climate_sensitivity_sampling.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import packages\n",
    "from itertools import combinations\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import random\n",
    "\n",
    "import itertools\n",
    "from matplotlib.cbook import _reshape_2D"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Getting the climate sensitivity values for each model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "climate_sensitives = {\n",
    "    'ACCESS-CM2': 4.7,\n",
    "    'ACCESS-ESM1-5': 3.9,\n",
    "    'BCC-CSM2-MR': 3.0,\n",
    "    'CAMS-CSM1-0': 2.3,\n",
    "    'CESM2': 5.2,\n",
    "    'CESM2-WACCM': 4.8,\n",
    "    'CMCC-CM2-SR5': 3.52,\n",
    "    'CMCC-ESM2': 3.57,\n",
    "    'CanESM5': 5.6,\n",
    "    'E3SM-1-1': np.nan,\n",
    "    'EC-Earth3-Veg-LR': np.nan,\n",
    "    'FGOALS-f3-L': 3.0,\n",
    "    'FGOALS-g3': np.nan,\n",
    "    'FIO-ESM-2-0': np.nan,\n",
    "    'GFDL-CM4': 3.9,\n",
    "    'GFDL-ESM4': 2.6,\n",
    "    'IITM-ESM': np.nan,\n",
    "    'INM-CM4-8': 1.8,\n",
    "    'INM-CM5-0': 1.9,\n",
    "    'IPSL-CM5A2-INCA': np.nan,\n",
    "    'IPSL-CM6A-LR': 4.6,\n",
    "    'KIOST-ESM': np.nan,\n",
    "    'MIROC6': 2.6,\n",
    "    'MPI-ESM1-2-HR': 3.0,\n",
    "    'MPI-ESM1-2-LR': 3.0,\n",
    "    'MRI-ESM2-0': 3.2,\n",
    "    'NESM3': 4.7,\n",
    "    'NorESM2-LM': 2.5,\n",
    "    'NorESM2-MM': 2.5,\n",
    "    'UKESM1-0-LL': 5.3\n",
    "}\n",
    "climate_sensitivies_df = pd.DataFrame(climate_sensitives.items(), columns = ['esm', 'climate_sensitivity'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Getting the number of ensembles and different experiments available for each of the models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>esm</th>\n",
       "      <th>experiment</th>\n",
       "      <th>ensemble</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>ssp126</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>ssp245</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>ssp370</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>ssp585</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>ssp126</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>ssp126</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>ssp245</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>ssp370</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>ssp434</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>ssp585</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>133 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               esm experiment  ensemble\n",
       "0       ACCESS-CM2     ssp126         5\n",
       "1       ACCESS-CM2     ssp245         5\n",
       "2       ACCESS-CM2     ssp370         5\n",
       "3       ACCESS-CM2     ssp585         5\n",
       "4    ACCESS-ESM1-5     ssp126        40\n",
       "..             ...        ...       ...\n",
       "128    UKESM1-0-LL     ssp126        13\n",
       "129    UKESM1-0-LL     ssp245        14\n",
       "130    UKESM1-0-LL     ssp370        13\n",
       "131    UKESM1-0-LL     ssp434         5\n",
       "132    UKESM1-0-LL     ssp585         5\n",
       "\n",
       "[133 rows x 3 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics_df = pd.read_csv('extracted_timeseries/extracted_metrics/IPCC_land_regions_metrics.csv')\n",
    "my_combinations = metrics_df[['esm', 'experiment', 'ensemble']].drop_duplicates()\n",
    "combinations_count = my_combinations.groupby(['esm', 'experiment']).count().reset_index()\n",
    "combinations_count"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Defining a function to decide whether to include the model in the running. Has to have a certain set of experiments available, and a certain number of total ensemble runs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def inclusion_criterion(df):\n",
    "    contains_all = np.all(np.isin(['ssp126', 'ssp245', 'ssp370', 'ssp585'], df.experiment))\n",
    "    many_runs = np.sum(df.ensemble) >= 4\n",
    "    try:\n",
    "        known_climate_sensitivity = ~np.isnan(climate_sensitives[df['esm'].values[0]])\n",
    "    except KeyError:\n",
    "        known_climate_sensitivity = False\n",
    "\n",
    "    return contains_all & many_runs  & known_climate_sensitivity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These are the models which meet the criteria set in the above function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>esm</th>\n",
       "      <th>include</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CAMS-CSM1-0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CESM2</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>CESM2-WACCM</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>CMCC-ESM2</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>CanESM5</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>FGOALS-f3-L</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>INM-CM4-8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>INM-CM5-0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>IPSL-CM6A-LR</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>MIROC6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>MPI-ESM1-2-HR</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>NorESM2-LM</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>NorESM2-MM</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              esm  include\n",
       "0      ACCESS-CM2     True\n",
       "1   ACCESS-ESM1-5     True\n",
       "2     BCC-CSM2-MR     True\n",
       "3     CAMS-CSM1-0     True\n",
       "4           CESM2     True\n",
       "5     CESM2-WACCM     True\n",
       "7    CMCC-CM2-SR5     True\n",
       "8       CMCC-ESM2     True\n",
       "9         CanESM5     True\n",
       "12    FGOALS-f3-L     True\n",
       "16      GFDL-ESM4     True\n",
       "18      INM-CM4-8     True\n",
       "19      INM-CM5-0     True\n",
       "21   IPSL-CM6A-LR     True\n",
       "23         MIROC6     True\n",
       "24  MPI-ESM1-2-HR     True\n",
       "25  MPI-ESM1-2-LR     True\n",
       "26     MRI-ESM2-0     True\n",
       "28     NorESM2-LM     True\n",
       "29     NorESM2-MM     True\n",
       "30    UKESM1-0-LL     True"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inclusion_df = combinations_count.groupby(['esm']).apply(inclusion_criterion)\n",
    "inclusion_df = pd.DataFrame(inclusion_df).reset_index().rename(columns={0: 'include'})\n",
    "inclusion_df[inclusion_df['include']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(inclusion_df[inclusion_df['include']]['esm'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is how the distribution of climate sensitivities changes when we only include models that meet the criteria"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(313.39267676767673, 0.5, '')"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "included_esms = inclusion_df[inclusion_df['include']]['esm'].values\n",
    "included_cs_df = climate_sensitivies_df[np.isin(climate_sensitivies_df['esm'], included_esms)]\n",
    "\n",
    "fig, (ax1,ax2) = plt.subplots(1,2)\n",
    "sns.boxplot(ax = ax1, y='climate_sensitivity', data=climate_sensitivies_df, showfliers = False, color='#89a2c4')\n",
    "sns.swarmplot(ax = ax1, y='climate_sensitivity', data=climate_sensitivies_df, color='0.25')\n",
    "sns.boxplot(ax = ax2, y='climate_sensitivity', data=included_cs_df, showfliers = False, color='#89a2c4')\n",
    "sns.swarmplot(ax = ax2, y='climate_sensitivity', data=included_cs_df, color='0.25')\n",
    "\n",
    "ax1.set_title('All Models')\n",
    "ax2.set_title('Criteria Trimmed')\n",
    "ax1.tick_params(axis='x', bottom=False, top=False, which='both', labelbottom=False)\n",
    "ax2.tick_params(axis='x', bottom=False, top=False, which='both', labelbottom=False)\n",
    "ax1.set_ylabel('Climate Sensitivity')\n",
    "ax2.set_ylabel('')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cs_group(x):\n",
    "    if (x < 2) or (x > 5):\n",
    "        return 'unlikely'\n",
    "    elif (x < 4) and (x > 2.5):\n",
    "        return 'likely'\n",
    "\n",
    "    return 'very likely'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/03/82f8dprj20zdn3zkgcd_p1280000gn/T/ipykernel_4619/448732164.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  included_cs_df['group'] = [cs_group(x) for x in included_cs_df.climate_sensitivity]\n"
     ]
    },
    {
     "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>esm</th>\n",
       "      <th>climate_sensitivity</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>4.70</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>3.90</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CAMS-CSM1-0</td>\n",
       "      <td>2.30</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CESM2</td>\n",
       "      <td>5.20</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>CESM2-WACCM</td>\n",
       "      <td>4.80</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>3.52</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>CMCC-ESM2</td>\n",
       "      <td>3.57</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>CanESM5</td>\n",
       "      <td>5.60</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>FGOALS-f3-L</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>2.60</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>INM-CM4-8</td>\n",
       "      <td>1.80</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>INM-CM5-0</td>\n",
       "      <td>1.90</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>IPSL-CM6A-LR</td>\n",
       "      <td>4.60</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>MIROC6</td>\n",
       "      <td>2.60</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>MPI-ESM1-2-HR</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.20</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>NorESM2-LM</td>\n",
       "      <td>2.50</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>NorESM2-MM</td>\n",
       "      <td>2.50</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>5.30</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              esm  climate_sensitivity        group\n",
       "0      ACCESS-CM2                 4.70  very likely\n",
       "1   ACCESS-ESM1-5                 3.90       likely\n",
       "2     BCC-CSM2-MR                 3.00       likely\n",
       "3     CAMS-CSM1-0                 2.30  very likely\n",
       "4           CESM2                 5.20     unlikely\n",
       "5     CESM2-WACCM                 4.80  very likely\n",
       "6    CMCC-CM2-SR5                 3.52       likely\n",
       "7       CMCC-ESM2                 3.57       likely\n",
       "8         CanESM5                 5.60     unlikely\n",
       "9     FGOALS-f3-L                 3.00       likely\n",
       "10      GFDL-ESM4                 2.60       likely\n",
       "11      INM-CM4-8                 1.80     unlikely\n",
       "12      INM-CM5-0                 1.90     unlikely\n",
       "13   IPSL-CM6A-LR                 4.60  very likely\n",
       "14         MIROC6                 2.60       likely\n",
       "15  MPI-ESM1-2-HR                 3.00       likely\n",
       "16  MPI-ESM1-2-LR                 3.00       likely\n",
       "17     MRI-ESM2-0                 3.20       likely\n",
       "18     NorESM2-LM                 2.50  very likely\n",
       "19     NorESM2-MM                 2.50  very likely\n",
       "20    UKESM1-0-LL                 5.30     unlikely"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df['group'] = [cs_group(x) for x in included_cs_df.climate_sensitivity]\n",
    "included_cs_df = included_cs_df.reset_index(drop=True)\n",
    "included_cs_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[2.6, 3.0, 3.2, 3.52, 3.57, 3.9]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[x for x in np.unique(np.array(included_cs_df.climate_sensitivity.values)) if (x < 4) and (x > 2.5)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1,  2,  6,  7,  9, 10, 14, 15, 16, 17])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df[included_cs_df['group'] == 'likely'].index.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from itertools import combinations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>esm</th>\n",
       "      <th>climate_sensitivity</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>3.90</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>3.52</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>CMCC-ESM2</td>\n",
       "      <td>3.57</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>FGOALS-f3-L</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>2.60</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>MIROC6</td>\n",
       "      <td>2.60</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>MPI-ESM1-2-HR</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.20</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              esm  climate_sensitivity   group\n",
       "1   ACCESS-ESM1-5                 3.90  likely\n",
       "2     BCC-CSM2-MR                 3.00  likely\n",
       "6    CMCC-CM2-SR5                 3.52  likely\n",
       "7       CMCC-ESM2                 3.57  likely\n",
       "9     FGOALS-f3-L                 3.00  likely\n",
       "10      GFDL-ESM4                 2.60  likely\n",
       "14         MIROC6                 2.60  likely\n",
       "15  MPI-ESM1-2-HR                 3.00  likely\n",
       "16  MPI-ESM1-2-LR                 3.00  likely\n",
       "17     MRI-ESM2-0                 3.20  likely"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df[included_cs_df['group'] == 'likely']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>esm1</th>\n",
       "      <th>esm2</th>\n",
       "      <th>esm3</th>\n",
       "      <th>esm4</th>\n",
       "      <th>esm5</th>\n",
       "      <th>cs1</th>\n",
       "      <th>cs2</th>\n",
       "      <th>cs3</th>\n",
       "      <th>cs4</th>\n",
       "      <th>cs5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>4.7</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>CAMS-CSM1-0</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.3</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>CESM2-WACCM</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.8</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>IPSL-CM6A-LR</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.6</td>\n",
       "      <td>4.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>NorESM2-LM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.6</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CAMS-CSM1-0</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2.3</td>\n",
       "      <td>2.6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>176</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CESM2-WACCM</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.9</td>\n",
       "      <td>4.8</td>\n",
       "      <td>2.6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>IPSL-CM6A-LR</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.9</td>\n",
       "      <td>4.6</td>\n",
       "      <td>2.6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>NorESM2-LM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2.6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.2</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>179</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>NorESM2-MM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2.6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.2</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>180 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              esm1           esm2           esm3           esm4          esm5  \\\n",
       "0       ACCESS-CM2  ACCESS-ESM1-5    BCC-CSM2-MR   CMCC-CM2-SR5     GFDL-ESM4   \n",
       "1    ACCESS-ESM1-5    BCC-CSM2-MR    CAMS-CSM1-0   CMCC-CM2-SR5     GFDL-ESM4   \n",
       "2    ACCESS-ESM1-5    BCC-CSM2-MR    CESM2-WACCM   CMCC-CM2-SR5     GFDL-ESM4   \n",
       "3    ACCESS-ESM1-5    BCC-CSM2-MR   CMCC-CM2-SR5      GFDL-ESM4  IPSL-CM6A-LR   \n",
       "4    ACCESS-ESM1-5    BCC-CSM2-MR   CMCC-CM2-SR5      GFDL-ESM4    NorESM2-LM   \n",
       "..             ...            ...            ...            ...           ...   \n",
       "175  ACCESS-ESM1-5    CAMS-CSM1-0         MIROC6  MPI-ESM1-2-LR    MRI-ESM2-0   \n",
       "176  ACCESS-ESM1-5    CESM2-WACCM         MIROC6  MPI-ESM1-2-LR    MRI-ESM2-0   \n",
       "177  ACCESS-ESM1-5   IPSL-CM6A-LR         MIROC6  MPI-ESM1-2-LR    MRI-ESM2-0   \n",
       "178  ACCESS-ESM1-5         MIROC6  MPI-ESM1-2-LR     MRI-ESM2-0    NorESM2-LM   \n",
       "179  ACCESS-ESM1-5         MIROC6  MPI-ESM1-2-LR     MRI-ESM2-0    NorESM2-MM   \n",
       "\n",
       "     cs1  cs2   cs3   cs4  cs5  \n",
       "0    4.7  3.9   3.0  3.52  2.6  \n",
       "1    3.9  3.0   2.3  3.52  2.6  \n",
       "2    3.9  3.0   4.8  3.52  2.6  \n",
       "3    3.9  3.0  3.52   2.6  4.6  \n",
       "4    3.9  3.0  3.52   2.6  2.5  \n",
       "..   ...  ...   ...   ...  ...  \n",
       "175  3.9  2.3   2.6   3.0  3.2  \n",
       "176  3.9  4.8   2.6   3.0  3.2  \n",
       "177  3.9  4.6   2.6   3.0  3.2  \n",
       "178  3.9  2.6   3.0   3.2  2.5  \n",
       "179  3.9  2.6   3.0   3.2  2.5  \n",
       "\n",
       "[180 rows x 10 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "likely_combs = list(combinations(included_cs_df[included_cs_df['group'] == 'likely'].index.values, 4))\n",
    "df = pd.DataFrame(columns=['esm1', 'esm2', 'esm3', 'esm4', 'esm5', 'cs1', 'cs2', 'cs3', 'cs4', 'cs5'])\n",
    "for comb in likely_combs:\n",
    "    esm_sample = included_cs_df.iloc[np.array(comb)]\n",
    "    sample_cs = np.array(esm_sample.climate_sensitivity.values)\n",
    "    consecutive_diff = np.diff(np.sort(sample_cs))\n",
    "    if ~np.isin(0, consecutive_diff) and np.isin(2.6, sample_cs) and np.isin(3.9, sample_cs):\n",
    "        for ind in included_cs_df[included_cs_df['group'] == 'very likely'].index:\n",
    "            vl_row = included_cs_df.iloc[ind]\n",
    "            esms_sorted = np.sort(np.concatenate((np.array(esm_sample.esm.values), np.array(vl_row.esm).reshape(1))))\n",
    "            sample_cs = [climate_sensitives[esms_sorted[j]] for j in range(len(esms_sorted))]\n",
    "            new_row = np.concatenate((esms_sorted, np.array(sample_cs)))\n",
    "            df = pd.concat([df, pd.DataFrame(new_row.reshape(1,-1), columns=list(df))], ignore_index=True)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['ACCESS-CM2'], dtype='<U10')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(included_cs_df.iloc[0].esm).reshape(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Int64Index([0, 3, 5, 13, 18, 19], dtype='int64')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df[included_cs_df['group'] == 'very likely'].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 2, 6, 7])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(likely_combs[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>esm</th>\n",
       "      <th>climate_sensitivity</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>3.90</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>3.52</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>CMCC-ESM2</td>\n",
       "      <td>3.57</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             esm  climate_sensitivity   group\n",
       "1  ACCESS-ESM1-5                 3.90  likely\n",
       "2    BCC-CSM2-MR                 3.00  likely\n",
       "6   CMCC-CM2-SR5                 3.52  likely\n",
       "7      CMCC-ESM2                 3.57  likely"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df.iloc[np.array(likely_combs[0])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df[included_cs_df['group'] == 'likely'].shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df[included_cs_df['group'] == 'very likely'].shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_sample(n):\n",
    "    max_l = included_cs_df[included_cs_df['group'] == 'likely'].shape[0]\n",
    "    max_vl = included_cs_df[included_cs_df['group'] == 'very likely'].shape[0]\n",
    "    if n > (max_l + max_vl):\n",
    "        raise Exception('Given sample size larger than available models')\n",
    "\n",
    "    # Number of sample from each group\n",
    "    likely_size = int(np.ceil(n * 2/3))\n",
    "    very_likely_size = n - likely_size\n",
    "\n",
    "    # If 2/3*n is more models than are in likely category, will have to get more from unlikely\n",
    "    # And vice versa\n",
    "    if very_likely_size > max_vl:\n",
    "        very_likely_size = max_vl\n",
    "        likely_size = n - very_likely_size\n",
    "    elif likely_size > max_l:\n",
    "        likely_size = max_l\n",
    "        very_likely_size = n - likely_size\n",
    "    \n",
    "    # Sample ESMs\n",
    "    ssl = included_cs_df[included_cs_df['group'] == 'likely'].sample(likely_size)\n",
    "    ssvl = included_cs_df[included_cs_df['group'] == 'very likely'].sample(very_likely_size)\n",
    "\n",
    "    # Return\n",
    "    return ssl, ssvl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_good_sample(n, max_iters: int = 100):\n",
    "    best_ssl = None\n",
    "    best_ssvl = None\n",
    "    best_ssl_2diff = 0\n",
    "    best_ssvl_2diff = 0\n",
    "\n",
    "    for i in range(max_iters):\n",
    "        ssl, ssvl = get_sample(n)\n",
    "\n",
    "        avg_ssl_2diff = np.sum((np.diff(np.sort(ssl.climate_sensitivity))))\n",
    "        avg_ssvl_2diff = np.sum((np.diff(np.sort(ssvl.climate_sensitivity))))\n",
    "\n",
    "        if ~np.isin(0, np.diff(np.sort(ssl.climate_sensitivity))):\n",
    "            # Likely range sample\n",
    "            if (avg_ssl_2diff > best_ssl_2diff) or (best_ssl is None):\n",
    "                best_ssl_2diff = avg_ssl_2diff\n",
    "                best_ssl = ssl\n",
    "            elif (avg_ssl_2diff == best_ssl_2diff ) and random.choice([True, False]):\n",
    "                best_ssl_2diff = avg_ssl_2diff\n",
    "                best_ssl = ssl\n",
    "            # Very Likely range sample\n",
    "            if (avg_ssvl_2diff > best_ssvl_2diff) or (best_ssvl is None):\n",
    "                best_ssvl_2diff = avg_ssvl_2diff\n",
    "                best_ssvl = ssvl\n",
    "\n",
    "    return best_ssl, best_ssvl, best_ssl_2diff, best_ssvl_2diff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>esm</th>\n",
       "      <th>climate_sensitivity</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>3.90</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>3.52</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>MIROC6</td>\n",
       "      <td>2.60</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>FGOALS-f3-L</td>\n",
       "      <td>3.00</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              esm  climate_sensitivity   group\n",
       "1   ACCESS-ESM1-5                 3.90  likely\n",
       "6    CMCC-CM2-SR5                 3.52  likely\n",
       "14         MIROC6                 2.60  likely\n",
       "9     FGOALS-f3-L                 3.00  likely"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ssl, ssvl, _, _ = get_good_sample(5)\n",
    "ssl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>esm</th>\n",
       "      <th>climate_sensitivity</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>CESM2-WACCM</td>\n",
       "      <td>4.8</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           esm  climate_sensitivity        group\n",
       "5  CESM2-WACCM                  4.8  very likely"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ssvl"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get combinations of good samples of size 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[25], line 4\u001b[0m\n\u001b[1;32m      2\u001b[0m esm_combos \u001b[39m=\u001b[39m pd\u001b[39m.\u001b[39mDataFrame(columns\u001b[39m=\u001b[39m[\u001b[39m'\u001b[39m\u001b[39mesm1\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mesm2\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mesm3\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mesm4\u001b[39m\u001b[39m'\u001b[39m, \u001b[39m'\u001b[39m\u001b[39mesm5\u001b[39m\u001b[39m'\u001b[39m])\n\u001b[1;32m      3\u001b[0m \u001b[39mfor\u001b[39;00m i \u001b[39min\u001b[39;00m \u001b[39mrange\u001b[39m(max_iters):\n\u001b[0;32m----> 4\u001b[0m     ssl, ssvl, _, _ \u001b[39m=\u001b[39m get_good_sample(\u001b[39m5\u001b[39;49m)\n\u001b[1;32m      5\u001b[0m     sample_arr \u001b[39m=\u001b[39m np\u001b[39m.\u001b[39msort(np\u001b[39m.\u001b[39mconcatenate((np\u001b[39m.\u001b[39marray(ssl\u001b[39m.\u001b[39mesm\u001b[39m.\u001b[39mvalues), np\u001b[39m.\u001b[39marray(ssvl\u001b[39m.\u001b[39mesm\u001b[39m.\u001b[39mvalues))))\n\u001b[1;32m      6\u001b[0m     esm_combos \u001b[39m=\u001b[39m pd\u001b[39m.\u001b[39mconcat([esm_combos, pd\u001b[39m.\u001b[39mDataFrame(sample_arr\u001b[39m.\u001b[39mreshape(\u001b[39m1\u001b[39m,\u001b[39m-\u001b[39m\u001b[39m1\u001b[39m), columns\u001b[39m=\u001b[39m\u001b[39mlist\u001b[39m(esm_combos))], ignore_index\u001b[39m=\u001b[39m\u001b[39mTrue\u001b[39;00m)\n",
      "Cell \u001b[0;32mIn[22], line 8\u001b[0m, in \u001b[0;36mget_good_sample\u001b[0;34m(n, max_iters)\u001b[0m\n\u001b[1;32m      5\u001b[0m best_ssvl_2diff \u001b[39m=\u001b[39m \u001b[39m0\u001b[39m\n\u001b[1;32m      7\u001b[0m \u001b[39mfor\u001b[39;00m i \u001b[39min\u001b[39;00m \u001b[39mrange\u001b[39m(max_iters):\n\u001b[0;32m----> 8\u001b[0m     ssl, ssvl \u001b[39m=\u001b[39m get_sample(n)\n\u001b[1;32m     10\u001b[0m     avg_ssl_2diff \u001b[39m=\u001b[39m np\u001b[39m.\u001b[39msum((np\u001b[39m.\u001b[39mdiff(np\u001b[39m.\u001b[39msort(ssl\u001b[39m.\u001b[39mclimate_sensitivity))))\n\u001b[1;32m     11\u001b[0m     avg_ssvl_2diff \u001b[39m=\u001b[39m np\u001b[39m.\u001b[39msum((np\u001b[39m.\u001b[39mdiff(np\u001b[39m.\u001b[39msort(ssvl\u001b[39m.\u001b[39mclimate_sensitivity))))\n",
      "Cell \u001b[0;32mIn[21], line 2\u001b[0m, in \u001b[0;36mget_sample\u001b[0;34m(n)\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mget_sample\u001b[39m(n):\n\u001b[0;32m----> 2\u001b[0m     max_l \u001b[39m=\u001b[39m included_cs_df[included_cs_df[\u001b[39m'\u001b[39;49m\u001b[39mgroup\u001b[39;49m\u001b[39m'\u001b[39;49m] \u001b[39m==\u001b[39;49m \u001b[39m'\u001b[39;49m\u001b[39mlikely\u001b[39;49m\u001b[39m'\u001b[39;49m]\u001b[39m.\u001b[39mshape[\u001b[39m0\u001b[39m]\n\u001b[1;32m      3\u001b[0m     max_vl \u001b[39m=\u001b[39m included_cs_df[included_cs_df[\u001b[39m'\u001b[39m\u001b[39mgroup\u001b[39m\u001b[39m'\u001b[39m] \u001b[39m==\u001b[39m \u001b[39m'\u001b[39m\u001b[39mvery likely\u001b[39m\u001b[39m'\u001b[39m]\u001b[39m.\u001b[39mshape[\u001b[39m0\u001b[39m]\n\u001b[1;32m      4\u001b[0m     \u001b[39mif\u001b[39;00m n \u001b[39m>\u001b[39m (max_l \u001b[39m+\u001b[39m max_vl):\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/site-packages/pandas/core/ops/common.py:72\u001b[0m, in \u001b[0;36m_unpack_zerodim_and_defer.<locals>.new_method\u001b[0;34m(self, other)\u001b[0m\n\u001b[1;32m     68\u001b[0m             \u001b[39mreturn\u001b[39;00m \u001b[39mNotImplemented\u001b[39m\n\u001b[1;32m     70\u001b[0m other \u001b[39m=\u001b[39m item_from_zerodim(other)\n\u001b[0;32m---> 72\u001b[0m \u001b[39mreturn\u001b[39;00m method(\u001b[39mself\u001b[39;49m, other)\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/site-packages/pandas/core/arraylike.py:42\u001b[0m, in \u001b[0;36mOpsMixin.__eq__\u001b[0;34m(self, other)\u001b[0m\n\u001b[1;32m     40\u001b[0m \u001b[39m@unpack_zerodim_and_defer\u001b[39m(\u001b[39m\"\u001b[39m\u001b[39m__eq__\u001b[39m\u001b[39m\"\u001b[39m)\n\u001b[1;32m     41\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m__eq__\u001b[39m(\u001b[39mself\u001b[39m, other):\n\u001b[0;32m---> 42\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_cmp_method(other, operator\u001b[39m.\u001b[39;49meq)\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/site-packages/pandas/core/series.py:6245\u001b[0m, in \u001b[0;36mSeries._cmp_method\u001b[0;34m(self, other, op)\u001b[0m\n\u001b[1;32m   6242\u001b[0m \u001b[39mwith\u001b[39;00m np\u001b[39m.\u001b[39merrstate(\u001b[39mall\u001b[39m\u001b[39m=\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mignore\u001b[39m\u001b[39m\"\u001b[39m):\n\u001b[1;32m   6243\u001b[0m     res_values \u001b[39m=\u001b[39m ops\u001b[39m.\u001b[39mcomparison_op(lvalues, rvalues, op)\n\u001b[0;32m-> 6245\u001b[0m \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_construct_result(res_values, name\u001b[39m=\u001b[39;49mres_name)\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/site-packages/pandas/core/series.py:3223\u001b[0m, in \u001b[0;36mSeries._construct_result\u001b[0;34m(self, result, name)\u001b[0m\n\u001b[1;32m   3219\u001b[0m     \u001b[39mreturn\u001b[39;00m (res1, res2)\n\u001b[1;32m   3221\u001b[0m \u001b[39m# We do not pass dtype to ensure that the Series constructor\u001b[39;00m\n\u001b[1;32m   3222\u001b[0m \u001b[39m#  does inference in the case where `result` has object-dtype.\u001b[39;00m\n\u001b[0;32m-> 3223\u001b[0m out \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_constructor(result, index\u001b[39m=\u001b[39;49m\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mindex)\n\u001b[1;32m   3224\u001b[0m out \u001b[39m=\u001b[39m out\u001b[39m.\u001b[39m__finalize__(\u001b[39mself\u001b[39m)\n\u001b[1;32m   3226\u001b[0m \u001b[39m# Set the result's name after __finalize__ is called because __finalize__\u001b[39;00m\n\u001b[1;32m   3227\u001b[0m \u001b[39m#  would set it back to self.name\u001b[39;00m\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/site-packages/pandas/core/series.py:470\u001b[0m, in \u001b[0;36mSeries.__init__\u001b[0;34m(self, data, index, dtype, name, copy, fastpath)\u001b[0m\n\u001b[1;32m    468\u001b[0m         data \u001b[39m=\u001b[39m data\u001b[39m.\u001b[39mcopy()\n\u001b[1;32m    469\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[0;32m--> 470\u001b[0m     data \u001b[39m=\u001b[39m sanitize_array(data, index, dtype, copy)\n\u001b[1;32m    472\u001b[0m     manager \u001b[39m=\u001b[39m get_option(\u001b[39m\"\u001b[39m\u001b[39mmode.data_manager\u001b[39m\u001b[39m\"\u001b[39m)\n\u001b[1;32m    473\u001b[0m     \u001b[39mif\u001b[39;00m manager \u001b[39m==\u001b[39m \u001b[39m\"\u001b[39m\u001b[39mblock\u001b[39m\u001b[39m\"\u001b[39m:\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/site-packages/pandas/core/construction.py:597\u001b[0m, in \u001b[0;36msanitize_array\u001b[0;34m(data, index, dtype, copy, raise_cast_failure, allow_2d)\u001b[0m\n\u001b[1;32m    594\u001b[0m             subarr \u001b[39m=\u001b[39m np\u001b[39m.\u001b[39marray(data, copy\u001b[39m=\u001b[39mcopy)\n\u001b[1;32m    595\u001b[0m     \u001b[39melse\u001b[39;00m:\n\u001b[1;32m    596\u001b[0m         \u001b[39m# we will try to copy by-definition here\u001b[39;00m\n\u001b[0;32m--> 597\u001b[0m         subarr \u001b[39m=\u001b[39m _try_cast(data, dtype, copy, raise_cast_failure)\n\u001b[1;32m    599\u001b[0m \u001b[39melif\u001b[39;00m \u001b[39misinstance\u001b[39m(data, ABCExtensionArray):\n\u001b[1;32m    600\u001b[0m     \u001b[39m# it is already ensured above this is not a PandasArray\u001b[39;00m\n\u001b[1;32m    601\u001b[0m     subarr \u001b[39m=\u001b[39m data\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/site-packages/pandas/core/construction.py:773\u001b[0m, in \u001b[0;36m_try_cast\u001b[0;34m(arr, dtype, copy, raise_cast_failure)\u001b[0m\n\u001b[1;32m    770\u001b[0m \u001b[39mif\u001b[39;00m dtype \u001b[39mis\u001b[39;00m \u001b[39mNone\u001b[39;00m:\n\u001b[1;32m    771\u001b[0m     \u001b[39m# perf shortcut as this is the most common case\u001b[39;00m\n\u001b[1;32m    772\u001b[0m     \u001b[39mif\u001b[39;00m is_ndarray:\n\u001b[0;32m--> 773\u001b[0m         arr \u001b[39m=\u001b[39m cast(np\u001b[39m.\u001b[39;49mndarray, arr)\n\u001b[1;32m    774\u001b[0m         \u001b[39mif\u001b[39;00m arr\u001b[39m.\u001b[39mdtype \u001b[39m!=\u001b[39m \u001b[39mobject\u001b[39m:\n\u001b[1;32m    775\u001b[0m             \u001b[39mreturn\u001b[39;00m sanitize_to_nanoseconds(arr, copy\u001b[39m=\u001b[39mcopy)\n",
      "File \u001b[0;32m~/opt/miniconda3/envs/stitches_env/lib/python3.9/typing.py:1375\u001b[0m, in \u001b[0;36mcast\u001b[0;34m(typ, val)\u001b[0m\n\u001b[1;32m   1371\u001b[0m     \u001b[39mcls\u001b[39m\u001b[39m.\u001b[39m_is_runtime_protocol \u001b[39m=\u001b[39m \u001b[39mTrue\u001b[39;00m\n\u001b[1;32m   1372\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mcls\u001b[39m\n\u001b[0;32m-> 1375\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mcast\u001b[39m(typ, val):\n\u001b[1;32m   1376\u001b[0m \u001b[39m    \u001b[39m\u001b[39m\"\"\"Cast a value to a type.\u001b[39;00m\n\u001b[1;32m   1377\u001b[0m \n\u001b[1;32m   1378\u001b[0m \u001b[39m    This returns the value unchanged.  To the type checker this\u001b[39;00m\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   1381\u001b[0m \u001b[39m    to be as fast as possible).\u001b[39;00m\n\u001b[1;32m   1382\u001b[0m \u001b[39m    \"\"\"\u001b[39;00m\n\u001b[1;32m   1383\u001b[0m     \u001b[39mreturn\u001b[39;00m val\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "max_iters = 500\n",
    "esm_combos = pd.DataFrame(columns=['esm1', 'esm2', 'esm3', 'esm4', 'esm5'])\n",
    "for i in range(max_iters):\n",
    "    ssl, ssvl, _, _ = get_good_sample(5)\n",
    "    sample_arr = np.sort(np.concatenate((np.array(ssl.esm.values), np.array(ssvl.esm.values))))\n",
    "    esm_combos = pd.concat([esm_combos, pd.DataFrame(sample_arr.reshape(1,-1), columns=list(esm_combos))], ignore_index=True)\n",
    "    esm_combos = esm_combos.drop_duplicates()\n",
    "esm_combos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_combos = pd.DataFrame(columns=['cs1', 'cs2', 'cs3', 'cs4', 'cs5'])\n",
    "for i in range(esm_combos.shape[0]):\n",
    "    esm_sample = esm_combos.iloc[i,:]\n",
    "    sample_cs = [climate_sensitives[esm_sample.values[j]] for j in range(5)]\n",
    "    cs_combos = pd.concat([cs_combos, pd.DataFrame(np.array(sample_cs).reshape(1,-1), columns=list(cs_combos))], ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "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>esm1</th>\n",
       "      <th>esm2</th>\n",
       "      <th>esm3</th>\n",
       "      <th>esm4</th>\n",
       "      <th>esm5</th>\n",
       "      <th>cs1</th>\n",
       "      <th>cs2</th>\n",
       "      <th>cs3</th>\n",
       "      <th>cs4</th>\n",
       "      <th>cs5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>CESM2-WACCM</td>\n",
       "      <td>CMCC-ESM2</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.00</td>\n",
       "      <td>4.80</td>\n",
       "      <td>3.57</td>\n",
       "      <td>2.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>FGOALS-f3-L</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>NorESM2-MM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.52</td>\n",
       "      <td>3.00</td>\n",
       "      <td>2.60</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CAMS-CSM1-0</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2.30</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.60</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>FGOALS-f3-L</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>NorESM2-LM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.52</td>\n",
       "      <td>3.00</td>\n",
       "      <td>2.60</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>CMCC-ESM2</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>NorESM2-LM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.52</td>\n",
       "      <td>3.57</td>\n",
       "      <td>2.60</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>NorESM2-LM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.60</td>\n",
       "      <td>3.00</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>GFDL-ESM4</td>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>NorESM2-MM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.60</td>\n",
       "      <td>3.00</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>BCC-CSM2-MR</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>4.7</td>\n",
       "      <td>3.90</td>\n",
       "      <td>3.00</td>\n",
       "      <td>3.52</td>\n",
       "      <td>2.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CMCC-CM2-SR5</td>\n",
       "      <td>CMCC-ESM2</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>NorESM2-MM</td>\n",
       "      <td>3.9</td>\n",
       "      <td>3.52</td>\n",
       "      <td>3.57</td>\n",
       "      <td>2.60</td>\n",
       "      <td>2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>CAMS-CSM1-0</td>\n",
       "      <td>FGOALS-f3-L</td>\n",
       "      <td>MIROC6</td>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2.30</td>\n",
       "      <td>3.00</td>\n",
       "      <td>2.60</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>120 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              esm1           esm2          esm3           esm4        esm5  \\\n",
       "0    ACCESS-ESM1-5    BCC-CSM2-MR   CESM2-WACCM      CMCC-ESM2   GFDL-ESM4   \n",
       "1    ACCESS-ESM1-5   CMCC-CM2-SR5   FGOALS-f3-L      GFDL-ESM4  NorESM2-MM   \n",
       "2    ACCESS-ESM1-5    CAMS-CSM1-0  CMCC-CM2-SR5         MIROC6  MRI-ESM2-0   \n",
       "3    ACCESS-ESM1-5   CMCC-CM2-SR5   FGOALS-f3-L      GFDL-ESM4  NorESM2-LM   \n",
       "4    ACCESS-ESM1-5   CMCC-CM2-SR5     CMCC-ESM2      GFDL-ESM4  NorESM2-LM   \n",
       "..             ...            ...           ...            ...         ...   \n",
       "115  ACCESS-ESM1-5   CMCC-CM2-SR5     GFDL-ESM4  MPI-ESM1-2-LR  NorESM2-LM   \n",
       "116  ACCESS-ESM1-5   CMCC-CM2-SR5     GFDL-ESM4  MPI-ESM1-2-LR  NorESM2-MM   \n",
       "117     ACCESS-CM2  ACCESS-ESM1-5   BCC-CSM2-MR   CMCC-CM2-SR5      MIROC6   \n",
       "118  ACCESS-ESM1-5   CMCC-CM2-SR5     CMCC-ESM2         MIROC6  NorESM2-MM   \n",
       "119  ACCESS-ESM1-5    CAMS-CSM1-0   FGOALS-f3-L         MIROC6  MRI-ESM2-0   \n",
       "\n",
       "     cs1   cs2   cs3   cs4  cs5  \n",
       "0    3.9  3.00  4.80  3.57  2.6  \n",
       "1    3.9  3.52  3.00  2.60  2.5  \n",
       "2    3.9  2.30  3.52  2.60  3.2  \n",
       "3    3.9  3.52  3.00  2.60  2.5  \n",
       "4    3.9  3.52  3.57  2.60  2.5  \n",
       "..   ...   ...   ...   ...  ...  \n",
       "115  3.9  3.52  2.60  3.00  2.5  \n",
       "116  3.9  3.52  2.60  3.00  2.5  \n",
       "117  4.7  3.90  3.00  3.52  2.6  \n",
       "118  3.9  3.52  3.57  2.60  2.5  \n",
       "119  3.9  2.30  3.00  2.60  3.2  \n",
       "\n",
       "[120 rows x 10 columns]"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "full_combos = pd.concat([esm_combos, cs_combos], axis=1)\n",
    "full_combos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.to_csv('/Users/prim232/gitrepos/stitches_in_r/R/inst/shinyApp/python_curation/misc_data/five_sample_esm_climate_sensitivity.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3.9 , 3.52, 3.  , 2.6 , 2.5 ])"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(sample_cs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.LineCollection at 0x1664da130>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "min_avg_diff = 1000\n",
    "for i in range(esm_combos.shape[0]):\n",
    "    esm_sample = esm_combos.iloc[i,:]\n",
    "    sample_cs = [climate_sensitives[esm_sample.values[j]] for j in range(5)]\n",
    "    avg_diff = np.sum((np.diff(np.sort(sample_cs))))\n",
    "\n",
    "    if avg_diff < min_avg_diff:\n",
    "        min_avg_diff = avg_diff\n",
    "        min_sample = esm_sample\n",
    "        min_sample_cs = sample_cs\n",
    "\n",
    "plt.scatter(y=np.random.normal(0, 0.05, len(min_sample_cs)), x = min_sample_cs, color='black', alpha=0.7) \n",
    "plt.vlines(x=[2, 2.5, 3, 4, 5], ymin=5, ymax=14, colors='gray', ls=':', lw=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# for i in range(esm_combos.shape[0]):\n",
    "for i in range(20):\n",
    "    esm_sample = esm_combos.iloc[i,:]\n",
    "    sample_cs = [climate_sensitives[esm_sample.values[j]] for j in range(5)]\n",
    "    plt.scatter(y=np.random.normal(i, 0, len(sample_cs)), x = sample_cs, color='black', alpha=0.7)\n",
    "    plt.vlines(x=[2, 2.5, 3, 4, 5], ymin=5, ymax=14, colors='gray', ls=':', lw=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[3.9, 3.0, 2.6, 3.2, 2.5]"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_cs = [climate_sensitives[esm_sample.values[j]] for j in range(5)]\n",
    "sample_cs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10,4))\n",
    "for n in range(5, 10, 2):\n",
    "    ssl, ssvl, _, _ = get_good_sample(n)\n",
    "    plt.scatter(y=np.random.normal(n, 0.00, ssvl.shape[0]), x = ssvl.climate_sensitivity, color = 'red', alpha=0.7)\n",
    "    plt.scatter(y=np.random.normal(n, 0.00, ssl.shape[0]), x = ssl.climate_sensitivity, color='black', alpha=0.7)\n",
    "    plt.vlines(x=[2, 2.5, 3, 4, 5], ymin=4.5, ymax=9.5, colors='gray', ls=':', lw=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE with all criteria samples: 0.6654777777777781\n"
     ]
    },
    {
     "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>climate_sensitivity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.050000</th>\n",
       "      <td>1.940000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.166667</th>\n",
       "      <td>2.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.500000</th>\n",
       "      <td>3.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.833333</th>\n",
       "      <td>4.733333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.950000</th>\n",
       "      <td>5.290000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          climate_sensitivity\n",
       "0.050000             1.940000\n",
       "0.166667             2.500000\n",
       "0.500000             3.200000\n",
       "0.833333             4.733333\n",
       "0.950000             5.290000"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "full_quants = climate_sensitivies_df.quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)['climate_sensitivity'].values\n",
    "full_mse = np.sum(np.square(ipcc_quants - full_quants))\n",
    "print(f'MSE with all criteria samples: {full_mse}')\n",
    "climate_sensitivies_df.quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE with all criteria samples: 2.4559999999999995\n",
      "              esm  climate_sensitivity        group\n",
      "0      ACCESS-CM2                  4.7  very likely\n",
      "1   ACCESS-ESM1-5                  3.9       likely\n",
      "3     CAMS-CSM1-0                  2.3  very likely\n",
      "4           CESM2                  5.2     unlikely\n",
      "5     CESM2-WACCM                  4.8  very likely\n",
      "8         CanESM5                  5.6     unlikely\n",
      "18      INM-CM5-0                  1.9     unlikely\n",
      "20   IPSL-CM6A-LR                  4.6  very likely\n",
      "22         MIROC6                  2.6       likely\n",
      "23  MPI-ESM1-2-HR                  3.0       likely\n",
      "24  MPI-ESM1-2-LR                  3.0       likely\n",
      "25     MRI-ESM2-0                  3.2       likely\n",
      "29    UKESM1-0-LL                  5.3     unlikely\n"
     ]
    },
    {
     "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>climate_sensitivity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.050000</th>\n",
       "      <td>2.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.166667</th>\n",
       "      <td>2.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.500000</th>\n",
       "      <td>3.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.833333</th>\n",
       "      <td>5.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.950000</th>\n",
       "      <td>5.42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          climate_sensitivity\n",
       "0.050000                 2.14\n",
       "0.166667                 2.60\n",
       "0.500000                 3.90\n",
       "0.833333                 5.20\n",
       "0.950000                 5.42"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_mse = included_cs_df.quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)['climate_sensitivity'].values\n",
    "all_mse = np.sum(np.square(ipcc_quants - all_mse))\n",
    "print(f'MSE with all criteria samples: {all_mse}')\n",
    "print(included_cs_df)\n",
    "included_cs_df.quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_err_eight = 10000\n",
    "best_index_eight = np.nan\n",
    "for index in list(combinations(included_cs_df.index,8)):\n",
    "    new_quants = included_cs_df.loc[index,:].quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)['climate_sensitivity'].values\n",
    "    mse = np.sum(np.square((new_quants - ipcc_quants)))\n",
    "    if mse < best_err_eight:\n",
    "        best_index_eight = index\n",
    "        best_err_eight = mse\n",
    "\n",
    "best_err_five = 10000\n",
    "best_index_five = np.nan\n",
    "for index in list(combinations(included_cs_df.index,5)):\n",
    "    new_quants = included_cs_df.loc[index,:].quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)['climate_sensitivity'].values\n",
    "    mse = np.sum(np.square((new_quants - ipcc_quants)))\n",
    "    if mse < best_err_five:\n",
    "        best_index_five = index\n",
    "        best_err_five = mse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE with 8 samples: 0.0710694444444444\n",
      "              esm  climate_sensitivity        group\n",
      "1   ACCESS-ESM1-5                  3.9       likely\n",
      "3     CAMS-CSM1-0                  2.3  very likely\n",
      "8         CanESM5                  5.6     unlikely\n",
      "18      INM-CM5-0                  1.9     unlikely\n",
      "22         MIROC6                  2.6       likely\n",
      "23  MPI-ESM1-2-HR                  3.0       likely\n",
      "24  MPI-ESM1-2-LR                  3.0       likely\n",
      "25     MRI-ESM2-0                  3.2       likely\n"
     ]
    },
    {
     "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>climate_sensitivity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.050000</th>\n",
       "      <td>2.040000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.166667</th>\n",
       "      <td>2.350000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.500000</th>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.833333</th>\n",
       "      <td>3.783333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.950000</th>\n",
       "      <td>5.005000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          climate_sensitivity\n",
       "0.050000             2.040000\n",
       "0.166667             2.350000\n",
       "0.500000             3.000000\n",
       "0.833333             3.783333\n",
       "0.950000             5.005000"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(f'MSE with 8 samples: {best_err_eight}')\n",
    "print(included_cs_df.loc[best_index_eight,:])\n",
    "included_cs_df.loc[best_index_eight,:].quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE with 5 samples: 0.03377777777777759\n",
      "              esm  climate_sensitivity     group\n",
      "8         CanESM5                  5.6  unlikely\n",
      "18      INM-CM5-0                  1.9  unlikely\n",
      "22         MIROC6                  2.6    likely\n",
      "23  MPI-ESM1-2-HR                  3.0    likely\n",
      "25     MRI-ESM2-0                  3.2    likely\n"
     ]
    },
    {
     "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>climate_sensitivity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.050000</th>\n",
       "      <td>2.040000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.166667</th>\n",
       "      <td>2.366667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.500000</th>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.833333</th>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.950000</th>\n",
       "      <td>5.120000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          climate_sensitivity\n",
       "0.050000             2.040000\n",
       "0.166667             2.366667\n",
       "0.500000             3.000000\n",
       "0.833333             4.000000\n",
       "0.950000             5.120000"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(f'MSE with 5 samples: {best_err_five}')\n",
    "print(included_cs_df.loc[best_index_five,:])\n",
    "included_cs_df.loc[best_index_five,:].quantile([0.05, 1/6, 0.5, 5/6, 0.95], numeric_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "included_esms = inclusion_df[inclusion_df['include']]['esm'].values\n",
    "included_cs_df = climate_sensitivies_df[np.isin(climate_sensitivies_df['esm'], included_esms)]\n",
    "eight_cs_df = included_cs_df.loc[best_index_eight,:]\n",
    "five_cs_df = included_cs_df.loc[best_index_five,:]\n",
    "\n",
    "fig, (ax1,ax2, ax3, ax4) = plt.subplots(1,4)\n",
    "sns.boxplot(ax = ax1, y='climate_sensitivity', data=climate_sensitivies_df, showfliers = False, color='#89a2c4')\n",
    "sns.swarmplot(ax = ax1, y='climate_sensitivity', data=climate_sensitivies_df, color='0.25')\n",
    "sns.boxplot(ax = ax2, y='climate_sensitivity', data=included_cs_df, showfliers = False, color='#89a2c4')\n",
    "sns.swarmplot(ax = ax2, y='climate_sensitivity', data=included_cs_df, color='0.25')\n",
    "sns.boxplot(ax = ax3, y='climate_sensitivity', data=eight_cs_df, showfliers = False, color='#89a2c4')\n",
    "sns.swarmplot(ax = ax3, y='climate_sensitivity', data=eight_cs_df, color='0.25')\n",
    "sns.boxplot(ax = ax4, y='climate_sensitivity', data=five_cs_df, showfliers = False, color='#89a2c4')\n",
    "sns.swarmplot(ax = ax4, y='climate_sensitivity', data=five_cs_df, color='0.25')\n",
    "\n",
    "ax1.set_title('All Models')\n",
    "ax2.set_title('Criteria Trimmed')\n",
    "ax3.set_title('Subset of Eight')\n",
    "ax4.set_title('Subset of Five')\n",
    "ax1.tick_params(axis='x', bottom=False, top=False, which='both', labelbottom=False)\n",
    "ax2.tick_params(axis='x', bottom=False, top=False, which='both', labelbottom=False)\n",
    "ax3.tick_params(axis='x', bottom=False, top=False, which='both', labelbottom=False)\n",
    "ax4.tick_params(axis='x', bottom=False, top=False, which='both', labelbottom=False)\n",
    "ax1.set_ylabel('Climate Sensitivity')\n",
    "ax2.set_ylabel('')\n",
    "ax3.set_ylabel('')\n",
    "ax4.set_ylabel('')\n",
    "\n",
    "fig.tight_layout(pad=1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function adapted from matplotlib.cbook\n",
    "def my_boxplot_stats(X, whis=1.5, bootstrap=None, labels=None,\n",
    "                  autorange=False, percents=[0.25, 0.75], whis_percents=[0.05,0.95]):\n",
    "\n",
    "    def _bootstrap_median(data, N=5000):\n",
    "        # determine 95% confidence intervals of the median\n",
    "        M = len(data)\n",
    "        percentiles = [2.5, 97.5]\n",
    "\n",
    "        bs_index = np.random.randint(M, size=(N, M))\n",
    "        bsData = data[bs_index]\n",
    "        estimate = np.median(bsData, axis=1, overwrite_input=True)\n",
    "\n",
    "        CI = np.percentile(estimate, percentiles)\n",
    "        return CI\n",
    "\n",
    "    def _compute_conf_interval(data, med, iqr, bootstrap):\n",
    "        if bootstrap is not None:\n",
    "            # Do a bootstrap estimate of notch locations.\n",
    "            # get conf. intervals around median\n",
    "            CI = _bootstrap_median(data, N=bootstrap)\n",
    "            notch_min = CI[0]\n",
    "            notch_max = CI[1]\n",
    "        else:\n",
    "\n",
    "            N = len(data)\n",
    "            notch_min = med - 1.57 * iqr / np.sqrt(N)\n",
    "            notch_max = med + 1.57 * iqr / np.sqrt(N)\n",
    "\n",
    "        return notch_min, notch_max\n",
    "\n",
    "    # output is a list of dicts\n",
    "    bxpstats = []\n",
    "\n",
    "    # convert X to a list of lists\n",
    "    X = _reshape_2D(X, \"X\")\n",
    "\n",
    "    # ncols = len(X)\n",
    "    # if labels is None:\n",
    "    #     labels = itertools.repeat(None)\n",
    "    # elif len(labels) != ncols:\n",
    "    #     raise ValueError(\"Dimensions of labels and X must be compatible\")\n",
    "\n",
    "    input_whis = whis\n",
    "    for ii, (x, label) in enumerate(zip(X, labels)):\n",
    "\n",
    "        # empty dict\n",
    "        stats = {}\n",
    "        if labels is not None:\n",
    "            stats['label'] = labels\n",
    "\n",
    "        # restore whis to the input values in case it got changed in the loop\n",
    "        whis = input_whis\n",
    "\n",
    "        # note tricksyness, append up here and then mutate below\n",
    "        bxpstats.append(stats)\n",
    "\n",
    "        # if empty, bail\n",
    "        if len(x) == 0:\n",
    "            stats['fliers'] = np.array([])\n",
    "            stats['mean'] = np.nan\n",
    "            stats['med'] = np.nan\n",
    "            stats['q1'] = np.nan\n",
    "            stats['q3'] = np.nan\n",
    "            stats['cilo'] = np.nan\n",
    "            stats['cihi'] = np.nan\n",
    "            stats['whislo'] = np.nan\n",
    "            stats['whishi'] = np.nan\n",
    "            stats['med'] = np.nan\n",
    "            continue\n",
    "\n",
    "        # up-convert to an array, just to be safe\n",
    "        x = np.asarray(x)\n",
    "\n",
    "        # arithmetic mean\n",
    "        stats['mean'] = np.mean(x)\n",
    "\n",
    "        # median\n",
    "        med = np.quantile(x, 0.5)\n",
    "        ## Altered line\n",
    "        q1, q3 = np.quantile(x, (percents[0], percents[1]))\n",
    "        # Whiskers\n",
    "        stats['whislo'] = np.quantile(x, whis_percents[0])\n",
    "        stats['whishi'] = np.quantile(x, whis_percents[1])\n",
    "\n",
    "        # interquartile range\n",
    "        stats['iqr'] = q3 - q1\n",
    "        if stats['iqr'] == 0 and autorange:\n",
    "            whis = 'range'\n",
    "\n",
    "        # conf. interval around median\n",
    "        stats['cilo'], stats['cihi'] = _compute_conf_interval(\n",
    "            x, med, stats['iqr'], bootstrap\n",
    "        )\n",
    "\n",
    "        # compute a single array of outliers\n",
    "        stats['fliers'] = np.hstack([\n",
    "            np.compress(x < stats['whislo'], x),\n",
    "            np.compress(x > stats['whishi'], x)\n",
    "        ])\n",
    "\n",
    "\n",
    "        # add in the remaining stats\n",
    "        stats['q1'], stats['med'], stats['q3'] = q1, med, q3\n",
    "\n",
    "    return bxpstats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "full_no_nan = climate_sensitivies_df['climate_sensitivity'].values[~np.isnan(climate_sensitivies_df['climate_sensitivity'].values)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "stats = {}\n",
    "stats['E'] = my_boxplot_stats(full_no_nan, labels='Available Models', percents=[1/6, 5/6])[0]\n",
    "stats['A'] = my_boxplot_stats(included_cs_df['climate_sensitivity'].values, labels='Meeting Criterion', percents=[1/6, 5/6])[0]\n",
    "stats['B'] = my_boxplot_stats(eight_cs_df['climate_sensitivity'].values, labels='Best Sample of 8', percents=[1/6, 5/6])[0]\n",
    "stats['C'] = my_boxplot_stats(five_cs_df['climate_sensitivity'].values, labels='Best Sample of 5', percents=[1/6, 5/6])[0]\n",
    "stats['D'] = stats['C'].copy()\n",
    "stats['D']['med'] = 3\n",
    "stats['D']['whishi'] = 5\n",
    "stats['D']['whislo'] = 2\n",
    "stats['D']['q1'] = 2.5\n",
    "stats['D']['q3'] = 4\n",
    "stats['D']['label'] = 'IPCC Ranges'\n",
    "\n",
    "fig, ax = plt.subplots(1, 1)\n",
    "\n",
    "widths = 0.4\n",
    "# Plot boxplots from our computed statistics\n",
    "mf = ax.bxp([stats['E'], stats['A'], stats['B'], stats['C'], stats['D']], positions=np.arange(0,5), showfliers=False, widths = widths, \n",
    "            patch_artist=True, vert=True)\n",
    "\n",
    "for patch in mf['boxes']:\n",
    "    patch.set_facecolor('black')\n",
    "for median in mf['medians']:\n",
    "    median.set_color('#94faff')\n",
    "    median.set_linewidth(3)\n",
    "\n",
    "plt.hlines(y=[2, 2.5, 3, 4, 5], xmin=0, xmax=4, colors='gray', ls=':', lw=1)\n",
    "\n",
    "alpha = 0.8\n",
    "sns.swarmplot(ax = ax, x=1.2, y = full_no_nan, color='#de1b3b', alpha=alpha)\n",
    "sns.swarmplot(ax = ax, x=2.2, y = 'climate_sensitivity', data=included_cs_df, color='#de1b3b', alpha=alpha)\n",
    "sns.swarmplot(ax = ax, x=3.2, y = 'climate_sensitivity', data=eight_cs_df, color='#de1b3b', alpha=alpha)\n",
    "sns.swarmplot(ax = ax, x=4.2, y = 'climate_sensitivity', data=five_cs_df, color='#de1b3b', alpha=alpha)\n",
    "\n",
    "plt.xticks(rotation = 30)\n",
    "\n",
    "ax.set_ylabel('Climate Sensitivity')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "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>esm</th>\n",
       "      <th>climate_sensitivity</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACCESS-CM2</td>\n",
       "      <td>4.7</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ACCESS-ESM1-5</td>\n",
       "      <td>3.9</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CAMS-CSM1-0</td>\n",
       "      <td>2.3</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CESM2</td>\n",
       "      <td>5.2</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>CESM2-WACCM</td>\n",
       "      <td>4.8</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>CanESM5</td>\n",
       "      <td>5.6</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>INM-CM5-0</td>\n",
       "      <td>1.9</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>IPSL-CM6A-LR</td>\n",
       "      <td>4.6</td>\n",
       "      <td>very likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>MIROC6</td>\n",
       "      <td>2.6</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>MPI-ESM1-2-HR</td>\n",
       "      <td>3.0</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>MPI-ESM1-2-LR</td>\n",
       "      <td>3.0</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>MRI-ESM2-0</td>\n",
       "      <td>3.2</td>\n",
       "      <td>likely</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>UKESM1-0-LL</td>\n",
       "      <td>5.3</td>\n",
       "      <td>unlikely</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              esm  climate_sensitivity        group\n",
       "0      ACCESS-CM2                  4.7  very likely\n",
       "1   ACCESS-ESM1-5                  3.9       likely\n",
       "3     CAMS-CSM1-0                  2.3  very likely\n",
       "4           CESM2                  5.2     unlikely\n",
       "5     CESM2-WACCM                  4.8  very likely\n",
       "8         CanESM5                  5.6     unlikely\n",
       "18      INM-CM5-0                  1.9     unlikely\n",
       "20   IPSL-CM6A-LR                  4.6  very likely\n",
       "22         MIROC6                  2.6       likely\n",
       "23  MPI-ESM1-2-HR                  3.0       likely\n",
       "24  MPI-ESM1-2-LR                  3.0       likely\n",
       "25     MRI-ESM2-0                  3.2       likely\n",
       "29    UKESM1-0-LL                  5.3     unlikely"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "included_cs_df"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.16 ('stitches_env')",
   "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": "3.9.16"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "230986d229b4a83088d849fb01e13746c2da10ab433b44ddeeba095e5efb26bd"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}