\n",
    "        ax[1].plot( [xval-1, xval], [RR[order[i-1]].toavg*86400.0, RR[order[i]].toavg*86400.0], color=model_colors[model], linewidth=1)\n",
    "        ax[2].plot( [xval-1, xval], [RR[order[i-1]].p95*86400.0, RR[order[i]].p95*86400.0], color=model_colors[model], linewidth=1)\n",
    "        ax[3].plot( [xval-1, xval], [RR[order[i-1]].xsum/RR[order[i-1]].tsum, RR[order[i]].xsum/RR[order[i]].tsum], color=model_colors[model], linewidth=1)\n",
    "    xval += 1\n",
    "\n",
    "ax[0].set_title(\"Global Mean Precipitation (mm d$^{-1}$)\", loc='left')\n",
    "ax[1].set_title(\"Tropical Mean Precipitation (mm d$^{-1}$)\", loc='left')\n",
    "ax[2].set_title(\"Tropical Ocean Precipitation 95$^{\\mathrm{th}}$ Percentile (mm d$^{-1}$)\", loc='left')\n",
    "ax[3].set_title(\"Tropical Ocean Extreme Fraction $\\\\left( \\\\varphi \\\\right)$\", loc='left')\n",
    "ax[-1].set_xticks(np.arange(1, len(RR)+1))\n",
    "ax[-1].set_xticklabels(RR.keys(), rotation=90)\n",
    "\n",
    "[a.spines['top'].set_visible(False) for a in ax]\n",
    "[a.spines['right'].set_visible(False) for a in ax]\n",
    "#\n",
    "\n",
    "    \n",
    "# FOR FINAL VERSION ADD PANEL LABELS:\n",
    "axr = ax.ravel()\n",
    "for i, label in enumerate(('a', 'b', 'c', 'd')):\n",
    "    addto = axr[i]\n",
    "    addto.text(-0.03, 1.18, label, transform=addto.transAxes,\n",
    "      fontsize=16, fontweight='bold', va='top', ha='right')\n",
    "\n",
    "\n",
    "#\n",
    "from matplotlib.lines import Line2D\n",
    "legend_elements_last = [Line2D([0], [0], marker=\"^\", linestyle=\"None\", fillstyle='full', color=\"darkgray\", label='AMIP Ocean Gridpoints'),\n",
    "                        Line2D([0], [0], marker=\"^\", linestyle=\"None\", fillstyle='none', color=\"darkgray\", label='AMIP All Gridpoints'),            \n",
    "                        Line2D([0], [0], marker=\"o\", linestyle=\"None\", fillstyle='full', color=\"darkgray\", label='Aqua'),\n",
    "                        Line2D([0], [0], marker=\"s\", linestyle=\"None\", fillstyle='full', color=\"darkgray\", label='Locking Ocean Gridpoints'),\n",
    "                        Line2D([0], [0], marker=\"s\", linestyle=\"None\", fillstyle='none', color=\"darkgray\", label='Locking All Gridpoints'),\n",
    "                  ]\n",
    "\n",
    "fig.legend(handles=legend_elements_last, loc='lower center', bbox_to_anchor=(.5, -0.09), frameon=True, ncol=2)\n",
    "\n",
    "# plotname = Path(\"/Users/brianpm/Dropbox/Manuscripts/M_cre_pex/summary_stats_v4.pdf\")\n",
    "plotname = Path(\"/Users/brianpm/Dropbox/Apps/Overleaf/cre_pex/Figure_2_final.pdf\")\n",
    "OVERWRITE = True\n",
    "if (not plotname.exists()) or OVERWRITE:\n",
    "    fig.savefig(plotname, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Figure 3\n",
    "The fractional change in extreme precipitation fraction ($\\delta \\varphi$) against tropical ocean mean precipitation ($\\delta P$). Dashed diagonal line in upper left shows the one-to-one line, vertical line at $\\delta P = 0$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('f-cntl', 'f-lock'), ('c-cntl', 'c-lock'), ('cesm-amip', 'cesm-amip-off'), ('cesm-aqua', 'cesm-aqua-off'), ('cnrm-amip', 'cnrm-amip-off'), ('cnrm-aqua', 'cnrm-aqua-off'), ('hadgem-amip', 'hadgem-amip-off'), ('hadgem-aqua', 'hadgem-aqua-off'), ('ipsl-amip', 'ipsl-amip-off'), ('ipsl-aqua', 'ipsl-aqua-off'), ('mri-amip', 'mri-amip-off')]\n",
      "Model is f-cntl : Compare RR[f-lock] to RR[f-cntl] ... deltaMean = -0.02200096350921898\n",
      "Model is c-cntl : Compare RR[c-lock] to RR[c-cntl] ... deltaMean = -0.017941238008637065\n",
      "Model is cesm : Compare RR[cesm-amip-off] to RR[cesm-amip] ... deltaMean = 0.16821125976608337\n",
      "Model is cesm : Compare RR[cesm-aqua-off] to RR[cesm-aqua] ... deltaMean = -0.0684929522607502\n",
      "Model is cnrm : Compare RR[cnrm-amip-off] to RR[cnrm-amip] ... deltaMean = 0.12588251498188904\n",
      "Model is cnrm : Compare RR[cnrm-aqua-off] to RR[cnrm-aqua] ... deltaMean = 0.0032487218798287077\n",
      "Model is hadgem : Compare RR[hadgem-amip-off] to RR[hadgem-amip] ... deltaMean = 0.06522749117296536\n",
      "Model is hadgem : Compare RR[hadgem-aqua-off] to RR[hadgem-aqua] ... deltaMean = 0.00409839971038231\n",
      "Model is ipsl : Compare RR[ipsl-amip-off] to RR[ipsl-amip] ... deltaMean = 0.08487057143579882\n",
      "Model is ipsl : Compare RR[ipsl-aqua-off] to RR[ipsl-aqua] ... deltaMean = 0.0165068403011556\n",
      "Model is mri : Compare RR[mri-amip-off] to RR[mri-amip] ... deltaMean = 0.09130974594060176\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# this just gets the pairs of simulations to form \"experiments\"\n",
    "\n",
    "epair = []\n",
    "for i, c in enumerate(RR):\n",
    "    m = RR[c].case[0:4]\n",
    "    if i%2 == 1:\n",
    "        epair.append((prior_case, c))\n",
    "    else:\n",
    "        prior_case = c\n",
    "print(epair)\n",
    "# now for each experiment, I want to get:\n",
    "# tavg[1]-tavg[0] / tavg[0]\n",
    "# xfrac[1]-xfrac[0]/xfrac[0]\n",
    "with plt.style.context('ggplot'):\n",
    "    fig, ax = plt.subplots(nrows=1, sharex=True, constrained_layout=True)\n",
    "    xval = 1\n",
    "    for c in epair:\n",
    "        pieces = c[0].split(\"-\")\n",
    "        if c[0] not in ['f-cntl', 'f-lock', 'c-cntl', 'c-lock']:\n",
    "            model = pieces[0]\n",
    "        else:\n",
    "            model = c[0]\n",
    "        if 'aqua' in c[0]:\n",
    "            m = 'o'\n",
    "        elif c[0] in ['f-cntl', 'f-lock', 'c-cntl', 'c-lock']:\n",
    "            m = \"s\"\n",
    "        else:\n",
    "            m = '^'\n",
    "        mfill = 'full'\n",
    "        mafill = 'none'\n",
    "        delta_mean = ((RR[c[1]].toavg - RR[c[0]].toavg)/RR[c[0]].toavg)\n",
    "        xfrac0 = RR[c[0]].xsum/RR[c[0]].tsum\n",
    "        xfrac1 = RR[c[1]].xsum/RR[c[1]].tsum\n",
    "        delta_xfrac = (xfrac1-xfrac0)/xfrac0\n",
    "        print(f\"Model is {model} : Compare RR[{c[1]}] to RR[{c[0]}] ... deltaMean = {delta_mean}\")\n",
    "\n",
    "\n",
    "    #     ax[0].plot(xval, delta_mean, marker=m, color='tab:blue', fillstyle=mafill)\n",
    "    #     ax[1].plot(xval, delta_xfrac, marker=m, color='tab:blue', fillstyle=mafill)\n",
    "    #     xval += 1\n",
    "        ax.plot(delta_mean, delta_xfrac, marker=m, color=model_colors[model], fillstyle=mfill, markersize=10)\n",
    "\n",
    "    ax.plot([-1, 1], [-1, 1], linestyle=\":\", color='gray')\n",
    "    ax.set_xlabel(\"$\\u03b4 P$\", loc='center', fontsize=18)\n",
    "    ax.set_ylabel(\"$\\u03b4 \\u03c6$\", loc='center', fontsize=18)\n",
    "    ax.spines['top'].set_visible(False)\n",
    "    ax.spines['right'].set_visible(False)\n",
    "    ax.set_xlim([-0.25, 0.25])\n",
    "    ax.set_ylim([-0.5, 0.])\n",
    "    ax.set_aspect('equal')\n",
    "\n",
    "\n",
    "    from matplotlib.lines import Line2D\n",
    "\n",
    "    legend_elements = [Line2D([0], [0], marker=\"None\", linestyle=\"None\", color=model_colors['cesm'], label='CESM2'),\n",
    "                       Line2D([0], [0], marker=\"None\", linestyle=\"None\", color=model_colors['cnrm'], label='CNRM'),\n",
    "                       Line2D([0], [0], marker=\"None\", linestyle=\"None\", color=model_colors['ipsl'], label='IPSL'),\n",
    "                       Line2D([0], [0], marker=\"None\", linestyle=\"None\", color=model_colors['mri'], label='MRI'),\n",
    "                       Line2D([0], [0], marker='s', color='gray', linestyle=\"None\", label='Locking'),\n",
    "                       Line2D([0], [0], marker='^', color='gray', linestyle=\"None\", label='AMIP LWoff'),\n",
    "                       Line2D([0], [0], marker='o', color='gray', linestyle=\"None\", label='AQUA LWoff'),\n",
    "\n",
    "                      ]\n",
    "\n",
    "    ax.legend(handles=legend_elements, loc=(0.02, 0.425), labelcolor='linecolor', facecolor=[1, 1, 1, 0])\n",
    "    ax.axvline(0.0, linestyle=':', color='gray')\n",
    "    plotname = Path(\"/Users/brianpm/Dropbox/Manuscripts/M_cre_pex/mean_versus_extreme_v3.pdf\")\n",
    "    if not plotname.exists():\n",
    "        fig.savefig(plotname, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extra work\n",
    "\n",
    "This section provides some additional analysis. \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4.669582324942678e-05\n",
      "4.349748845682736e-05\n",
      "-0.0684929522607502\n",
      "4.4454349775372726e-05\n",
      "5.005037512698122e-05\n",
      "0.12588251498188904\n"
     ]
    }
   ],
   "source": [
    "print(RR['cesm-aqua'].tavg)\n",
    "print(RR['cesm-aqua-off'].tavg)\n",
    "print((RR['cesm-aqua-off'].tavg - RR['cesm-aqua'].tavg)/RR['cesm-aqua'].tavg  )\n",
    "\n",
    "\n",
    "print(RR['cnrm-amip'].toavg)\n",
    "print(RR['cnrm-amip-off'].toavg)\n",
    "print((RR['cnrm-amip-off'].toavg - RR['cnrm-amip'].toavg)/RR['cnrm-amip'].toavg  )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brianpm/anaconda3/lib/python3.7/site-packages/xarray/conventions.py:494: SerializationWarning: variable 'sftlf' has multiple fill values {1e+20, 1e+20}, decoding all values to NaN.\n",
      "  use_cftime=use_cftime,\n"
     ]
    }
   ],
   "source": [
    "# for the amip runs, we need the land fraction:\n",
    "land = {}\n",
    "for s in sims:\n",
    "    if 'amip' in s:\n",
    "        land[s] = quick_load_var(data_loc, \"sftlf\", \"fx\", *sims[s])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove land points from the amip runs:\n",
    "for c in pr:\n",
    "    if 'amip' in c:\n",
    "        pr[c] = pr[c].where(land[c] <= 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_tropics_quantiles(da, threshold=None, q=0.99):\n",
    "    da_tropics = da.sel(lat=slice(-30, 30))  # SHOULD BE ADJUSTABLE\n",
    "    da_tropics_val = np.nanquantile(da_tropics, q)\n",
    "    if threshold is not None:\n",
    "        da_climo = da.mean(dim=\"time\")\n",
    "        da_process = np.where(\n",
    "            (da_climo >= threshold).sel(lat=slice(-30, 30)).broadcast_like(da_tropics),\n",
    "            da_tropics,\n",
    "            np.nan,\n",
    "        )\n",
    "        da_threshold_val = np.nanquantile(da_process, q)\n",
    "        print(\n",
    "            f\"Compare the tropics region ({da_tropics_val}) to the rainy tropics ({da_threshold_val})\"\n",
    "        )\n",
    "        return da_threshold_val\n",
    "    else:\n",
    "        print(f\"Quantile value: {da_tropics_val}\")\n",
    "        return da_tropics_val\n",
    "\n",
    "\n",
    "def conf_interval(data_series, **kwargs):\n",
    "    \"\"\"\n",
    "    Calculate the confidence interval for the data distribution under the assumptions that it can be calculated using \\\n",
    "    a student-t distribution.\n",
    "    \n",
    "    This function expects just a simple series. To apply to an array, we will need to adjust.\n",
    "    \n",
    "    :return conf_int: confidence interval, to apply around mean value\n",
    "    \"\"\"\n",
    "    if 'confidence_level' in kwargs and (kwargs['confidence_level'] is not None):\n",
    "        confidence_level = kwargs[\"confidence_level\"]\n",
    "    else:\n",
    "        confidence_level = 0.95\n",
    "        \n",
    "    n = len(data_series)\n",
    "    mean = np.mean(data_series)\n",
    "    # Allow a known \"population standard deviation\" to be passed in.\n",
    "    if 'known_std' in kwargs:\n",
    "        sem = kwargs['known_std'] / np.sqrt(n)\n",
    "    else:\n",
    "        sem = stats.sem(data_series)\n",
    "    \n",
    "    # The population standard deviation will be estimated by the sample standard deviation.\n",
    "    # Therefore Students t distribution is used as the critical value:\n",
    "    # confidence level: C\n",
    "    # degrees of freedom = (N - 1)\n",
    "    # critical value t* = t_α(r) where r is dof (N-1) and α = (1 - C)/2\n",
    "    # This follows the same form as Forster et al. 2016:\n",
    "    # CI = t_value * SE\n",
    "    # with SE = σ / sqrt[N]\n",
    "    conf_int = sem * stats.t.ppf((1 + confidence_level) / 2, n - 1)\n",
    "\n",
    "    # If you wanted to give the actual interval, it's just around the mean:\n",
    "    # start = mean - conf_int\n",
    "    # end = mean + conf_int\n",
    "\n",
    "    return conf_int\n",
    "\n",
    "\n",
    "def report_global_mean_with_ci(arr, doprint=False, confidence_level=None):\n",
    "    \"\"\"Calculate the global average and confidence interval for DataArray arr.\"\"\"\n",
    "    wgt = np.cos(np.radians(arr['lat']))\n",
    "    w, _ = xr.broadcast(wgt, arr)\n",
    "    global_average = esmlab.weighted_mean(arr, dim=[\"lat\",\"lon\"], weights=w)\n",
    "    ci = conf_interval(global_average, **{\"confidence_level\":confidence_level})\n",
    "    if doprint:\n",
    "        print(f\"The mean (95% ci): {global_average.mean().values.item()} \\u00B1 {ci}\")\n",
    "    return global_average, ci"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Quantile value: 0.0003849748443462888\n",
      "Quantile value: 0.0003567487851250904\n",
      "Quantile value: 0.00042201049131108427\n",
      "Quantile value: 0.00026625057216733694\n",
      "Quantile value: 0.00042169346357695406\n",
      "Quantile value: 0.00023998573160497524\n",
      "Quantile value: 0.00044758025964256433\n",
      "Quantile value: 0.0001830892966245307\n",
      "Quantile value: 0.0005445090576540693\n",
      "Quantile value: 0.000518587381811814\n",
      "Quantile value: 0.0004815489955944949\n",
      "Quantile value: 0.00047797238366911266\n"
     ]
    }
   ],
   "source": [
    "# get the 99th percentile of precipitation for tropical oceans\n",
    "pr99 = {}\n",
    "prAV = {}\n",
    "prAVci = {}\n",
    "for c in pr:\n",
    "    pr99[c] = get_tropics_quantiles(pr[c])\n",
    "    prAV[c], prAVci[c] = report_global_mean_with_ci(pr[c])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=2, sharex=True, constrained_layout=True)\n",
    "xval = 1\n",
    "for c in pr99:\n",
    "    if 'amip' in c:\n",
    "        m = 'o'\n",
    "    else:\n",
    "        m = '^'\n",
    "    ax[0].plot(xval, prAV[c].mean()*86400.0, marker=m, color='darkgray')\n",
    "    ax[1].plot(xval, pr99[c].mean()*86400.0, marker=m, color='darkgray')\n",
    "    xval += 1\n",
    "\n",
    "ax[0].set_title(\"Mean Precipitation (mm/day)\", loc='left')\n",
    "ax[1].set_title(\"99th Percentile (mm/day)\", loc='left')\n",
    "ax[1].set_xticks(np.arange(1, len(pr99)+1))\n",
    "ax[1].set_xticklabels(pr99.keys(), rotation=90)\n",
    "ax[0].spines['top'].set_visible(False)\n",
    "ax[0].spines['right'].set_visible(False)\n",
    "ax[1].spines['top'].set_visible(False)\n",
    "ax[1].spines['right'].set_visible(False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.0022061409212938516,\n",
       " 0.0028942716559749036,\n",
       " 0.004375811522960663,\n",
       " 0.0035248446633134273,\n",
       " 0.002688452406247731,\n",
       " 0.002457532420799608,\n",
       " 0.0035701220627936017,\n",
       " 0.002981181726090687,\n",
       " 0.0031770461478454336,\n",
       " 0.0029570117492605436,\n",
       " 0.004094379932565495,\n",
       " 0.0037504784400485443]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[prAVci[c]*86400. for c in prAVci]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This was a preliminary table of precipitation event statistics.\n",
    "\n",
    "MODEL & Simulation               & [N_events]   & N_total    & [Duration]  & [Size]      & [Size_max]    Intensity //\n",
    "IPSL  & aqua-control             & 20.60 ± 3.72 & 43432.00,  & 1.88 ± 1.62 & 2.40 ± 1.89 & 3.15 ± 3.24  & 38.67093334198003 //\n",
    "IPSL  & aqua-control-lwoff       & 15.42 ± 3.54 & 33228.00,  & 1.84 ± 1.69 & 2.59 ± 2.82 & 3.59 ± 5.12  & 15.818915185928397 //\n",
    "CNRM-CM6-1 & aqua-control        & 14.55 ± 3.15 & 33565.00,  & 1.58 ± 1.77 & 4.36 ± 4.33 & 5.54 ± 7.26  & 41.60583381652839 //\n",
    "CNRM-CM6-1 & aqua-control-lwoff  & 12.74 ± 2.93 & 33339.00,  & 1.40 ± 1.21 & 5.12 ± 5.96 & 6.05 ± 8.43  & 41.29681282043464 //\n",
    "CESM2 & aqua-control             & 13.66 ± 3.37 & 29816.00,  & 1.67 ± 1.95 & 6.73 ± 7.95 & 8.91 ± 14.11 & 36.46170570373533 //\n",
    "CESM2 & aqua-control-lwoff       & 17.30 ± 3.37 & 35752.00,  & 1.77 ± 1.63 & 7.04 ± 7.40 & 9.49 ± 12.45 & 23.00404930114746 //\n",
    "CNRM-CM6-1 & amip                & 12.08 ± 2.90 & 89594.00,  & 1.77 ± 2.06 & 3.49 ± 3.59 & 4.54 ± 6.21  & 47.04558120727552 //\n",
    "CNRM-CM6-1 & amip-lwoff          & 12.48 ± 3.06 & 106396.00, & 1.54 ± 2.03 & 3.69 ± 3.81 & 4.49 ± 5.85  & 44.80594955444349 //\n",
    "IPSL-CM6A-LR &  amip             & 13.44 ± 3.51 & 79114.00,  & 2.42 ± 2.97 & 2.19 ± 1.84 & 2.99 ± 3.41  & 36.43431427001923 //\n",
    "IPSL-CM6A-LR & amip-lwoff        & 12.61 ± 3.74 & 81788.00,  & 2.03 ± 2.18 & 2.14 ± 2.30 & 2.92 ± 4.28  & 20.73476715087895 //\n",
    "CESM2 & amip                     & 13.81 ± 3.28 & 175658.00, & 1.87 ± 1.99 & 5.45 ± 6.14 & 7.54 ± 10.98 & 33.26182556152344 //\n",
    "CESM2 & amip-lwoff               & 12.34 ± 3.15 & 91516.00,  & 1.77 ± 2.00 & 5.88 ± 6.83 & 7.97 ± 12.27 & 30.823095607757466\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Average Number')"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Exploratory plot\n",
    "# stick diagram with yaxis being 99th percentile pr and xaxis being any of these clustering metrics\n",
    "epairs = ((\"cesm-amip\", \"cesm-amip-off\"), (\"ipsl-amip\", \"ipsl-amip-off\"), (\"cnrm-amip\", \"cnrm-amip-off\"), (\"cesm-aqua\", \"cesm-aqua-off\"), (\"ipsl-aqua\", \"ipsl-aqua-off\"), (\"cnrm-aqua\", \"cnrm-aqua-off\"))\n",
    "avg_n = {\"cesm-amip\": 13.81,\n",
    "        \"cesm-amip-off\": 12.34,\n",
    "        \"cesm-aqua\": 13.66,\n",
    "        \"cesm-aqua-off\": 17.30,\n",
    "        \"ipsl-amip\": 13.44,\n",
    "        \"ipsl-amip-off\": 12.61,\n",
    "        \"ipsl-aqua\": 20.60,\n",
    "        \"ipsl-aqua-off\": 15.42,\n",
    "        \"cnrm-amip\": 12.08,\n",
    "        \"cnrm-amip-off\": 12.48,\n",
    "        \"cnrm-aqua\": 14.55,\n",
    "        \"cnrm-aqua-off\": 12.74,\n",
    "       }\n",
    "\n",
    "avg_d = {\"cesm-amip\": 1.87,\n",
    "        \"cesm-amip-off\": 1.77,\n",
    "        \"cesm-aqua\": 1.67,\n",
    "        \"cesm-aqua-off\": 1.77,\n",
    "        \"ipsl-amip\": 2.42,\n",
    "        \"ipsl-amip-off\": 2.03,\n",
    "        \"ipsl-aqua\": 1.88,\n",
    "        \"ipsl-aqua-off\": 1.84,\n",
    "        \"cnrm-amip\": 1.77,\n",
    "        \"cnrm-amip-off\": 1.54,\n",
    "        \"cnrm-aqua\": 1.58,\n",
    "        \"cnrm-aqua-off\": 1.40,\n",
    "       }\n",
    "\n",
    "mclr = {\"cesm-amip\": 'blue',\n",
    "        \"cesm-amip-off\": 'blue',\n",
    "        \"cesm-aqua\": 'dodgerblue',\n",
    "        \"cesm-aqua-off\": 'dodgerblue',\n",
    "        \"ipsl-amip\": 'red',\n",
    "        \"ipsl-amip-off\": 'red',\n",
    "        \"ipsl-aqua\": 'pink',\n",
    "        \"ipsl-aqua-off\": 'pink',\n",
    "        \"cnrm-amip\": 'seagreen',\n",
    "        \"cnrm-amip-off\": 'seagreen',\n",
    "        \"cnrm-aqua\": 'green',\n",
    "        \"cnrm-aqua-off\": 'green',\n",
    "       }\n",
    "\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(ncols=2, constrained_layout=True, sharey=True)\n",
    "for p in epairs:\n",
    "    ax[0].plot([avg_d[p[0]], avg_d[p[1]]], [pr99[p[0]], pr99[p[1]]], color=mclr[p[0]])\n",
    "    ax[0].plot(avg_d[p[0]], pr99[p[0]], marker='o', color=mclr[p[0]])\n",
    "    ax[0].plot(avg_d[p[1]], pr99[p[1]], marker='.', color=mclr[p[0]])\n",
    "    ax[1].plot([avg_n[p[0]], avg_n[p[1]]], [pr99[p[0]], pr99[p[1]]], color=mclr[p[0]])\n",
    "    ax[1].plot(avg_n[p[0]], pr99[p[0]], marker='o', color=mclr[p[0]])\n",
    "    ax[1].plot(avg_n[p[1]], pr99[p[1]], marker='.', color=mclr[p[0]])\n",
    "ax[0].set_title(\"Average Duration\")\n",
    "ax[1].set_title(\"Average Number\")"
   ]
  },
  {
   "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": "3.9.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}