notebooks/scenarios.py.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6631e361",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pypsa\n",
    "import yaml\n",
    "import cartopy\n",
    "import sys\n",
    "import re\n",
    "import os\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import geopandas as gpd\n",
    "import xarray as xr\n",
    "import cartopy.crs as ccrs\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "import matplotlib as mpl\n",
    "\n",
    "from itertools import product\n",
    "from matplotlib.lines import Line2D\n",
    "from matplotlib.patches import FancyArrowPatch\n",
    "import matplotlib.patches as mpatches\n",
    "from matplotlib.transforms import Bbox\n",
    "\n",
    "from vresutils.costdata import annuity\n",
    "\n",
    "PATH = \"../workflows/pypsa-eur-sec/\"\n",
    "\n",
    "sys.path.append(os.path.join(PATH, \"scripts/\"))\n",
    "from plot_summary import rename_techs\n",
    "\n",
    "plt.style.use([\"bmh\", \"matplotlibrc\"])\n",
    "xr.set_options(display_style=\"html\")\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "87586e1c-4b6c-46bf-92a6-e49ea27c3165",
   "metadata": {},
   "outputs": [],
   "source": [
    "CLUSTERS = 181\n",
    "LV_OPTS = \"Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10\"\n",
    "OUTPUT = \"../results/graphics-20221227/\"\n",
    "\n",
    "MAIN_SCENARIOS = PATH + \"results/20221227-main\"\n",
    "DEC_SCENARIOS = PATH + \"results/20221227-decentral\"\n",
    "LV_SCENARIOS = PATH + \"results/20221227-lv\"\n",
    "ONW_SCENARIOS = PATH + \"results/20221227-onw\"\n",
    "GAS_SCENARIOS = PATH + \"results/20221227-gas\"\n",
    "IMP_SCENARIOS = PATH + \"results/20221227-import\"\n",
    "SHP_SCENARIOS = PATH + \"results/20221227-shipping\"\n",
    "COST_SCENARIOS = PATH + \"results/20221227-costs\"\n",
    "TIME_SCENARIOS = PATH + \"results/20221227-time\"\n",
    "SPACE_SCENARIOS = PATH + \"results/20221227-spatial\"\n",
    "OLD_SCENARIOS = \"../workflows-rev0/pypsa-eur-sec/results/20211218-181-h2\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "496f7566-e197-4b89-8cad-b21794126112",
   "metadata": {},
   "outputs": [],
   "source": [
    "latex = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6897226d-1d2d-4d99-83fc-b9b9686cab54",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(PATH + \"config.main.yaml\") as file:\n",
    "    config = yaml.safe_load(file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fc763ecb-1847-43b1-86e0-f1ee5a8a06fc",
   "metadata": {},
   "outputs": [],
   "source": [
    "tech_colors = config[\"plotting\"][\"tech_colors\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d35f982f-c5c2-4d37-8c81-b3edb4362eef",
   "metadata": {},
   "outputs": [],
   "source": [
    "tech_colors[\"battery electric vehicles\"] = tech_colors[\"BEV charger\"]\n",
    "tech_colors[\"other\"] = \"#454545\"\n",
    "tech_colors[\"building heat demand\"] = tech_colors[\"heat\"]\n",
    "tech_colors[\"ambient heat\"] = tech_colors[\"heat pump\"]\n",
    "tech_colors[\"residential electricity demand\"] = \"#72709c\"\n",
    "tech_colors[\"industry electricity demand\"] = tech_colors[\"electricity\"]\n",
    "tech_colors[\"hydrogen demand\"] = tech_colors[\"land transport fuel cell\"]\n",
    "tech_colors[\"agriculture machinery\"] = tech_colors[\"land transport fuel cell\"]\n",
    "tech_colors[\"methane demand\"] = tech_colors[\"helmeth\"]\n",
    "tech_colors[\"liquid hydrocarbon demand\"] = tech_colors[\"kerosene for aviation\"]\n",
    "tech_colors[\"aviation fuels\"] = tech_colors[\"kerosene for aviation\"]\n",
    "tech_colors[\"shipping fuels\"] = tech_colors[\"shipping methanol\"]\n",
    "tech_colors[\"biomass demand\"] = tech_colors[\"biogas\"]\n",
    "tech_colors[\"biogas upgrading\"] = tech_colors[\"biogas\"]\n",
    "tech_colors[\"hydrogen for industry\"] = tech_colors[\"H2 for industry\"]\n",
    "tech_colors[\"hydrogen-to-power/heat\"] = tech_colors[\"gas-to-power/heat\"]\n",
    "tech_colors[\"hydrogen for land transport\"] = \"#8487e8\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b24c98bb-757b-4e73-8ad7-be5a48c075a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "if not os.path.exists(OUTPUT):\n",
    "    os.makedirs(OUTPUT)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "97dd1be2-2218-40a4-8799-16ddd3d78889",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rename_techs_tyndp(tech):\n",
    "    tech = rename_techs(tech)\n",
    "    if \"heat pump\" in tech or \"resistive heater\" in tech:\n",
    "        return \"power-to-heat\"\n",
    "    elif tech in [\"H2 Electrolysis\"]:  # , \"H2 liquefaction\"]:\n",
    "        return \"power-to-hydrogen\"\n",
    "    elif \"H2 pipeline\" in tech:\n",
    "        return \"H2 pipeline\"\n",
    "    elif tech == \"H2\":\n",
    "        return \"H2 storage\"\n",
    "    elif tech in [\"OCGT\", \"CHP\", \"gas boiler\", \"H2 Fuel Cell\"]:\n",
    "        return \"gas-to-power/heat\"\n",
    "    # elif \"solar\" in tech:\n",
    "    #    return \"solar\"\n",
    "    elif tech in [\"Fischer-Tropsch\", \"methanolisation\"]:\n",
    "        return \"power-to-liquid\"\n",
    "    elif \"offshore wind\" in tech:\n",
    "        return \"offshore wind\"\n",
    "    elif \"SMR\" in tech:\n",
    "        return tech.replace(\"SMR\", \"steam methane reforming\")\n",
    "    elif \"DAC\" in tech:\n",
    "        return \"direct air capture\"\n",
    "    elif \"CC\" in tech or \"sequestration\" in tech:\n",
    "        return \"carbon capture\"\n",
    "    elif tech == \"oil\" or tech == \"gas\":\n",
    "        return \"fossil oil and gas\"\n",
    "    else:\n",
    "        return tech"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2570121-81a7-4107-a6fd-d50dcc94617e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rename_techs_balances(tech):\n",
    "    tech = rename_techs(tech)\n",
    "    if \"heat pump\" in tech:\n",
    "        return \"ambient heat\"\n",
    "    elif tech in [\"H2 Electrolysis\"]:  # , \"H2 liquefaction\"]:\n",
    "        return \"power-to-hydrogen\"\n",
    "    elif \"solar\" in tech:\n",
    "        return \"solar\"\n",
    "    elif tech in [\"Fischer-Tropsch\", \"methanolisation\"]:\n",
    "        return \"power-to-liquid\"\n",
    "    elif tech == \"DAC\":\n",
    "        return \"direct air capture\"\n",
    "    elif \"offshore wind\" in tech:\n",
    "        return \"offshore wind\"\n",
    "    elif tech == \"oil\" or tech == \"gas\":\n",
    "        return \"fossil oil and gas\"\n",
    "    elif tech in [\"BEV charger\", \"V2G\", \"Li ion\", \"land transport EV\"]:\n",
    "        return \"battery electric vehicles\"\n",
    "    elif tech in [\"biogas\", \"solid biomass\"]:\n",
    "        return \"biomass\"\n",
    "    elif tech in [\"electricity\"]:\n",
    "        return \"residential electricity demand\"\n",
    "    elif tech in [\"industry electricity\", \"agriculture electricity\"]:\n",
    "        return \"industry electricity demand\"\n",
    "    elif tech in [\"agriculture heat\", \"heat\", \"low-temperature heat for industry\"]:\n",
    "        return \"heat demand\"\n",
    "    elif \"solid biomass for industry\" in tech:\n",
    "        return \"biomass demand\"\n",
    "    elif \"gas for industry\" in tech:\n",
    "        return \"methane demand\"\n",
    "    elif tech in [\"H2 for industry\", \"land transport fuel cell\"]:\n",
    "        return \"hydrogen demand\"\n",
    "    elif tech in [\n",
    "        \"kerosene for aviation\",\n",
    "        \"naphtha for industry\",\n",
    "        \"shipping methanol\",\n",
    "        \"agriculture machinery oil\",\n",
    "    ]:\n",
    "        return \"liquid hydrocarbon demand\"\n",
    "    elif tech in [\n",
    "        \"transmission lines\",\n",
    "        \"H2 pipeline\",\n",
    "        \"H2 pipeline retrofitted\",\n",
    "        \"H2\",\n",
    "        \"electricity distribution grid\",\n",
    "        \"SMR\",\n",
    "        \"SMR CC\",\n",
    "        \"OCGT\",\n",
    "        \"CHP\",\n",
    "        \"gas boiler\",\n",
    "        \"H2 Fuel Cell\",\n",
    "        \"resistive heater\",\n",
    "        \"battery storage\",\n",
    "        \"methanation\",\n",
    "    ]:\n",
    "        return \"other\"\n",
    "    else:\n",
    "        return tech"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e30b1a58-d4d2-4d69-a324-e8b6fe463305",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rename_techs_carbon_balances(tech):\n",
    "    prefix_to_remove = [\n",
    "        \"residential \",\n",
    "        \"services \",\n",
    "        \"urban \",\n",
    "        \"rural \",\n",
    "        \"central \",\n",
    "        \"decentral \",\n",
    "    ]\n",
    "    for ptr in prefix_to_remove:\n",
    "        if tech[: len(ptr)] == ptr:\n",
    "            tech = tech[len(ptr) :]\n",
    "    if tech == \"biogas to gas\":\n",
    "        return \"biogas upgrading\"\n",
    "    elif tech == \"agriculture machinery oil emissions\":\n",
    "        return \"agriculture machinery\"\n",
    "    elif tech == \"shipping methanol emissions\":\n",
    "        return \"shipping fuels\"\n",
    "    elif tech == \"DAC\":\n",
    "        return \"direct air capture\"\n",
    "    elif \"SMR\" in tech:\n",
    "        return tech.replace(\"SMR\", \"steam methane reforming\")\n",
    "    else:\n",
    "        return tech"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "143ef3c4-7385-4480-a870-1f3f5f9cd278",
   "metadata": {},
   "outputs": [],
   "source": [
    "preferred_order = pd.Index(\n",
    "    [\n",
    "        \"transmission lines\",\n",
    "        \"electricity distribution grid\",\n",
    "        \"fossil oil and gas\",\n",
    "        \"hydroelectricity\",\n",
    "        \"hydro reservoir\",\n",
    "        \"run of river\",\n",
    "        \"pumped hydro storage\",\n",
    "        \"solid biomass\",\n",
    "        \"biogas\",\n",
    "        \"onshore wind\",\n",
    "        \"offshore wind\",\n",
    "        \"offshore wind (AC)\",\n",
    "        \"offshore wind (DC)\",\n",
    "        \"solar PV\",\n",
    "        \"solar thermal\",\n",
    "        \"solar rooftop\",\n",
    "        \"solar\",\n",
    "        \"building retrofitting\",\n",
    "        \"ground heat pump\",\n",
    "        \"air heat pump\",\n",
    "        \"heat pump\",\n",
    "        \"resistive heater\",\n",
    "        \"power-to-heat\",\n",
    "        \"gas-to-power/heat\",\n",
    "        \"CHP\",\n",
    "        \"OCGT\",\n",
    "        \"gas boiler\",\n",
    "        \"gas\",\n",
    "        \"natural gas\",\n",
    "        \"helmeth\",\n",
    "        \"methanation\",\n",
    "        \"power-to-gas\",\n",
    "        \"power-to-hydrogen\",\n",
    "        \"H2 pipeline\",\n",
    "        \"H2 liquefaction\",\n",
    "        \"H2 storage\",\n",
    "        \"hydrogen storage\",\n",
    "        \"power-to-liquid\",\n",
    "        \"battery storage\",\n",
    "        \"hot water storage\",\n",
    "        \"CO2 sequestration\",\n",
    "        \"CCS\",\n",
    "        \"carbon capture and sequestration\",\n",
    "        \"DAC\",\n",
    "        \"direct air capture\",\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "935bad6a-3c7d-4ab5-a18b-88f0bea24505",
   "metadata": {},
   "outputs": [],
   "source": [
    "def parse_index(c, with_resolution=False):\n",
    "    clusters = c[0]\n",
    "\n",
    "    lv = c[1]\n",
    "\n",
    "    match = re.search(r\"onwind\\+p([0-9.]*)\", c[2])\n",
    "    onw = 100.0 if match is None else 100 * float(match.groups()[0])\n",
    "\n",
    "    h2 = \"no H2 grid\" if \"noH2network\" in c[2] else \"H2 grid\"\n",
    "\n",
    "    to_return = (clusters, lv, onw, h2)\n",
    "\n",
    "    if with_resolution:\n",
    "        match = re.findall(r\"(\\d+)H\", c[2])\n",
    "        to_return += (int(match[0]),)\n",
    "\n",
    "    return to_return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "436112de-0bd3-4e27-b508-1174e1dacc64",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_decentral():\n",
    "    costs = pd.read_csv(\n",
    "        DEC_SCENARIOS + \"/csvs/costs.csv\", header=[0, 1, 2, 3], index_col=[0, 1, 2]\n",
    "    )\n",
    "\n",
    "    costs = costs.xs(\"2050\", level=\"planning_horizon\", axis=1)\n",
    "\n",
    "    costs.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c) for c in costs.columns], names=[\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    )\n",
    "\n",
    "    costs = costs.xs(str(CLUSTERS), level=\"clusters\", axis=1)\n",
    "\n",
    "    df = costs.groupby(level=2).sum().div(1e9)\n",
    "\n",
    "    df = df.groupby(df.index.map(rename_techs_tyndp)).sum()\n",
    "\n",
    "    df = df.xs(100, level=\"onw\", axis=1, drop_level=False)\n",
    "\n",
    "    df.sum() / df.sum().min()\n",
    "\n",
    "    to_drop = df.index[df.max(axis=1).fillna(0.0) < 1.2]\n",
    "    print(to_drop)\n",
    "    df.drop(to_drop, inplace=True)\n",
    "\n",
    "    order = preferred_order.intersection(df.index).append(\n",
    "        df.index.difference(preferred_order)\n",
    "    )\n",
    "    df = df.loc[order]\n",
    "\n",
    "    tech_colors = config[\"plotting\"][\"tech_colors\"]\n",
    "    colors = [tech_colors[i] for i in df.index]\n",
    "\n",
    "    df.columns = df.columns.get_level_values(2)\n",
    "    df.columns.name = \"\"\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "15323752-e515-4fb8-8084-baaea4e80a4e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "19d7d348-786c-45e1-97a2-cd7b94f0fa52",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_main(\n",
    "    scenarios=None, clusters=None, rename=True, with_resolution=False, with_space=False\n",
    "):\n",
    "    if scenarios is None:\n",
    "        scenarios = MAIN_SCENARIOS\n",
    "\n",
    "    if clusters is None:\n",
    "        clusters = CLUSTERS\n",
    "\n",
    "    horizon = \"2030\" if \"rev0\" in scenarios else \"2050\"\n",
    "\n",
    "    costs = pd.read_csv(\n",
    "        scenarios + f\"/csvs/costs.csv\", header=[0, 1, 2, 3], index_col=[0, 1, 2]\n",
    "    )\n",
    "\n",
    "    costs = costs.xs(horizon, level=\"planning_horizon\", axis=1)\n",
    "\n",
    "    names = [\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    if with_resolution:\n",
    "        names += (\"res\",)\n",
    "\n",
    "    costs.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c, with_resolution) for c in costs.columns], names=names\n",
    "    )\n",
    "\n",
    "    if not with_space:\n",
    "        costs = costs.xs(str(clusters), level=\"clusters\", axis=1)\n",
    "\n",
    "    df = costs.groupby(level=2).sum().div(1e9)\n",
    "\n",
    "    if rename:\n",
    "        df = df.groupby(df.index.map(rename_techs_tyndp)).sum()\n",
    "\n",
    "    to_drop = df.index[df.max(axis=1).fillna(0.0) < 1.2]\n",
    "    print(to_drop)\n",
    "    df.drop(to_drop, inplace=True)\n",
    "\n",
    "    order = preferred_order.intersection(df.index).append(\n",
    "        df.index.difference(preferred_order)\n",
    "    )\n",
    "    df = df.loc[order]\n",
    "\n",
    "    if \"-imp\" in scenarios:\n",
    "        # imports for methanol, kerosene and naphtha at 120 €/MWh\n",
    "        print(\"add import costs\")\n",
    "        df.loc[\"green e-fuel imports\"] = (1026.64 + 546.36) * 120e6 / 1e9  # bn€/a\n",
    "        tech_colors[\"green e-fuel imports\"] = \"#46caf0\"\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5366bedc-c76f-4309-83f7-55afeb7df6b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_main_capacities(\n",
    "    scenarios=None,\n",
    "    clusters=None,\n",
    "    rename=True,\n",
    "    with_resolution=False,\n",
    "    with_space=False,\n",
    "    merge=True,\n",
    "):\n",
    "    if scenarios is None:\n",
    "        scenarios = MAIN_SCENARIOS\n",
    "\n",
    "    if clusters is None:\n",
    "        clusters = CLUSTERS\n",
    "\n",
    "    horizon = \"2030\" if \"rev0\" in scenarios else \"2050\"\n",
    "\n",
    "    df = pd.read_csv(\n",
    "        scenarios + f\"/csvs/capacities.csv\", header=[0, 1, 2, 3], index_col=[0, 1]\n",
    "    )\n",
    "\n",
    "    df = df.xs(horizon, level=\"planning_horizon\", axis=1)\n",
    "\n",
    "    names = [\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    if with_resolution:\n",
    "        names += (\"res\",)\n",
    "\n",
    "    df.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c, with_resolution) for c in df.columns], names=names\n",
    "    )\n",
    "\n",
    "    if not with_space:\n",
    "        df = df.xs(str(clusters), level=\"clusters\", axis=1)\n",
    "\n",
    "    if rename:\n",
    "        grouper = [\n",
    "            df.index.get_level_values(0),\n",
    "            df.index.get_level_values(1).map(rename_techs_tyndp),\n",
    "        ]\n",
    "        df = df.groupby(grouper).sum()\n",
    "\n",
    "    to_drop = df.index[df.max(axis=1).fillna(0.0) < 10]\n",
    "    df.drop(to_drop, inplace=True)\n",
    "\n",
    "    twh = df.xs(\"stores\", level=0).div(1e6)  # TWh\n",
    "\n",
    "    to_drop = [\n",
    "        \"CCS\",\n",
    "        \"biogas\",\n",
    "        \"co2\",\n",
    "        \"fossil oil and gas\",\n",
    "        \"solid biomass\",\n",
    "    ]\n",
    "    twh.drop(twh.index.intersection(to_drop), inplace=True)\n",
    "\n",
    "    gw = df.drop([\"stores\", \"lines\"]).div(1e3)  # GW\n",
    "\n",
    "    if merge:\n",
    "        gw = gw.groupby(level=1).sum()\n",
    "        techs = gw.index\n",
    "        kwargs = dict()\n",
    "    else:\n",
    "        techs = gw.index.levels[1]\n",
    "        kwargs = dict(level=1)\n",
    "\n",
    "    to_drop = [\n",
    "        \"fossil oil and gas\",\n",
    "        \"transmission lines\",\n",
    "        \"DAC\",\n",
    "        \"direct air capture\",\n",
    "        \"H2 pipeline\",\n",
    "        \"H2 pipeline retrofitted\",\n",
    "        \"CCS\",\n",
    "        \"carbon capture\" \"biogas\",\n",
    "        \"gas for industry\",\n",
    "        \"hot water storage\",\n",
    "        \"solid biomass for industry\",\n",
    "        \"process emissions\",\n",
    "    ]\n",
    "    gw.drop(techs.intersection(to_drop), **kwargs, inplace=True)\n",
    "\n",
    "    return gw, twh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "619d0b63-3392-4ad3-8c79-78b2e56fe41e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_main_energy(\n",
    "    scenarios=None, clusters=None, rename=True, with_resolution=False, with_space=False\n",
    "):\n",
    "    if scenarios is None:\n",
    "        scenarios = MAIN_SCENARIOS\n",
    "\n",
    "    if clusters is None:\n",
    "        clusters = CLUSTERS\n",
    "\n",
    "    horizon = \"2030\" if \"rev0\" in scenarios else \"2050\"\n",
    "\n",
    "    df = pd.read_csv(\n",
    "        scenarios + f\"/csvs/energy.csv\", header=[0, 1, 2, 3], index_col=[0, 1]\n",
    "    )\n",
    "\n",
    "    df = df.xs(horizon, level=\"planning_horizon\", axis=1)\n",
    "\n",
    "    names = [\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    if with_resolution:\n",
    "        names += (\"res\",)\n",
    "\n",
    "    df.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c, with_resolution) for c in df.columns], names=names\n",
    "    )\n",
    "\n",
    "    if not with_space:\n",
    "        df = df.xs(str(clusters), level=\"clusters\", axis=1)\n",
    "\n",
    "    df = df.groupby(level=1).sum().div(1e6)  # TWh\n",
    "\n",
    "    if rename:\n",
    "        df = df.groupby(df.index.map(rename_techs_tyndp)).sum()\n",
    "\n",
    "    to_drop = df.index[df.abs().max(axis=1).fillna(0.0) < 10]\n",
    "    df.drop(to_drop, inplace=True)\n",
    "\n",
    "    order = preferred_order.intersection(df.index).append(\n",
    "        df.index.difference(preferred_order)\n",
    "    )\n",
    "    df = df.loc[order]\n",
    "\n",
    "    if \"-imp\" in scenarios:\n",
    "        # imports for methanol, kerosene and naphtha\n",
    "        df.loc[\"green e-fuel imports\"] = 1026.64 + 546.36  # TWh\n",
    "        tech_colors[\"green e-fuel imports\"] = \"#46caf0\"\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "14cfd90e-8380-43b4-b0ab-1e652bc17c99",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_main_cfs(\n",
    "    scenarios=None, clusters=None, with_resolution=False, with_space=False\n",
    "):\n",
    "    if scenarios is None:\n",
    "        scenarios = MAIN_SCENARIOS\n",
    "\n",
    "    if clusters is None:\n",
    "        clusters = CLUSTERS\n",
    "\n",
    "    horizon = \"2030\" if \"rev0\" in scenarios else \"2050\"\n",
    "\n",
    "    df = pd.read_csv(\n",
    "        scenarios + f\"/csvs/cfs.csv\", header=[0, 1, 2, 3], index_col=[0, 1]\n",
    "    )\n",
    "\n",
    "    df = df.xs(horizon, level=\"planning_horizon\", axis=1)\n",
    "\n",
    "    names = [\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    if with_resolution:\n",
    "        names += (\"res\",)\n",
    "\n",
    "    df.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c, with_resolution) for c in df.columns], names=names\n",
    "    )\n",
    "\n",
    "    if not with_space:\n",
    "        df = df.xs(str(clusters), level=\"clusters\", axis=1)\n",
    "\n",
    "    df = df.groupby(level=1).sum()\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "66230d82-ad4a-4707-97a9-1bdd1e05bb51",
   "metadata": {},
   "outputs": [],
   "source": [
    "def fix_df(df):\n",
    "    column_dict = {\n",
    "        \"1.0\": \"w/o power grid expansion\",\n",
    "        \"opt\": \"w power grid expansion\",\n",
    "        \"H2 grid\": \"w hydrogen network\",\n",
    "        \"no H2 grid\": \"w/o hydrogen network\",\n",
    "    }\n",
    "    df.rename(columns=column_dict, inplace=True)\n",
    "    df.columns = [\"\\n\".join(col).strip() for col in df.columns.values]\n",
    "    df.sort_index(axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "109256d6-c557-468c-9653-6c780bfbd788",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_main_supply_energy(\n",
    "    scenarios=None,\n",
    "    clusters=None,\n",
    "    rename=True,\n",
    "    with_resolution=False,\n",
    "    with_space=False,\n",
    "    carrier=\"energy\",\n",
    "):\n",
    "    if scenarios is None:\n",
    "        scenarios = MAIN_SCENARIOS\n",
    "\n",
    "    if clusters is None:\n",
    "        clusters = CLUSTERS\n",
    "\n",
    "    horizon = \"2030\" if \"rev0\" in scenarios else \"2050\"\n",
    "\n",
    "    df = pd.read_csv(\n",
    "        scenarios + \"/csvs/supply_energy.csv\", index_col=[0, 1, 2], header=[0, 1, 2, 3]\n",
    "    )\n",
    "\n",
    "    co2_carriers = [\"co2\", \"co2 stored\", \"process emissions\"]\n",
    "    if carrier == \"energy\":\n",
    "        carrier = [i for i in df.index.levels[0] if i not in co2_carriers]\n",
    "\n",
    "    df = df.loc[carrier].groupby(level=2).sum().div(1e6)  # TWh / MtCO2\n",
    "    df.index = [\n",
    "        i[:-1]\n",
    "        if ((i not in [\"co2\", \"NH3\", \"H2\"]) and (i[-1:] in [\"0\", \"1\", \"2\", \"3\"]))\n",
    "        else i\n",
    "        for i in df.index\n",
    "    ]\n",
    "\n",
    "    df = df.xs(horizon, level=\"planning_horizon\", axis=1)\n",
    "\n",
    "    names = [\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    if with_resolution:\n",
    "        names += (\"res\",)\n",
    "\n",
    "    df.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c, with_resolution) for c in df.columns], names=names\n",
    "    )\n",
    "\n",
    "    if not with_space:\n",
    "        df = df.xs(str(clusters), level=\"clusters\", axis=1)\n",
    "\n",
    "    if rename or callable(rename):\n",
    "        func = rename if callable(rename) else rename_techs_tyndp\n",
    "        df = df.groupby(df.index.map(func)).sum()\n",
    "\n",
    "    to_drop = df.index[df.abs().max(axis=1).fillna(0.0) < 10]\n",
    "    df.drop(to_drop, inplace=True)\n",
    "\n",
    "    order = preferred_order.intersection(df.index).append(\n",
    "        df.index.difference(preferred_order)\n",
    "    )\n",
    "    df = df.loc[order]\n",
    "\n",
    "    if \"-imp\" in scenarios and carrier == \"energy\":\n",
    "        # imports for methanol, kerosene and naphtha\n",
    "        df.loc[\"green e-fuel imports\"] = 1026.64 + 546.36  # TWh\n",
    "        tech_colors[\"green e-fuel imports\"] = \"#46caf0\"\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1d2f9460-3acb-464e-a1e4-71353e804cc6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def energy_balances(scenarios, onw=100):\n",
    "    co2_carriers = [\"co2\", \"co2 stored\", \"process emissions\"]\n",
    "\n",
    "    balances_df = pd.read_csv(\n",
    "        scenarios + \"/csvs/supply_energy.csv\", index_col=[0, 1, 2], header=[0, 1, 2, 3]\n",
    "    )\n",
    "\n",
    "    balances = {i.replace(\" \", \"_\"): [i] for i in balances_df.index.levels[0]}\n",
    "    balances[\"energy\"] = [\n",
    "        i for i in balances_df.index.levels[0] if i not in co2_carriers\n",
    "    ]\n",
    "    balances[\"carbon\"] = [i for i in balances_df.index.levels[0] if i in co2_carriers]\n",
    "\n",
    "    key = \"energy\"\n",
    "\n",
    "    df = balances_df.loc[balances[key]]\n",
    "\n",
    "    df = df.groupby(level=2).sum().div(1e6)\n",
    "\n",
    "    df.index = [\n",
    "        i[:-1]\n",
    "        if ((i not in [\"co2\", \"NH3\", \"H2\"]) and (i[-1:] in [\"0\", \"1\", \"2\", \"3\"]))\n",
    "        else i\n",
    "        for i in df.index\n",
    "    ]\n",
    "\n",
    "    df = df.groupby(rename_techs_balances).sum()\n",
    "\n",
    "    df.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c) for c in df.columns], names=[\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    )\n",
    "\n",
    "    df = df.xs((onw, str(181)), level=[\"onw\", \"clusters\"], axis=1)\n",
    "\n",
    "    order = pd.Index(\n",
    "        [\n",
    "            \"fossil oil and gas\",\n",
    "            \"hydroelectricity\",\n",
    "            \"biomass\",\n",
    "            \"offshore wind\",\n",
    "            \"onshore wind\",\n",
    "            \"solar\",\n",
    "            \"ambient heat\",\n",
    "            \"residential electricity demand\",\n",
    "            \"industry electricity demand\",\n",
    "            \"electricity demand\",\n",
    "            \"battery electric vehicles\",\n",
    "            \"heat demand\",\n",
    "            \"hydrogen demand\",\n",
    "            \"biomass demand\",\n",
    "            \"methane demand\",\n",
    "            \"liquid hydrocarbon demand\",\n",
    "            \"power-to-liquid\",\n",
    "            \"methanation\",\n",
    "            \"power-to-hydrogen\",\n",
    "            \"hot water storage\",\n",
    "            \"direct air capture\",\n",
    "            \"other\",\n",
    "        ]\n",
    "    )\n",
    "\n",
    "    order = order.intersection(df.index).append(df.index.difference(order))\n",
    "    df = df.loc[order]\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a80c3cba-dfd1-4eaa-8307-470367e20dff",
   "metadata": {},
   "outputs": [],
   "source": [
    "def carbon_balances(scenarios, onw=100):\n",
    "    co2_carriers = [\"co2\", \"co2 stored\", \"process emissions\"]\n",
    "\n",
    "    balances_df = pd.read_csv(\n",
    "        scenarios + \"/csvs/supply_energy.csv\", index_col=[0, 1, 2], header=[0, 1, 2, 3]\n",
    "    )\n",
    "\n",
    "    balances = {i.replace(\" \", \"_\"): [i] for i in balances_df.index.levels[0]}\n",
    "    balances[\"energy\"] = [\n",
    "        i for i in balances_df.index.levels[0] if i not in co2_carriers\n",
    "    ]\n",
    "    balances[\"carbon\"] = [i for i in balances_df.index.levels[0] if i in co2_carriers]\n",
    "\n",
    "    key = \"co2\"\n",
    "\n",
    "    df = balances_df.loc[balances[key]]\n",
    "\n",
    "    df = df.groupby(level=2).sum().div(1e6)\n",
    "\n",
    "    df.index = [\n",
    "        i[:-1]\n",
    "        if ((i not in [\"co2\", \"NH3\", \"H2\"]) and (i[-1:] in [\"0\", \"1\", \"2\", \"3\"]))\n",
    "        else i\n",
    "        for i in df.index\n",
    "    ]\n",
    "\n",
    "    df = df.groupby(rename_techs_carbon_balances).sum()\n",
    "\n",
    "    df.columns = pd.MultiIndex.from_tuples(\n",
    "        [parse_index(c) for c in df.columns], names=[\"clusters\", \"lv\", \"onw\", \"h2\"]\n",
    "    )\n",
    "\n",
    "    df = df.xs((onw, str(181)), level=[\"onw\", \"clusters\"], axis=1)\n",
    "\n",
    "    df.drop(\"co2\", inplace=True)\n",
    "\n",
    "    order = pd.Index(\n",
    "        [\n",
    "            \"liquid hydrocarbons emissions\",\n",
    "            \"methanol emissions\",\n",
    "            \"process emissions CC\",\n",
    "            \"gas for industry CC\",\n",
    "            \"gas CHP CC\",\n",
    "            \"gas CHP\",\n",
    "            \"OCGT\",\n",
    "            \"gas boiler\",\n",
    "            \"steam methane reforming\",\n",
    "            \"steam methane reforming CC\",\n",
    "            \"biogas upgrading\",\n",
    "            \"solid biomass CHP CC\",\n",
    "            \"solid biomass for industry CC\",\n",
    "            \"direct air capture\",\n",
    "        ]\n",
    "    )\n",
    "\n",
    "    order = order.intersection(df.index).append(df.index.difference(order))\n",
    "    df = df.loc[order]\n",
    "\n",
    "    df = df.loc[df.abs().max(axis=1) > 0.01]\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "23d06d7b-e848-4dbe-8bbf-4573d1f78492",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rename_techs_h2_balances(tech):\n",
    "    if tech == \"H2 for industry\":\n",
    "        return \"hydrogen for industry\"\n",
    "    elif tech == \"Sabatier\":\n",
    "        return \"methanation\"\n",
    "    elif tech == \"H2 Electrolysis\":\n",
    "        return \"power-to-hydrogen\"\n",
    "    elif tech == \"land transport fuel cell\":\n",
    "        return \"hydrogen for land transport\"\n",
    "    elif tech == \"H2 Fuel Cell\":\n",
    "        return \"hydrogen-to-power/heat\"\n",
    "    elif \"SMR\" in tech:\n",
    "        return tech.replace(\"SMR\", \"steam methane reforming\")\n",
    "    else:\n",
    "        return tech"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19a5fc9d-5021-4cd8-b29e-24f98f9471b8",
   "metadata": {},
   "source": [
    "## Dashboard Export"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "431e7eed-4e31-4a76-b788-8197c2bed274",
   "metadata": {},
   "outputs": [],
   "source": [
    "SCENARIOS = {\n",
    "    (0, 0, 0, 0): (MAIN_SCENARIOS, 100),\n",
    "    (1, 0, 0, 0): (COST_SCENARIOS, 100),\n",
    "    (0, 1, 0, 0): (IMP_SCENARIOS, 100),\n",
    "    (0, 0, 1, 0): (SHP_SCENARIOS, 100),\n",
    "    (0, 0, 0, 1): (MAIN_SCENARIOS, 0),\n",
    "}\n",
    "\n",
    "NAMES = [\"optimistic_costs\", \"imports\", \"hydrogen_in_shipping\", \"no_onwind\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "11af736d-64d4-45ef-94d0-dbeda3579812",
   "metadata": {},
   "outputs": [],
   "source": [
    "tsc = pd.concat(\n",
    "    {\n",
    "        k: load_main(scenarios).xs(onw, level=\"onw\", axis=1)\n",
    "        for k, (scenarios, onw) in SCENARIOS.items()\n",
    "    },\n",
    "    names=NAMES,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f2a3cbef-822a-4734-9638-88fd84108631",
   "metadata": {},
   "outputs": [],
   "source": [
    "tsc.index.names = tsc.index.names[:-1] + [\"carrier\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "706599f1-ca28-4f75-8be1-e2174d0d1dff",
   "metadata": {},
   "outputs": [],
   "source": [
    "tsc = tsc.stack([0, 1]).to_xarray()\n",
    "tsc.name = \"costs\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e6f813fc-db25-4752-b673-f9218a442b87",
   "metadata": {},
   "outputs": [],
   "source": [
    "energy = pd.concat(\n",
    "    {\n",
    "        k: load_main_supply_energy(scenarios).xs(onw, level=\"onw\", axis=1)\n",
    "        for k, (scenarios, onw) in SCENARIOS.items()\n",
    "    },\n",
    "    names=NAMES,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9f17b62b-a48b-45df-969d-11cdf3bd28ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "energy.index.names = energy.index.names[:-1] + [\"carrier\"]\n",
    "energy = energy.stack([0, 1]).to_xarray()\n",
    "energy.name = \"energy\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8aff7fe0-6866-485b-beb3-11fcebc899f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "co2 = pd.concat(\n",
    "    {k: carbon_balances(scenarios, onw) for k, (scenarios, onw) in SCENARIOS.items()},\n",
    "    names=NAMES,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8efb7fc5-3a6f-448c-91dc-fc9bbf7c0684",
   "metadata": {},
   "outputs": [],
   "source": [
    "co2.index.names = co2.index.names[:-1] + [\"carrier\"]\n",
    "co2 = co2.stack([0, 1]).to_xarray()\n",
    "co2.name = \"co2\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eddbf8ad-273a-4dd3-8870-feea847c9532",
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_capacities(scenarios, onw):\n",
    "    gw, twh = load_main_capacities(scenarios, merge=False)\n",
    "\n",
    "    gw = gw.xs(onw, level=\"onw\", axis=1)\n",
    "    twh = twh.xs(onw, level=\"onw\", axis=1)\n",
    "\n",
    "    gen = gw.loc[[\"generators\", \"storage_units\"]].groupby(level=1).sum()\n",
    "\n",
    "    con = gw.loc[[\"links\"]].groupby(level=1).sum()\n",
    "    return gen, con, twh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "db38475f-4105-4031-b6a4-fec7b3c12a2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "cap = pd.concat(\n",
    "    {\n",
    "        k: pd.concat(\n",
    "            read_capacities(scenarios, onw),\n",
    "            keys=[\"generation\", \"conversion\", \"storage\"],\n",
    "        )\n",
    "        for k, (scenarios, onw) in SCENARIOS.items()\n",
    "    },\n",
    "    names=NAMES,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c1b93297-d5e7-4b01-bfc5-ae0522fe3ab3",
   "metadata": {},
   "outputs": [],
   "source": [
    "cap.index.names = cap.index.names[:-2] + [\"category\", \"carrier\"]\n",
    "cap = cap.stack([0, 1]).unstack(\"category\").to_xarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "64c5da46-6cea-40e0-b47f-505d89a9513a",
   "metadata": {},
   "outputs": [],
   "source": [
    "ds = xr.merge([tsc, energy, co2, cap]).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "563f4833-dd1e-416b-a297-b4092e4e074c",
   "metadata": {},
   "outputs": [],
   "source": [
    "comp = dict(zlib=True, complevel=9)\n",
    "encoding = {var: comp for var in ds.data_vars}\n",
    "ds.to_netcdf(\"scenarios.nc\", encoding=encoding)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52ad4435-aa31-46d3-8e13-d181a9956511",
   "metadata": {},
   "source": [
    "## Balances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b95e25e6-922e-4d22-a0d7-063d69ec6c86",
   "metadata": {},
   "outputs": [],
   "source": [
    "scenarios = MAIN_SCENARIOS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3322b888-7697-4808-a7ce-c3588e712449",
   "metadata": {},
   "outputs": [],
   "source": [
    "e = energy_balances(scenarios)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "85d0e7dc-dab4-4089-aa41-1a08bb9e132c",
   "metadata": {},
   "outputs": [],
   "source": [
    "e"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "65ce1a64-a7bd-4c6f-be01-3cc1bf92943e",
   "metadata": {},
   "outputs": [],
   "source": [
    "h = load_main_supply_energy(\n",
    "    scenarios, carrier=[\"H2\"], rename=rename_techs_h2_balances\n",
    ").xs(100, level=\"onw\", axis=1)\n",
    "\n",
    "order = pd.Index(\n",
    "    [\n",
    "        \"power-to-hydrogen\",\n",
    "        \"steam methane reforming CC\",\n",
    "        \"Fischer-Tropsch\",\n",
    "        \"methanolisation\",\n",
    "        \"hydrogen for industry\",\n",
    "        \"hydrogen for land transport\",\n",
    "        \"hydrogen-to-power/heat\",\n",
    "    ]\n",
    ")\n",
    "\n",
    "order = order.intersection(h.index).append(h.index.difference(order))\n",
    "h = h.loc[order]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c5e4ebb3-7140-4b0b-bcf6-9c3a283d2ad3",
   "metadata": {},
   "outputs": [],
   "source": [
    "h"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9ebf93ad-87fb-4b41-9d6d-c7d5cf9841fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "c = carbon_balances(scenarios)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa92ec9f-6786-47a2-92c6-80692fd18cf7",
   "metadata": {},
   "outputs": [],
   "source": [
    "fn = (\n",
    "    f\"../workflows/pypsa-eur-sec/resources/industrial_energy_demand_elec_s_181_2050.csv\"\n",
    ")\n",
    "industrial_demand = pd.read_csv(fn, index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "763a9726-f8cf-41ea-b890-46244a73a4d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "plastics = (\n",
    "    industrial_demand[\"naphtha\"].sum() * 0.26\n",
    "    - industrial_demand[\"process emission from feedstock\"].sum()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ac3a5cd9-232d-48bf-a8c9-62f292ed5614",
   "metadata": {},
   "outputs": [],
   "source": [
    "fn = f\"../workflows/pypsa-eur-sec/resources/energy_totals.csv\"\n",
    "energy_totals = pd.read_csv(fn, index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a97e0532-72e3-492f-8301-33ffb3b0202b",
   "metadata": {},
   "outputs": [],
   "source": [
    "aviation = (\n",
    "    energy_totals[[\"total international aviation\", \"total domestic aviation\"]]\n",
    "    .sum()\n",
    "    .sum()\n",
    "    * 0.26\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa6edf1a-9170-4658-8c35-b506259e4698",
   "metadata": {},
   "outputs": [],
   "source": [
    "c.drop(\"oil emissions\", inplace=True)\n",
    "c.loc[\"aviation fuels\"] = aviation\n",
    "c.loc[\"high-value chemicals\"] = plastics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6032f2c-9f9f-441f-ab32-66c613b1bf4c",
   "metadata": {},
   "outputs": [],
   "source": [
    "other = c.abs().max(axis=1) < 10\n",
    "to_drop = c.loc[other].index\n",
    "c.loc[\"other\"] = c.loc[other].sum()\n",
    "c.drop(to_drop, axis=0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ca358657-95eb-4847-81f2-59287eb9775f",
   "metadata": {},
   "outputs": [],
   "source": [
    "order = pd.Index(\n",
    "    [\n",
    "        \"aviation fuels\",\n",
    "        \"shipping fuels\",\n",
    "        \"high-value chemicals\",\n",
    "        \"agriculture machinery\",\n",
    "        \"process emissions CC\",\n",
    "        \"gas for industry CC\",\n",
    "        \"gas CHP\",\n",
    "        \"gas boiler\",\n",
    "        \"other\",\n",
    "        \"biogas upgrading\",\n",
    "        \"solid biomass CHP CC\",\n",
    "        \"solid biomass for industry CC\",\n",
    "        \"direct air capture\",\n",
    "    ]\n",
    ")\n",
    "\n",
    "order = order.intersection(c.index).append(c.index.difference(order))\n",
    "c = c.loc[order]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9ec6a73b-6a44-4ca7-b9f8-10886de53aaf",
   "metadata": {},
   "outputs": [],
   "source": [
    "column_dict = {\n",
    "    \"1.0\": \"w/o power\",\n",
    "    \"opt\": \"w power\",\n",
    "    \"H2 grid\": \"w H$_2$\",\n",
    "    \"no H2 grid\": \"w/o H$_2$\",\n",
    "}\n",
    "\n",
    "\n",
    "c.rename(columns=column_dict, inplace=True)\n",
    "e.rename(columns=column_dict, inplace=True)\n",
    "h.rename(columns=column_dict, inplace=True)\n",
    "\n",
    "sep = \"\\n\"\n",
    "c.columns = [sep.join(col).strip() for col in c.columns.values]\n",
    "e.columns = [sep.join(col).strip() for col in e.columns.values]\n",
    "h.columns = [sep.join(col).strip() for col in h.columns.values]\n",
    "\n",
    "c.sort_index(axis=1, inplace=True)\n",
    "e.sort_index(axis=1, inplace=True)\n",
    "h.sort_index(axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5e43b948-745b-4a8d-be49-1c154c1637e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(13, 5.2))\n",
    "\n",
    "lim = 15000\n",
    "e.T.plot.bar(ax=ax1, width=0.8, stacked=True, color=tech_colors, ylim=(-lim, lim))\n",
    "\n",
    "c.T.plot.bar(ax=ax3, width=0.8, stacked=True, cmap=\"tab20\", ylim=(-600, 600))\n",
    "\n",
    "h.T.plot.bar(ax=ax2, width=0.8, stacked=True, color=tech_colors, ylim=(-3000, 3000))\n",
    "\n",
    "ax1.legend(loc=(1.03, -0.5), ncol=1)\n",
    "ax2.legend(loc=(1.03, 0.33))\n",
    "ax3.legend(loc=(1.03, -0.0))\n",
    "\n",
    "for ax in [ax1, ax2, ax3]:\n",
    "    for tick in ax.get_xticklabels():\n",
    "        tick.set_rotation(90)\n",
    "    ax.axhline(0, color=\"k\", linewidth=0.75)\n",
    "    ax.xaxis.set_tick_params(pad=5)\n",
    "    ax.set_xlabel(\"network expansion\", fontsize=10)\n",
    "    ax.grid(axis=\"y\")\n",
    "\n",
    "\n",
    "ax1.set_ylabel(r\"demand $\\leftarrow$ TWh $\\rightarrow$ supply   \")\n",
    "ax3.set_ylabel(r\"withdrawal $\\leftarrow$ Mt$_{CO_2}$ $\\rightarrow$ emission  \")\n",
    "ax2.set_ylabel(r\"demand $\\leftarrow$ TWh $\\rightarrow$ supply   \")\n",
    "\n",
    "ax1.set_title(\"Energy\\nBalance\", fontsize=11)\n",
    "ax3.set_title(r\"Atmosphere CO$_2$\" + \"\\nBalance\", fontsize=11)\n",
    "ax2.set_title(\"Hydrogen\\nBalance\", fontsize=11)\n",
    "\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig(OUTPUT + \"balance.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c55ac95a-b6c4-4795-9299-c5525fd641bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, (ax1, ax2, ax3) = plt.subplots(3, 1, figsize=(13, 8))\n",
    "\n",
    "lim = 14.000\n",
    "e.T.div(1e3).plot.barh(\n",
    "    ax=ax1, width=0.7, stacked=True, color=tech_colors, xlim=(-lim, lim)\n",
    ")\n",
    "\n",
    "c.T.plot.barh(ax=ax3, width=0.7, stacked=True, cmap=\"tab20\", xlim=(-600, 600))\n",
    "\n",
    "h.T.plot.barh(ax=ax2, width=0.7, stacked=True, color=tech_colors, xlim=(-3000, 3000))\n",
    "\n",
    "ax1.legend(loc=(1.03, -0.15), ncol=2)\n",
    "ax2.legend(loc=(1.03, 0), ncol=1)\n",
    "ax3.legend(loc=(1.03, 0), ncol=2)\n",
    "\n",
    "ax1.set_xticks(np.arange(-lim, lim + 1, 2))\n",
    "ax1.set_xticks(np.arange(-lim, lim + 1, 1), minor=True)\n",
    "ax2.set_xticks(np.arange(-3000, 3001, 250), minor=True)\n",
    "ax3.set_xticks(np.arange(-600, 601, 50), minor=True)\n",
    "\n",
    "for ax, thresh, dec in zip([ax1, ax2, ax3], [1, 200, 30], [2, 0, 0]):\n",
    "    # for tick in ax.get_xticklabels():\n",
    "    #    tick.set_rotation(90)\n",
    "    ax.axvline(0, color=\"k\", linewidth=0.75)\n",
    "    ax.xaxis.set_tick_params(pad=5)\n",
    "    ax.set_ylabel(\"network expansion\", fontsize=11)\n",
    "    ax.grid(axis=\"x\")\n",
    "\n",
    "    for bars in ax.containers:\n",
    "        labels = [\n",
    "            np.round(abs(v), dec) if abs(v) > thresh else \"\" for v in bars.datavalues\n",
    "        ]\n",
    "        if not dec:\n",
    "            labels = [f\"{l:.0f}\" if isinstance(l, float) else l for l in labels]\n",
    "        ax.bar_label(\n",
    "            bars, labels=labels, label_type=\"center\", color=\"#333333\", fontsize=10\n",
    "        )\n",
    "\n",
    "ax1.set_xlabel(r\"consumption $\\leftarrow$ thousand TWh $\\rightarrow$ supply           \")\n",
    "ax3.set_xlabel(r\"withdrawal $\\leftarrow$ Mt$_{CO_2}$ $\\rightarrow$ emission  \")\n",
    "ax2.set_xlabel(r\"consumption $\\leftarrow$ TWh $\\rightarrow$ supply           \")\n",
    "\n",
    "ax1.set_title(\"Energy Balance\", fontsize=11)\n",
    "ax3.set_title(r\"Atmosphere CO$_2$ Balance\", fontsize=11)\n",
    "ax2.set_title(\"Hydrogen Balance\", fontsize=11)\n",
    "\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig(OUTPUT + \"balance-h.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b37a5cb-f106-4fad-be95-c4b22ad9caa0",
   "metadata": {},
   "source": [
    "## Curtailment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eb87e484-165d-4cbe-baf2-2faf70752136",
   "metadata": {},
   "outputs": [],
   "source": [
    "COMMON_PATH = \"../results/graphics-20221227/20221227-main\"\n",
    "\n",
    "curtailment = pd.concat(\n",
    "    {\n",
    "        (\"opt\", \"yes\"): pd.read_csv(\n",
    "            f\"{COMMON_PATH}/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/curtailment.csv\",\n",
    "            index_col=0,\n",
    "        ),\n",
    "        (\"opt\", \"no\"): pd.read_csv(\n",
    "            f\"{COMMON_PATH}/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/curtailment.csv\",\n",
    "            index_col=0,\n",
    "        ),\n",
    "        (\"1.0\", \"yes\"): pd.read_csv(\n",
    "            f\"{COMMON_PATH}/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/curtailment.csv\",\n",
    "            index_col=0,\n",
    "        ),\n",
    "        (\"1.0\", \"no\"): pd.read_csv(\n",
    "            f\"{COMMON_PATH}/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/curtailment.csv\",\n",
    "            index_col=0,\n",
    "        ),\n",
    "    },\n",
    "    names=[\"lv\", \"h2\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7604176e-9aa0-4974-9bfa-680794597512",
   "metadata": {},
   "outputs": [],
   "source": [
    "curtailment.loc[\"opt\", \"yes\"].curtailed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "45343753-0dbd-43f1-9655-72660824c2e1",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, axes = plt.subplots(1, 4, figsize=(8, 3), sharey=True)\n",
    "\n",
    "idx = [(\"opt\", \"yes\"), (\"opt\", \"no\"), (\"1.0\", \"yes\"), (\"1.0\", \"no\")]\n",
    "\n",
    "scen = [\n",
    "    \"w power grid expansion\\nw hydrogen network\",\n",
    "    \"w power grid expansion\\nw/o hydrogen network\",\n",
    "    \"w/o power grid expansion\\nw hydrogen network\",\n",
    "    \"w/o power grid expansion\\nw/o hydrogen network\",\n",
    "]\n",
    "\n",
    "\n",
    "for i, ax in enumerate(axes):\n",
    "    df = curtailment.loc[idx[i]]\n",
    "\n",
    "    pd.DataFrame(df.curtailed / df.available.sum() * 100).T.plot.bar(\n",
    "        ax=ax, stacked=True, color=tech_colors, legend=False if i else True\n",
    "    )\n",
    "    ax.set_ylabel(\"Curtailment [%]\")\n",
    "    ax.set_xticklabels([scen[i]], rotation=0)\n",
    "\n",
    "    twax = ax.twinx()\n",
    "    mn, mx = ax.get_ylim()\n",
    "    twax.set_ylim(mn * df.available.sum() / 100, mx * df.available.sum() / 100)\n",
    "    if i == 3:\n",
    "        twax.set_ylabel(\"Curtailment [TWh]\")\n",
    "plt.tight_layout()\n",
    "\n",
    "handles, labels = axes[0].get_legend_handles_labels()\n",
    "\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "\n",
    "nice_names = {\n",
    "    \"offwind-ac\": \"Offshore Wind (AC)\",\n",
    "    \"offwind-dc\": \"Offshore Wind (DC)\",\n",
    "    \"onwind\": \"Onshore Wind\",\n",
    "    \"solar\": \"Solar PV (utility)\",\n",
    "    \"solar rooftop\": \"Solar PV (rooftop)\",\n",
    "    \"ror\": \"Run of River\",\n",
    "}\n",
    "labels = [nice_names[k] for k in labels]\n",
    "\n",
    "axes[0].legend(handles, labels, ncol=3, frameon=False, loc=(0, 1.02))\n",
    "\n",
    "fig.savefig(OUTPUT + \"curtailment.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e38ec1bf-5280-4858-b033-4d26a278a10c",
   "metadata": {},
   "source": [
    "## For CSV Export"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d459c2bc-e7c5-4d15-aa2e-249ca187fb24",
   "metadata": {},
   "outputs": [],
   "source": [
    "def export_csv(scenario):\n",
    "    suffix = scenario.split(\"-\")[-1]\n",
    "\n",
    "    gw, twh = load_main_capacities(rename=False)\n",
    "\n",
    "    to_keep = [\n",
    "        \"H2\",\n",
    "        \"Li ion\",\n",
    "        \"battery\",\n",
    "        \"home battery\",\n",
    "        \"residential rural water tanks\",\n",
    "        \"residential urban decentral water tanks\",\n",
    "        \"services rural water tanks\",\n",
    "        \"services urban decentral water tanks\",\n",
    "        \"urban central water tanks\",\n",
    "    ]\n",
    "    twh.loc[to_keep].xs(100.0, level=\"onw\", axis=1).groupby(\n",
    "        rename_techs_tyndp\n",
    "    ).sum().to_csv(f\"{OUTPUT}/capacities-twh.csv\")\n",
    "\n",
    "    gw.xs(100.0, level=\"onw\", axis=1).groupby(rename_techs_tyndp).sum().to_csv(\n",
    "        f\"{OUTPUT}/capacities-gw.csv\"\n",
    "    )\n",
    "\n",
    "    load_main_energy().xs(100.0, level=\"onw\", axis=1).div(1e6).groupby(\n",
    "        rename_techs_tyndp\n",
    "    ).sum().to_csv(f\"{OUTPUT}/energy.csv\")\n",
    "\n",
    "    load_main().xs(100.0, level=\"onw\", axis=1).to_csv(f\"{OUTPUT}/costs.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aca9f6fc-2fe9-4745-8d05-211046f66be4",
   "metadata": {},
   "outputs": [],
   "source": [
    "for scenario in [\n",
    "    MAIN_SCENARIOS,\n",
    "    GAS_SCENARIOS,\n",
    "    IMP_SCENARIOS,\n",
    "    SHP_SCENARIOS,\n",
    "    COST_SCENARIOS,\n",
    "]:\n",
    "    export_csv(scenario)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67d40c02-ad31-4c81-85dc-05beb1e2c0a1",
   "metadata": {},
   "source": [
    "## Line Volume Sensitivity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eec11dc5-82dd-4ca6-8eaa-6e34fa20c995",
   "metadata": {},
   "outputs": [],
   "source": [
    "costs = pd.read_csv(\n",
    "    LV_SCENARIOS + \"/csvs/costs.csv\", header=[0, 1, 2, 3], index_col=[0, 1, 2]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c11f7929-cd5b-4cf7-9110-53429b5e036b",
   "metadata": {},
   "outputs": [],
   "source": [
    "costs = costs[str(CLUSTERS)].rename(lambda x: float(x), axis=1, level=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "30b4aa10-59fc-4542-b573-0a4cb2893158",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = (\n",
    "    costs.xs(LV_OPTS, level=\"opt\", axis=1)\n",
    "    .xs(\"2050\", level=\"planning_horizon\", axis=1)\n",
    "    .groupby(level=2)\n",
    "    .sum()\n",
    "    .div(1e9)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ebefebc9-4c9d-4e7e-9e59-b764896f5177",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.groupby(df.index.map(rename_techs_tyndp)).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d7ac1350-65ab-47bb-aded-cb885bdc2517",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "77f7ac22-4006-42ba-ac4d-147ab5bc8d04",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sum() - df.sum().max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f75033cf-75e8-46ea-88a8-46b713098e97",
   "metadata": {},
   "outputs": [],
   "source": [
    "benefit_abs = df.sum() - df.sum().max()\n",
    "benefit_abs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "63069215-6ac3-4ad6-8d6b-14004f0d7378",
   "metadata": {},
   "outputs": [],
   "source": [
    "latex[\"lvbenefitabs\"] = -benefit_abs[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "682d2534-ec4e-4916-996d-b1513d58333d",
   "metadata": {},
   "outputs": [],
   "source": [
    "benefit_rel = (1 - df.sum() / df.sum().max()) * 100\n",
    "benefit_rel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f273310e-9922-4312-9873-22823e6e3014",
   "metadata": {},
   "outputs": [],
   "source": [
    "latex[\"lvbenefitabs\"] = -benefit_rel[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "da95b708-443b-4790-8785-993a8228fcb2",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_drop = df.index[df.max(axis=1).fillna(0.0) < 1.2]\n",
    "to_drop = to_drop.difference(\n",
    "    [\"DAC\", \"hot water storage\"]\n",
    ")  # exclude dropping to align with onwind sensitivity...\n",
    "df.drop(to_drop, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0a5c7698-f65c-482a-8710-0242bf748b38",
   "metadata": {},
   "outputs": [],
   "source": [
    "order = preferred_order.intersection(df.index).append(\n",
    "    df.index.difference(preferred_order)\n",
    ")\n",
    "df = df.loc[order]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "629f0e80-ed30-4050-8656-c90d4cfada36",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.columns *= 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "56ecdd3c-6d71-4c17-b4da-7855e8f661a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "dec = load_decentral()[\"H2 grid\"]\n",
    "dec.name = \"\\nno power\\ntransmission\"\n",
    "dec = pd.DataFrame(dec).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9651f702-4c70-4c16-883b-85c3c3dabc16",
   "metadata": {},
   "outputs": [],
   "source": [
    "dec.sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1db9fea4-9268-4826-8549-9ea25bebd172",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, (ax0, ax1) = plt.subplots(\n",
    "    1,\n",
    "    2,\n",
    "    figsize=(5, 3),\n",
    "    sharey=True,\n",
    "    gridspec_kw={\"width_ratios\": [1, 6], \"wspace\": 0.05},\n",
    ")\n",
    "\n",
    "to_plot = df.T.sort_index()\n",
    "\n",
    "tech_colors = config[\"plotting\"][\"tech_colors\"]\n",
    "colors = [tech_colors[i] for i in df.index]\n",
    "\n",
    "to_plot.plot.area(ax=ax1, stacked=True, linewidth=0, color=colors)\n",
    "\n",
    "dec.plot.bar(ax=ax0, stacked=True, linewidth=0, legend=False, color=tech_colors)\n",
    "\n",
    "handles, labels = ax1.get_legend_handles_labels()\n",
    "\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "\n",
    "ax1.set_xlim(100, 200)\n",
    "ax1.set_xlabel(\"Power Grid Reinforcement Restriction\\n[% relative to today's volume]\")\n",
    "\n",
    "ax0.set_ylim([0, 1000])\n",
    "ax0.set_ylabel(\"System Cost\\n[EUR billion per year]\")\n",
    "ax0.grid(axis=\"y\")\n",
    "ax1.grid(axis=\"y\")\n",
    "\n",
    "ax1.axvline(125, color=\"k\", linestyle=\"--\", linewidth=1)\n",
    "\n",
    "ax1.text(127.5, 810, \"TYNDP equivalent\", size=11, color=\"k\")\n",
    "\n",
    "# legend on side\n",
    "ax1.legend(handles, labels, ncol=2, frameon=False, bbox_to_anchor=(1.02, 1.03))\n",
    "\n",
    "for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "    ax1.spines[i].set_visible(False)\n",
    "    ax0.spines[i].set_visible(False)\n",
    "\n",
    "# legend inside\n",
    "# ax1.legend(handles, labels, ncol=3, frameon=False, bbox_to_anchor=(1.1, 1.48))\n",
    "\n",
    "ax1.set_axisbelow(False)\n",
    "\n",
    "ax0.tick_params(rotation=0, labelsize=10)\n",
    "ax1.tick_params(rotation=0, labelsize=10)\n",
    "ax1.set_xticks([100, 125, 150, 175, 200])\n",
    "\n",
    "fig.savefig(OUTPUT + \"lv-sensitivity.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "63ffab26-ad74-4775-8ffb-d81493b50ecb",
   "metadata": {},
   "source": [
    "## Onshore Wind Sensitivity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c6c71dd4-3c5f-4a64-b355-3b4658d79ced",
   "metadata": {},
   "outputs": [],
   "source": [
    "costs = pd.read_csv(\n",
    "    ONW_SCENARIOS + \"/csvs/costs.csv\", header=[0, 1, 2, 3], index_col=[0, 1, 2]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "27e7e0a7-c1c3-4397-9513-2c0fe3081bcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "costs = costs.xs(\n",
    "    (str(CLUSTERS), \"1.25\", \"2050\"), level=[\"cluster\", \"lv\", \"planning_horizon\"], axis=1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2fd7b8ee-8d6c-4408-aa2b-70e0d5567414",
   "metadata": {},
   "outputs": [],
   "source": [
    "costs.columns = [\n",
    "    100 * float(re.search(r\"onwind\\+p([0-9.]*)\", c).groups()[0]) for c in costs.columns\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d0009c57-bf47-45ef-a5f5-9700ddfacadc",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = costs.groupby(level=2).sum().div(1e9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a4d48b62-a968-473f-b361-5ccf483de53e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.groupby(df.index.map(rename_techs_tyndp)).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dd39c1ce-5036-45b5-9c59-7aaebb86bf1e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d5132ecb-b7c9-4e2b-9f80-02c5e48f7a76",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sum() - df.sum().max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "76c00169-72cb-41ee-89d5-ad78a7461299",
   "metadata": {},
   "outputs": [],
   "source": [
    "(1 - df.sum() / df.sum().min()) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d182a105-f710-4d60-ae66-2f4b3e733abb",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_drop = df.index[df.max(axis=1).fillna(0.0) < 1.2]\n",
    "print(to_drop)\n",
    "df.drop(to_drop, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4c031a0-233a-4d47-806f-e0b7196ac6eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "order = preferred_order.intersection(df.index).append(\n",
    "    df.index.difference(preferred_order)\n",
    ")\n",
    "df = df.loc[order]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6d5d1d8c-056a-4ccb-8558-05daf0c2fd8b",
   "metadata": {},
   "outputs": [],
   "source": [
    "spec = load_main().T\n",
    "\n",
    "spec = spec.xs((\"1.0\", 0.0, \"H2 grid\"), level=[\"lv\", \"onw\", \"h2\"])\n",
    "\n",
    "spec.index = [\"no grid\\nexpansion\\nno onshore wind\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e21208cf-a338-4e14-a437-78d3eb933ab9",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, (ax0, ax1) = plt.subplots(\n",
    "    1,\n",
    "    2,\n",
    "    figsize=(5, 3),\n",
    "    sharey=True,\n",
    "    gridspec_kw={\"width_ratios\": [1, 6], \"wspace\": 0.05},\n",
    ")\n",
    "\n",
    "to_plot = df.T.sort_index()\n",
    "\n",
    "tech_colors = config[\"plotting\"][\"tech_colors\"]\n",
    "colors = [tech_colors[i] for i in df.index]\n",
    "\n",
    "to_plot.plot.area(ax=ax1, stacked=True, linewidth=0, color=colors)\n",
    "\n",
    "spec.plot.bar(ax=ax0, stacked=True, linewidth=0, legend=False, color=tech_colors)\n",
    "\n",
    "handles, labels = ax1.get_legend_handles_labels()\n",
    "\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "\n",
    "ax1.set_xlim(0, 100)\n",
    "ax1.set_xlabel(\"Fraction of technical onshore\\nwind potential available [%]\")\n",
    "\n",
    "ax0.set_ylim(0, 1000)\n",
    "ax0.set_ylabel(\"System Cost\\n[EUR billion per year]\")\n",
    "ax0.grid(axis=\"y\")\n",
    "ax1.grid(axis=\"y\")\n",
    "\n",
    "ax1.axvline(25, color=\"k\", linestyle=\"--\", linewidth=1)\n",
    "\n",
    "ax1.text(27.5, 900, \"compromise social potential\", size=11, color=\"k\")\n",
    "\n",
    "# legend on side\n",
    "ax1.legend(handles, labels, ncol=2, frameon=False, bbox_to_anchor=(1, 1.02))\n",
    "\n",
    "for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "    ax1.spines[i].set_visible(False)\n",
    "    ax0.spines[i].set_visible(False)\n",
    "\n",
    "# legend inside\n",
    "# ax.legend(handles, labels, ncol=3, frameon=False, bbox_to_anchor=(1.1, 1.48))\n",
    "\n",
    "ax1.set_axisbelow(False)\n",
    "\n",
    "ax0.tick_params(rotation=0, labelsize=10)\n",
    "ax1.tick_params(rotation=0, labelsize=10)\n",
    "ax1.set_xticks([0, 25, 50, 75, 100])\n",
    "\n",
    "fig.savefig(OUTPUT + \"onw-sensitivity.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "803b7d68-af2c-4bdb-9b74-fcfcf3b81207",
   "metadata": {},
   "source": [
    "\n",
    "## Inter-scenario diff charts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa7ee6b5-112a-40f7-be4f-fe02122c78c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_diff(\n",
    "    cost,\n",
    "    main,\n",
    "    df=None,\n",
    "    legend=False,\n",
    "    suffix=None,\n",
    "    label=\"cost\",\n",
    "    unit=\"bn€/a\",\n",
    "    scaler=50,\n",
    "    threshold=False,\n",
    "    label_scaler=15,\n",
    "    rename=True,\n",
    "    sign=1,\n",
    "):\n",
    "    union = cost.index.union(main.index)\n",
    "    cost = cost.reindex(index=union, fill_value=0.0)\n",
    "    main = main.reindex(index=union, fill_value=0.0)\n",
    "\n",
    "    if df is None:\n",
    "        df = cost - main\n",
    "\n",
    "    if rename:\n",
    "        column_dict = {\n",
    "            \"1.0\": \"w/o power grid expansion\",\n",
    "            \"opt\": \"w power grid expansion\",\n",
    "            \"H2 grid\": \"w hydrogen network\",\n",
    "            \"no H2 grid\": \"w/o hydrogen network\",\n",
    "        }\n",
    "\n",
    "        fix_df(df)\n",
    "        fix_df(cost)\n",
    "        fix_df(main)\n",
    "\n",
    "    fig, ax = plt.subplots(1, 1, figsize=(6, 2.7))\n",
    "\n",
    "    if threshold:\n",
    "        to_drop = df.index[df.abs().max(axis=1).fillna(0.0) < threshold]\n",
    "        df.drop(to_drop, inplace=True)\n",
    "\n",
    "    df.round(1).T.plot.barh(\n",
    "        ax=ax,\n",
    "        stacked=True,\n",
    "        color=tech_colors,\n",
    "        xlabel=f\"reduced\"\n",
    "        + r\" $\\leftarrow$ \"\n",
    "        + f\"{label} [{unit}]\"\n",
    "        + r\" $\\rightarrow$ \"\n",
    "        + f\"increased\",\n",
    "        legend=legend,\n",
    "    )\n",
    "    # ax.axvline(0, color='#454545', linewidth=1.5, zorder=-1)\n",
    "\n",
    "    net = df.sum()\n",
    "    rel = sign * (cost.sum() / main.sum() * 100 - 100)\n",
    "    netcolor = \"darkslategrey\"\n",
    "    ax.scatter(net.values, net.index, s=10, c=netcolor, alpha=0.6, edgecolor=\"none\")\n",
    "\n",
    "    for x, y in net.items():\n",
    "        ax.annotate(\n",
    "            f\"{y:.0f} | {rel[x]:.1f}%\",\n",
    "            (y, x),\n",
    "            xytext=(-12, 13),\n",
    "            textcoords=\"offset points\",\n",
    "            color=netcolor,\n",
    "            fontsize=9,\n",
    "            bbox=dict(fc=\"0.85\", boxstyle=\"square,pad=0.1\"),\n",
    "        )\n",
    "\n",
    "    plt.grid(axis=\"x\")\n",
    "\n",
    "    ylim = max(-df.where(df < 0).sum().min(), df.where(df > 0).sum().max())\n",
    "    ylim = np.ceil(ylim / scaler) * scaler\n",
    "\n",
    "    plt.xlim([-ylim, ylim])\n",
    "    plt.ylim([-0.5, 3.8])\n",
    "\n",
    "    for bars in ax.containers:\n",
    "        labels = [\n",
    "            f\"{abs(v):.0f}\" if abs(v) > ylim / label_scaler else \"\"\n",
    "            for v in bars.datavalues\n",
    "        ]\n",
    "        ax.bar_label(\n",
    "            bars, labels=labels, label_type=\"center\", color=\"#444444\", fontsize=9\n",
    "        )\n",
    "\n",
    "    handles = [\n",
    "        plt.Line2D(\n",
    "            [],\n",
    "            [],\n",
    "            color=mpl.colors.to_rgba(netcolor, 0.5),\n",
    "            marker=\".\",\n",
    "            markeredgecolor=\"none\",\n",
    "            linestyle=\"None\",\n",
    "            markersize=10,\n",
    "        )\n",
    "    ]\n",
    "    legend = ax.legend(\n",
    "        handles,\n",
    "        [\"net difference\"],\n",
    "        labelcolor=netcolor,\n",
    "        fontsize=11,\n",
    "        frameon=True,\n",
    "        bbox_to_anchor=(-0.015, -0.055),\n",
    "    )\n",
    "    ax.add_artist(legend)\n",
    "\n",
    "    plt.tight_layout()\n",
    "\n",
    "    if legend:\n",
    "        plt.legend(ncol=1, loc=(1.05, -0.25), labelspacing=0.2)\n",
    "\n",
    "    plt.savefig(\n",
    "        OUTPUT + f\"diff-{label}{suffix}.pdf\", bbox_inches=Bbox([[0, 0], [8.5, 2.8]])\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "02331a81-5f34-4d56-abbf-21a363fbf737",
   "metadata": {},
   "outputs": [],
   "source": [
    "for SCENARIO in [IMP_SCENARIOS, SHP_SCENARIOS, COST_SCENARIOS]:\n",
    "    suffix = \"-\" + SCENARIO.split(\"-\")[-1] if SCENARIO != MAIN_SCENARIOS else \"\"\n",
    "\n",
    "    sign = -1 if \"rev0\" in SCENARIO else 1\n",
    "\n",
    "    print(suffix)\n",
    "\n",
    "    main = load_main(MAIN_SCENARIOS).xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    cost = load_main(SCENARIO).xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    plot_diff(sign * cost, sign * main, suffix=suffix, threshold=5, sign=sign)\n",
    "\n",
    "    a = load_main_supply_energy(SCENARIO).xs(100.0, level=\"onw\", axis=1)\n",
    "    b = load_main_supply_energy(MAIN_SCENARIOS).xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    a_pos = a.where(a > 0)\n",
    "    b_pos = b.where(b > 0)\n",
    "\n",
    "    plot_diff(\n",
    "        sign * a_pos,\n",
    "        sign * b_pos,\n",
    "        suffix=suffix,\n",
    "        label=\"supply\",\n",
    "        unit=r\"TWh\",\n",
    "        scaler=200,\n",
    "        threshold=1,\n",
    "        label_scaler=14,\n",
    "        sign=sign,\n",
    "    )\n",
    "\n",
    "    a_neg = -a.where(a < 0)\n",
    "    b_neg = -b.where(b < 0)\n",
    "\n",
    "    plot_diff(\n",
    "        sign * a_neg,\n",
    "        sign * b_neg,\n",
    "        suffix=suffix,\n",
    "        label=\"consumption\",\n",
    "        unit=r\"TWh\",\n",
    "        scaler=500,\n",
    "        legend=True,\n",
    "        threshold=1,\n",
    "        label_scaler=8,\n",
    "        sign=sign,\n",
    "    )\n",
    "\n",
    "    a = load_main_supply_energy(SCENARIO, carrier=[\"co2\"], rename=False).xs(\n",
    "        100.0, level=\"onw\", axis=1\n",
    "    )\n",
    "    b = load_main_supply_energy(MAIN_SCENARIOS, carrier=[\"co2\"], rename=False).xs(\n",
    "        100.0, level=\"onw\", axis=1\n",
    "    )\n",
    "\n",
    "    a = a.groupby(rename_techs_carbon_balances).sum()\n",
    "    b = b.groupby(rename_techs_carbon_balances).sum()\n",
    "\n",
    "    a_pos = a.where(a > 0)\n",
    "    b_pos = b.where(b > 0)\n",
    "\n",
    "    plot_diff(\n",
    "        sign * a_pos,\n",
    "        sign * b_pos,\n",
    "        suffix=suffix,\n",
    "        label=\"emission\",\n",
    "        unit=r\"Mt$_{CO_2}$\",\n",
    "        scaler=200,\n",
    "        threshold=0.05,\n",
    "        label_scaler=14,\n",
    "        sign=sign,\n",
    "    )\n",
    "\n",
    "    a_neg = -a.where(a < 0)\n",
    "    b_neg = -b.where(b < 0)\n",
    "\n",
    "    plot_diff(\n",
    "        sign * a_neg,\n",
    "        sign * b_neg,\n",
    "        suffix=suffix,\n",
    "        label=\"withdrawal\",\n",
    "        unit=r\"Mt$_{CO_2}$\",\n",
    "        scaler=400,\n",
    "        legend=True,\n",
    "        threshold=0.05,\n",
    "        label_scaler=8,\n",
    "        sign=sign,\n",
    "    )\n",
    "\n",
    "    gw_a, twh_a = load_main_capacities(SCENARIO)\n",
    "    gw_b, twh_b = load_main_capacities(MAIN_SCENARIOS)\n",
    "\n",
    "    gw_a = gw_a.xs(100.0, level=\"onw\", axis=1)\n",
    "    gw_b = gw_b.xs(100.0, level=\"onw\", axis=1)\n",
    "    twh_a = twh_a.xs(100.0, level=\"onw\", axis=1)\n",
    "    twh_b = twh_b.xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    plot_diff(\n",
    "        sign * gw_a,\n",
    "        sign * gw_b,\n",
    "        suffix=suffix,\n",
    "        label=\"capacity\",\n",
    "        unit=\"GW\",\n",
    "        scaler=250,\n",
    "        threshold=20,\n",
    "        label_scaler=8,\n",
    "        sign=sign,\n",
    "    )\n",
    "\n",
    "    plot_diff(\n",
    "        sign * twh_a,\n",
    "        sign * twh_b,\n",
    "        suffix=suffix,\n",
    "        label=\"storage\",\n",
    "        unit=\"TWh\",\n",
    "        scaler=40,\n",
    "        threshold=1,\n",
    "        sign=sign,\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5c4fad60-20a3-464e-ade3-25ae9bee42ed",
   "metadata": {},
   "source": [
    "## Time diff charts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dc67b821-7fcc-4f99-b0de-a80d2b96d866",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_time_diff(\n",
    "    df,\n",
    "    reference,\n",
    "    legend=False,\n",
    "    label=\"cost\",\n",
    "    unit=\"bn€/a\",\n",
    "    scaler=50,\n",
    "    threshold=False,\n",
    "    label_scaler=15,\n",
    "    rename=True,\n",
    "    ylim=None,\n",
    "):\n",
    "    ref = df[reference]\n",
    "\n",
    "    df = (df.T - df[reference]).drop(reference).T\n",
    "\n",
    "    df.columns = [f\"{c}-hourly\\n{8760/c:.0f} snapshots\" for c in df.columns]\n",
    "\n",
    "    fig, ax = plt.subplots(1, 1, figsize=(6, 3.4))\n",
    "\n",
    "    if threshold:\n",
    "        to_drop = df.index[df.abs().max(axis=1).fillna(0.0) < threshold]\n",
    "        df.drop(to_drop, inplace=True)\n",
    "\n",
    "    ax.axhline(1, zorder=-1, alpha=0.1, linewidth=35, color=\"darkslategray\")\n",
    "    df.round(1).T.plot.barh(\n",
    "        ax=ax,\n",
    "        stacked=True,\n",
    "        color=tech_colors,\n",
    "        xlabel=f\"reduced\"\n",
    "        + r\" $\\leftarrow$ \"\n",
    "        + f\"{label} [{unit}]\"\n",
    "        + r\" $\\rightarrow$ \"\n",
    "        + f\"increased\",\n",
    "        legend=legend,\n",
    "    )\n",
    "\n",
    "    net = df.sum()\n",
    "    rel = df.sum() / ref.sum() * 100\n",
    "\n",
    "    netcolor = \"darkslategrey\"\n",
    "    ax.scatter(net.values, net.index, s=10, c=netcolor, alpha=0.6, edgecolor=\"none\")\n",
    "\n",
    "    for x, y in net.items():\n",
    "        ax.annotate(\n",
    "            f\"{y:.1f} | {rel[x]:.2f}%\",\n",
    "            (y, x),\n",
    "            xytext=(-12, 13),\n",
    "            textcoords=\"offset points\",\n",
    "            color=netcolor,\n",
    "            fontsize=9,\n",
    "            bbox=dict(fc=\"0.85\", boxstyle=\"square,pad=0.1\"),\n",
    "        )\n",
    "\n",
    "    plt.grid(axis=\"x\")\n",
    "\n",
    "    if ylim is None:\n",
    "        ylim = max(-df.where(df < 0).sum().min(), df.where(df > 0).sum().max())\n",
    "        ylim = np.ceil(ylim / scaler) * scaler\n",
    "\n",
    "    plt.xlim([-ylim, ylim])\n",
    "    plt.ylim([-0.5, 4.8])\n",
    "    plt.ylabel(\"\")\n",
    "\n",
    "    for bars in ax.containers:\n",
    "        labels = [\n",
    "            f\"{abs(v):.0f}\" if abs(v) > ylim / label_scaler else \"\"\n",
    "            for v in bars.datavalues\n",
    "        ]\n",
    "        ax.bar_label(\n",
    "            bars, labels=labels, label_type=\"center\", color=\"#444444\", fontsize=9\n",
    "        )\n",
    "\n",
    "    handles = [\n",
    "        plt.Line2D(\n",
    "            [],\n",
    "            [],\n",
    "            color=mpl.colors.to_rgba(netcolor, 0.5),\n",
    "            marker=\".\",\n",
    "            markeredgecolor=\"none\",\n",
    "            linestyle=\"None\",\n",
    "            markersize=10,\n",
    "        )\n",
    "    ]\n",
    "    legend = ax.legend(\n",
    "        handles,\n",
    "        [\"net difference\"],\n",
    "        labelcolor=netcolor,\n",
    "        fontsize=10,\n",
    "        frameon=False,\n",
    "        loc=(1.05, -0.1),\n",
    "    )\n",
    "    ax.add_artist(legend)\n",
    "\n",
    "    plt.tight_layout()\n",
    "\n",
    "    if legend:\n",
    "        plt.legend(ncol=1, loc=(1.05, 0), labelspacing=0.2)\n",
    "\n",
    "    plt.savefig(\n",
    "        OUTPUT + f\"diff-time-{grid.replace('.', 'p')}-{label}.pdf\",\n",
    "        bbox_inches=Bbox([[0, 0], [8.5, 3.4]]),\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "70efa894-7e7d-4da2-92a0-7bf122c4a2fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "grid = \"opt\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4c6d6836-75da-475a-b793-ca914925615f",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = (\n",
    "    load_main(TIME_SCENARIOS, clusters=90, with_resolution=True)\n",
    "    .xs(grid, level=\"lv\", axis=1)\n",
    "    .droplevel([\"onw\", \"h2\"], axis=1)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d0d2bb56-3e86-4c81-9ceb-9254f8f1c4c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_time_diff(df, 1, rename=False, scaler=20, label_scaler=50, threshold=0.1, ylim=60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c6ae3807-ed36-4d10-91a3-a82f2184a731",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = (\n",
    "    load_main_supply_energy(TIME_SCENARIOS, clusters=90, with_resolution=True)\n",
    "    .xs(grid, level=\"lv\", axis=1)\n",
    "    .droplevel([\"onw\", \"h2\"], axis=1)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8a7c94cf-1e71-49f2-aab0-8353acd2259b",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_time_diff(\n",
    "    df.where(df > 0),\n",
    "    1,\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"supply\",\n",
    "    unit=r\"TWh\",\n",
    "    ylim=700,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "56878f6a-e9c4-4a5b-ae12-26dbf10ec247",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_time_diff(\n",
    "    -df.where(df < 0),\n",
    "    1,\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"consumption\",\n",
    "    unit=r\"TWh\",\n",
    "    ylim=120,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cd9e1ce8-a480-4238-a4a9-058b18331951",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = (\n",
    "    load_main_supply_energy(\n",
    "        TIME_SCENARIOS, carrier=[\"co2\"], rename=False, clusters=90, with_resolution=True\n",
    "    )\n",
    "    .xs(grid, level=\"lv\", axis=1)\n",
    "    .droplevel([\"onw\", \"h2\"], axis=1)\n",
    ")"
   ]
  },
  {
   "cell_type": "raw",
   "id": "63092c2a-c81f-4480-a038-d6710a7ca613",
   "metadata": {},
   "source": [
    "df = df.groupby(rename_techs_carbon_balances).sum()"
   ]
  },
  {
   "cell_type": "raw",
   "id": "77306617-4cb1-4c5b-96d4-a54c4caf4104",
   "metadata": {},
   "source": [
    "plot_time_diff(\n",
    "    df.where(df > 0),\n",
    "    1,\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0,\n",
    "    label=\"emission\",\n",
    "    unit=r\"Mt$_{CO_2}$\",\n",
    "    ylim=700,\n",
    ")"
   ]
  },
  {
   "cell_type": "raw",
   "id": "6270a033-9476-407b-a86c-f77eaa00db0a",
   "metadata": {},
   "source": [
    "plot_time_diff(\n",
    "    -df.where(df < 0),\n",
    "    1,\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0,\n",
    "    label=\"withdrawal\",\n",
    "    unit=r\"TWh\",\n",
    "    ylim=120,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f961deb8-da73-4fca-9a04-6f57934fd993",
   "metadata": {},
   "outputs": [],
   "source": [
    "gw, twh = load_main_capacities(TIME_SCENARIOS, clusters=90, with_resolution=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d62333c8-684f-499f-90c9-a286b896cbb0",
   "metadata": {},
   "outputs": [],
   "source": [
    "gw = gw.xs(grid, level=\"lv\", axis=1).droplevel([\"onw\", \"h2\"], axis=1)\n",
    "twh = twh.xs(grid, level=\"lv\", axis=1).droplevel([\"onw\", \"h2\"], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8f5e0c37-bd9b-445f-9259-bf82bb29dbce",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_time_diff(\n",
    "    gw,\n",
    "    1,\n",
    "    scaler=50,\n",
    "    label_scaler=15,\n",
    "    threshold=0.1,\n",
    "    label=\"capacity\",\n",
    "    unit=r\"GW\",\n",
    "    ylim=600,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0590fdd0-effd-44ee-b7b9-0ef1e21590d3",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_time_diff(\n",
    "    twh,\n",
    "    1,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"storage\",\n",
    "    unit=r\"TWh\",\n",
    "    ylim=40,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b8e81a6-eb19-4aee-a4c2-4614fa6e6ae5",
   "metadata": {},
   "source": [
    "## Spatial diff charts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "650a3511-62da-42b9-868f-717d7c71beb6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_space_diff(\n",
    "    df,\n",
    "    reference,\n",
    "    legend=False,\n",
    "    label=\"cost\",\n",
    "    unit=\"bn€/a\",\n",
    "    scaler=50,\n",
    "    threshold=False,\n",
    "    label_scaler=15,\n",
    "    rename=True,\n",
    "    ylim=None,\n",
    "):\n",
    "    ref = df[reference]\n",
    "\n",
    "    df = (df.T - df[reference]).drop(reference).T\n",
    "\n",
    "    df.columns = [f\"{c} regions\" for c in df.columns]\n",
    "\n",
    "    fig, ax = plt.subplots(1, 1, figsize=(6, 2.8))\n",
    "\n",
    "    if threshold:\n",
    "        to_drop = df.index[df.abs().max(axis=1).fillna(0.0) < threshold]\n",
    "        df.drop(to_drop, inplace=True)\n",
    "\n",
    "    # ax.axhline(1, zorder=-1, alpha=0.1, linewidth=35, color='darkslategray')\n",
    "    df.round(1).T.plot.barh(\n",
    "        ax=ax,\n",
    "        stacked=True,\n",
    "        color=tech_colors,\n",
    "        xlabel=f\"reduced\"\n",
    "        + r\" $\\leftarrow$ \"\n",
    "        + f\"{label} [{unit}]\"\n",
    "        + r\" $\\rightarrow$ \"\n",
    "        + f\"increased\",\n",
    "        legend=legend,\n",
    "    )\n",
    "\n",
    "    net = df.sum()\n",
    "    rel = df.sum() / ref.sum() * 100\n",
    "\n",
    "    netcolor = \"darkslategrey\"\n",
    "    ax.scatter(net.values, net.index, s=10, c=netcolor, alpha=0.6, edgecolor=\"none\")\n",
    "\n",
    "    for x, y in net.items():\n",
    "        ax.annotate(\n",
    "            f\"{y:.1f} | {rel[x]:.2f}%\",\n",
    "            (y, x),\n",
    "            xytext=(-12, 13),\n",
    "            textcoords=\"offset points\",\n",
    "            color=netcolor,\n",
    "            fontsize=9,\n",
    "            bbox=dict(fc=\"0.85\", boxstyle=\"square,pad=0.1\"),\n",
    "        )\n",
    "\n",
    "    plt.grid(axis=\"x\")\n",
    "\n",
    "    if ylim is None:\n",
    "        ylim = max(-df.where(df < 0).sum().min(), df.where(df > 0).sum().max())\n",
    "        ylim = np.ceil(ylim / scaler) * scaler\n",
    "\n",
    "    plt.xlim([-ylim, ylim])\n",
    "    plt.ylim([-0.5, 3.8])\n",
    "    plt.ylabel(\"\")\n",
    "\n",
    "    for bars in ax.containers:\n",
    "        labels = [\n",
    "            f\"{abs(v):.0f}\" if abs(v) > ylim / label_scaler else \"\"\n",
    "            for v in bars.datavalues\n",
    "        ]\n",
    "        ax.bar_label(\n",
    "            bars, labels=labels, label_type=\"center\", color=\"#444444\", fontsize=9\n",
    "        )\n",
    "\n",
    "    handles = [\n",
    "        plt.Line2D(\n",
    "            [],\n",
    "            [],\n",
    "            color=mpl.colors.to_rgba(netcolor, 0.5),\n",
    "            marker=\".\",\n",
    "            markeredgecolor=\"none\",\n",
    "            linestyle=\"None\",\n",
    "            markersize=10,\n",
    "        )\n",
    "    ]\n",
    "    legend = ax.legend(\n",
    "        handles,\n",
    "        [\"net difference\"],\n",
    "        labelcolor=netcolor,\n",
    "        fontsize=10,\n",
    "        frameon=False,\n",
    "        loc=(1.05, -0.25),\n",
    "    )\n",
    "    ax.add_artist(legend)\n",
    "\n",
    "    plt.tight_layout()\n",
    "\n",
    "    if legend:\n",
    "        plt.legend(ncol=1, loc=(1.05, -0.15), labelspacing=0.2)\n",
    "\n",
    "    plt.savefig(\n",
    "        OUTPUT\n",
    "        + f\"diff-space-{grid.replace('.', 'p')}-{h2.replace(' ', '_')}-{label}.pdf\",\n",
    "        bbox_inches=Bbox([[0, 0], [8.5, 2.8]]),\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4d3707c2-44d9-4355-be2d-c744c8caadb6",
   "metadata": {},
   "outputs": [],
   "source": [
    "grid = \"opt\"\n",
    "h2 = \"H2 grid\"\n",
    "df = (\n",
    "    load_main(SPACE_SCENARIOS, with_space=True)\n",
    "    .xs((grid, h2), level=[\"lv\", \"h2\"], axis=1)\n",
    "    .droplevel(\"onw\", axis=1)\n",
    ")\n",
    "s = load_main(MAIN_SCENARIOS)[(grid, 100, h2)]\n",
    "s.name = \"181\"\n",
    "df = pd.concat([df, s], axis=1).fillna(0.0).loc[:, ::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e4e9ebc3-b4df-48e9-a429-8fc2fdbc0408",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_space_diff(\n",
    "    df, \"181\", rename=False, scaler=20, label_scaler=30, threshold=0.1, ylim=60\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b6aead24-174b-4a6c-8c63-3486376556e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = (\n",
    "    load_main_supply_energy(SPACE_SCENARIOS, with_space=True)\n",
    "    .xs((grid, h2), level=[\"lv\", \"h2\"], axis=1)\n",
    "    .droplevel(\"onw\", axis=1)\n",
    ")\n",
    "s = load_main_supply_energy(MAIN_SCENARIOS)[(grid, 100, h2)]\n",
    "s.name = \"181\"\n",
    "df = pd.concat([df, s], axis=1).fillna(0.0).loc[:, ::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c588f475-5d48-4abf-8679-45cc069b4f17",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_space_diff(\n",
    "    df.where(df > 0),\n",
    "    \"181\",\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"supply\",\n",
    "    unit=r\"TWh\",\n",
    "    ylim=700,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2981cb7b-452b-404e-a699-781758aefddb",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_space_diff(\n",
    "    -df.where(df < 0),\n",
    "    \"181\",\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"consumption\",\n",
    "    unit=r\"TWh\",\n",
    "    ylim=250,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "71244941-a4cd-42a3-a0ba-db483f132723",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = (\n",
    "    load_main_supply_energy(\n",
    "        SPACE_SCENARIOS, carrier=[\"co2\"], rename=False, with_space=True\n",
    "    )\n",
    "    .xs((grid, h2), level=[\"lv\", \"h2\"], axis=1)\n",
    "    .droplevel(\"onw\", axis=1)\n",
    ")\n",
    "s = load_main_supply_energy(MAIN_SCENARIOS, carrier=[\"co2\"], rename=False)[\n",
    "    (grid, 100, h2)\n",
    "]\n",
    "s.name = \"181\"\n",
    "df = pd.concat([df, s], axis=1).fillna(0.0).loc[:, ::-1]\n",
    "\n",
    "df = df.groupby(rename_techs_carbon_balances).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f23a5471-b768-4760-bf74-73264224330c",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_space_diff(\n",
    "    df.where(df > 0),\n",
    "    \"181\",\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"emission\",\n",
    "    unit=r\"Mt$_{CO_2}$\",\n",
    "    ylim=100,\n",
    ")"
   ]
  },
  {
   "cell_type": "raw",
   "id": "d25eb41e-0523-4718-8737-072d2f5fbf9f",
   "metadata": {},
   "source": [
    "plot_space_diff(\n",
    "    -df.where(df < 0),\n",
    "    \"181\",\n",
    "    rename=False,\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"withdrawal\",\n",
    "    unit=r\"Mt$_{CO_2}$\",\n",
    "    ylim=100,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fb78a3e7-131d-4a4e-8d8e-55e7f502d89b",
   "metadata": {},
   "outputs": [],
   "source": [
    "gw, twh = load_main_capacities(SPACE_SCENARIOS, with_space=True)\n",
    "gw_, twh_ = load_main_capacities(MAIN_SCENARIOS)\n",
    "\n",
    "gw = gw.xs((grid, h2), level=[\"lv\", \"h2\"], axis=1).droplevel(\"onw\", axis=1)\n",
    "gw_ = gw_[(grid, 100, h2)]\n",
    "\n",
    "twh = twh.xs((grid, h2), level=[\"lv\", \"h2\"], axis=1).droplevel(\"onw\", axis=1)\n",
    "twh_ = twh_[(grid, 100, h2)]\n",
    "\n",
    "gw_.name = \"181\"\n",
    "twh_.name = \"181\"\n",
    "\n",
    "gw = pd.concat([gw, gw_], axis=1).fillna(0.0).loc[:, ::-1]\n",
    "twh = pd.concat([twh, twh_], axis=1).fillna(0.0).loc[:, ::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8cea9914-b5ac-4fa4-bbcd-07dbd6f204c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_space_diff(\n",
    "    gw,\n",
    "    \"181\",\n",
    "    scaler=50,\n",
    "    label_scaler=15,\n",
    "    threshold=0.1,\n",
    "    label=\"capacity\",\n",
    "    unit=r\"GW\",\n",
    "    ylim=600,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "96ed96ab-6447-4ab0-bbad-07a0a5fae864",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_space_diff(\n",
    "    twh,\n",
    "    \"181\",\n",
    "    scaler=50,\n",
    "    label_scaler=25,\n",
    "    threshold=0.1,\n",
    "    label=\"storage\",\n",
    "    unit=r\"TWh\",\n",
    "    ylim=40,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49403b26-8233-4902-abe9-424a61a027c8",
   "metadata": {},
   "source": [
    "## Intra-scenario diff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6185cc02-a8a9-4302-abac-4fb73ef7d8a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_internal_diff(\n",
    "    df,\n",
    "    reference,\n",
    "    legend=False,\n",
    "    label=\"cost\",\n",
    "    unit=\"bn€/a\",\n",
    "    scaler=50,\n",
    "    threshold=False,\n",
    "    label_scaler=15,\n",
    "    rename=True,\n",
    "):\n",
    "    ref = df[reference]\n",
    "\n",
    "    df = (df.T - df[reference]).drop(reference).T\n",
    "\n",
    "    if rename:\n",
    "        column_dict = {\n",
    "            \"1.0\": \"w/o power grid expansion\",\n",
    "            \"opt\": \"\",\n",
    "            \"H2 grid\": \"\",\n",
    "            \"no H2 grid\": \"w/o hydrogen network\",\n",
    "        }\n",
    "\n",
    "        df.rename(columns=column_dict, inplace=True)\n",
    "        df.columns = [\"\\n\".join(col).strip() for col in df.columns.values]\n",
    "        df.sort_index(axis=1, inplace=True)\n",
    "\n",
    "    fig, ax = plt.subplots(1, 1, figsize=(6.7, 2.4))\n",
    "\n",
    "    if threshold:\n",
    "        to_drop = df.index[df.abs().max(axis=1).fillna(0.0) < threshold]\n",
    "        df.drop(to_drop, inplace=True)\n",
    "\n",
    "    df.round(1).T.plot.barh(\n",
    "        ax=ax,\n",
    "        stacked=True,\n",
    "        color=tech_colors,\n",
    "        xlabel=f\"reduced\"\n",
    "        + r\" $\\leftarrow$ \"\n",
    "        + f\"{label} [{unit}]\"\n",
    "        + r\" $\\rightarrow$ \"\n",
    "        + f\"increased\",\n",
    "        legend=legend,\n",
    "    )\n",
    "\n",
    "    net = df.sum()\n",
    "    rel = df.sum() / ref.sum() * 100\n",
    "\n",
    "    netcolor = \"darkslategrey\"\n",
    "    ax.scatter(net.values, net.index, s=10, c=netcolor, alpha=0.6, edgecolor=\"none\")\n",
    "\n",
    "    for x, y in net.items():\n",
    "        ax.annotate(\n",
    "            f\"{y:.1f} | {rel[x]:.1f}%\",\n",
    "            (y, x),\n",
    "            xytext=(-12, 13),\n",
    "            textcoords=\"offset points\",\n",
    "            color=netcolor,\n",
    "            fontsize=9,\n",
    "            bbox=dict(fc=\"0.85\", boxstyle=\"square,pad=0.1\"),\n",
    "        )\n",
    "\n",
    "    plt.grid(axis=\"x\")\n",
    "\n",
    "    ylim = max(-df.where(df < 0).sum().min(), df.where(df > 0).sum().max())\n",
    "    ylim = np.ceil(ylim / scaler) * scaler\n",
    "\n",
    "    plt.xlim([-ylim, ylim])\n",
    "    plt.ylim([-0.5, 2.8])\n",
    "    plt.ylabel(\"\")\n",
    "\n",
    "    for bars in ax.containers:\n",
    "        labels = [\n",
    "            f\"{abs(v):.0f}\" if abs(v) > ylim / label_scaler else \"\"\n",
    "            for v in bars.datavalues\n",
    "        ]\n",
    "        ax.bar_label(\n",
    "            bars, labels=labels, label_type=\"center\", color=\"#444444\", fontsize=9\n",
    "        )\n",
    "\n",
    "    handles = [\n",
    "        plt.Line2D(\n",
    "            [],\n",
    "            [],\n",
    "            color=mpl.colors.to_rgba(netcolor, 0.5),\n",
    "            marker=\".\",\n",
    "            markeredgecolor=\"none\",\n",
    "            linestyle=\"None\",\n",
    "            markersize=10,\n",
    "        )\n",
    "    ]\n",
    "    legend = ax.legend(\n",
    "        handles,\n",
    "        [\"net difference\"],\n",
    "        labelcolor=netcolor,\n",
    "        fontsize=11,\n",
    "        frameon=True,\n",
    "        loc=(-0.33, -0.3),\n",
    "    )\n",
    "    ax.add_artist(legend)\n",
    "\n",
    "    plt.tight_layout()\n",
    "\n",
    "    if legend:\n",
    "        plt.legend(ncol=3, loc=(-0.35, 1.01), labelspacing=0.2)\n",
    "\n",
    "    plt.savefig(\n",
    "        OUTPUT + f\"diff-internal-{label}{suffix}.pdf\",\n",
    "        bbox_inches=Bbox([[0, 0], [7.3, 3.8]]),\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ee166d4b-dd2f-4ec9-9623-520559e23e85",
   "metadata": {},
   "outputs": [],
   "source": [
    "reference = (\"opt\", \"H2 grid\")\n",
    "\n",
    "for SCENARIO in [MAIN_SCENARIOS, IMP_SCENARIOS, SHP_SCENARIOS, COST_SCENARIOS]:\n",
    "    suffix = \"-\" + SCENARIO.split(\"-\")[-1] if SCENARIO != MAIN_SCENARIOS else \"\"\n",
    "\n",
    "    df = load_main(SCENARIO).xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    plot_internal_diff(df, reference, scaler=40, label_scaler=25, threshold=0.5)\n",
    "\n",
    "    df = load_main_supply_energy(SCENARIO).xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    plot_internal_diff(\n",
    "        df.where(df > 0),\n",
    "        reference,\n",
    "        scaler=200,\n",
    "        label_scaler=15,\n",
    "        threshold=0.1,\n",
    "        label=\"supply\",\n",
    "        unit=r\"TWh\",\n",
    "    )\n",
    "\n",
    "    plot_internal_diff(\n",
    "        -df.where(df < 0),\n",
    "        reference,\n",
    "        scaler=200,\n",
    "        label_scaler=10,\n",
    "        threshold=0.1,\n",
    "        label=\"consumption\",\n",
    "        unit=r\"TWh\",\n",
    "    )\n",
    "\n",
    "    df = load_main_supply_energy(SCENARIO, carrier=[\"co2\"], rename=False).xs(\n",
    "        100.0, level=\"onw\", axis=1\n",
    "    )\n",
    "    df = df.groupby(rename_techs_carbon_balances).sum()\n",
    "\n",
    "    plot_internal_diff(\n",
    "        df.where(df > 0),\n",
    "        reference,\n",
    "        scaler=200,\n",
    "        label_scaler=15,\n",
    "        threshold=0.1,\n",
    "        label=\"emission\",\n",
    "        unit=r\"Mt$_{CO_2}$\",\n",
    "    )\n",
    "\n",
    "    plot_internal_diff(\n",
    "        -df.where(df < 0),\n",
    "        reference,\n",
    "        scaler=200,\n",
    "        label_scaler=10,\n",
    "        threshold=0.1,\n",
    "        label=\"withdrawal\",\n",
    "        unit=r\"Mt$_{CO_2}$\",\n",
    "    )\n",
    "\n",
    "    gw, twh = load_main_capacities(SCENARIO)\n",
    "\n",
    "    df = gw.xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    plot_internal_diff(\n",
    "        df,\n",
    "        reference,\n",
    "        threshold=1,\n",
    "        scaler=500,\n",
    "        label_scaler=10,\n",
    "        label=\"capacity\",\n",
    "        unit=\"GW\",\n",
    "    )\n",
    "\n",
    "    df = twh.xs(100.0, level=\"onw\", axis=1)\n",
    "\n",
    "    plot_internal_diff(\n",
    "        df, reference, threshold=1, scaler=60, label=\"storage\", unit=\"TWh\"\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f02abb82-20b2-4f42-9e8a-f4678914ebdb",
   "metadata": {},
   "source": [
    "## H2 Network Scenarios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b0ae57ed-61f5-4bfc-bcd3-98affdf79215",
   "metadata": {},
   "outputs": [],
   "source": [
    "SCENARIO = MAIN_SCENARIOS\n",
    "\n",
    "df = load_main(SCENARIO)\n",
    "\n",
    "suffix = \"-\" + SCENARIO.split(\"-\")[-1] if SCENARIO != MAIN_SCENARIOS else \"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e763c379-e8be-4683-99af-970d85487b2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sum().sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e9282840-e8fa-4b52-9458-85b22d0e4882",
   "metadata": {},
   "outputs": [],
   "source": [
    "dff = df.sum().unstack(\"h2\")\n",
    "\n",
    "dff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b417ca11-2141-4b9e-82fa-0241c3496779",
   "metadata": {},
   "outputs": [],
   "source": [
    "rel_advantage = (((dff[\"no H2 grid\"] / dff[\"H2 grid\"] - 1) * 100).round(1)).unstack().T\n",
    "rel_advantage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "314f3240-2b68-4780-8fc2-5c19a714d77e",
   "metadata": {},
   "outputs": [],
   "source": [
    "abs_advantage = ((dff[\"no H2 grid\"] - dff[\"H2 grid\"]).round().astype(int)).unstack().T\n",
    "abs_advantage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4c58c8e5-72e4-4dc9-afa1-5a5863f7cbc5",
   "metadata": {},
   "outputs": [],
   "source": [
    "(df.sum() - df.sum().max()).sort_values().round()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "604ff07d-e746-47a3-8f04-c6d101c59302",
   "metadata": {},
   "outputs": [],
   "source": [
    "((1 - df.sum() / df.sum().max()) * 100).sort_values().round(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "17254f81-3979-4980-9faa-e5ccd7bc702c",
   "metadata": {},
   "outputs": [],
   "source": [
    "x = df.sum().xs(100.0, level=\"onw\")\n",
    "xx = x / x.min()\n",
    "xx.unstack().round(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c85bbe5c-d83e-4c92-bdaa-01733e6f1ea5",
   "metadata": {},
   "outputs": [],
   "source": [
    "dff.xs(100.0, level=\"onw\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "da43808c-8243-46cb-b64b-bc9681bca6bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "tech_colors = config[\"plotting\"][\"tech_colors\"]\n",
    "colors = [tech_colors[i] for i in df.index]\n",
    "\n",
    "colors_list = [\n",
    "    \"hydroelectricity\",\n",
    "    \"onshore wind\",\n",
    "    \"offshore wind\",\n",
    "    \"solar PV\",\n",
    "    # \"solar thermal\",\n",
    "    \"solar rooftop\",\n",
    "    \"power-to-heat\",\n",
    "    \"gas-to-power/heat\",\n",
    "    \"power-to-hydrogen\",\n",
    "    \"power-to-liquid\",\n",
    "    \"battery storage\",\n",
    "    \"hot water storage\",\n",
    "    \"carbon capture\",\n",
    "    \"H2 storage\",\n",
    "    \"steam methane reforming\",\n",
    "    \"steam methane reforming CC\",\n",
    "    \"methanation\",\n",
    "    \"electricity distribution grid\",\n",
    "    \"fossil oil and gas\",\n",
    "    \"biogas\",\n",
    "    \"solid biomass\",\n",
    "    \"direct air capture\",\n",
    "]\n",
    "\n",
    "color_map = {i: tech_colors[i] for i in colors_list}\n",
    "color_map[\"HVDC transmission\"] = \"darkseagreen\"\n",
    "color_map[\"HVAC transmission\"] = \"rosybrown\"\n",
    "\n",
    "color_map = pd.Series(color_map).sort_index(key=lambda x: x.map(str.lower)).to_dict()\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 1.5))\n",
    "handles = [mpatches.Patch(color=v, label=k) for k, v in color_map.items()]\n",
    "ax.legend(handles=handles, ncol=6)\n",
    "ax.axis(\"off\")\n",
    "plt.savefig(OUTPUT + \"color_legend.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2207cc1-083f-4d67-a38c-f63c8e32ec7a",
   "metadata": {},
   "outputs": [],
   "source": [
    "xx = enumerate(df.columns.get_level_values(\"lv\").unique()[::-1])\n",
    "yy = enumerate(df.columns.get_level_values(\"onw\").unique())\n",
    "\n",
    "fig, axs = plt.subplots(2, 2, figsize=(5.5, 5.5), sharex=True, sharey=True)\n",
    "\n",
    "kwargs = dict(stacked=True, color=tech_colors, ylim=(0, 1100), legend=False)\n",
    "\n",
    "for x, y in product(xx, yy):\n",
    "    ax = axs[x[0], y[0]]\n",
    "\n",
    "    toprow_kwargs = (\n",
    "        dict(title=f\"onshore wind\\n{int(y[1])}% potential\\n\") if x[0] == 0 else {}\n",
    "    )\n",
    "\n",
    "    ylabel_value = \"optimal\" if x[1] == \"opt\" else f\"{100 * float(x[1]) - 100:.{0}f}%\"\n",
    "\n",
    "    to_plot = df.xs((x[1], y[1]), axis=1, level=[\"lv\", \"onw\"]).T.sort_index(\n",
    "        ascending=True\n",
    "    )\n",
    "    to_plot.plot.bar(\n",
    "        ax=ax,\n",
    "        **kwargs,\n",
    "        ylabel=f\"line expansion\\n{ylabel_value}\\n\\nbn€/a\",\n",
    "        **toprow_kwargs,\n",
    "    )\n",
    "\n",
    "    ax.set_xlabel(\"\", rotation=0)\n",
    "\n",
    "    ax.tick_params(labelrotation=0)\n",
    "\n",
    "    ax.grid(axis=\"y\")\n",
    "    ax.title.set_size(11)\n",
    "\n",
    "    # ax.patch.set_visible(False)\n",
    "\n",
    "    for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "        ax.spines[i].set_visible(False)\n",
    "\n",
    "    label = f\"+ {rel_advantage.loc[y[1],x[1]]}%\\n+ {abs_advantage.loc[y[1],x[1]]}\"\n",
    "    ax.text(0.31, 825, label, size=8.5, color=\"k\")\n",
    "\n",
    "\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "fig.legend(handles, labels, bbox_to_anchor=(1.45, 0.9))\n",
    "\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig(f\"{OUTPUT}sensitivity-h2{suffix}.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "59c29c3a-7bef-403b-b748-c9e8d15fed55",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_new = df.xs(100, level=\"onw\", axis=1)\n",
    "\n",
    "dff = df_new.sum().unstack(\"h2\")\n",
    "\n",
    "dff / dff.min(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7028b4d2-7677-4fab-9fec-e9220b64a53c",
   "metadata": {},
   "outputs": [],
   "source": [
    "h2_rel_benefit = (dff.T / dff.min(axis=1) * 100 - 100).iloc[1].reset_index(drop=True)\n",
    "h2_abs_benefit = (dff.T - dff.min(axis=1)).iloc[1].reset_index(drop=True)\n",
    "\n",
    "ac_rel_benefit = (dff / dff.min(axis=0) * 100 - 100).iloc[0].reset_index(drop=True)\n",
    "ac_abs_benefit = (dff - dff.min(axis=0)).iloc[0].reset_index(drop=True)\n",
    "\n",
    "max_rel_benefit = dff.max().max() / dff.min().min() * 100 - 100\n",
    "\n",
    "max_abs_benefit = int(dff.max().max() - dff.min().min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f776bc2c-7aac-40a6-97e6-2bec742a674f",
   "metadata": {},
   "outputs": [],
   "source": [
    "ac_grid_cost = df.xs(100.0, level=\"onw\", axis=1).loc[\"transmission lines\"].round(1)\n",
    "h2_grid_cost = df.xs(100.0, level=\"onw\", axis=1).loc[\"H2 pipeline\"].round(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1079de55-15c6-476c-9b1b-b447805a2139",
   "metadata": {},
   "outputs": [],
   "source": [
    "heavy_grouping = \"\"\n",
    "if heavy_grouping == \"-grouped\":\n",
    "\n",
    "    def heavy_aggregation(i):\n",
    "        if \"solar\" in i:\n",
    "            return \"solar\"\n",
    "        if \"bio\" in i:\n",
    "            return \"biomass\"\n",
    "\n",
    "        if i in [\"power-to-hydrogen\", \"methanation\", \"power-to-liquid\"]:\n",
    "            return \"power-to-gas/liquid\"\n",
    "        if \"capture\" in i or \"CC\" in i:\n",
    "            return \"carbon capture\"\n",
    "        if i in [\"H2 pipeline\", \"electricity distribution grid\", \"transmission lines\"]:\n",
    "            return \"transmission\"\n",
    "        if \"storage\" in i:\n",
    "            return \"storage\"\n",
    "        else:\n",
    "            return i\n",
    "\n",
    "    df_new = df_new.groupby(heavy_aggregation).sum()\n",
    "    tech_colors[\"transmission\"] = tech_colors[\"transmission lines\"]\n",
    "    tech_colors[\"storage\"] = tech_colors[\"battery storage\"]\n",
    "    tech_colors[\"power-to-gas/liquid\"] = tech_colors[\"power-to-gas\"]\n",
    "    df_new = df_new.sort_values(by=(\"opt\", \"H2 grid\"), ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f7bc8e7-e0ba-4737-a351-48dedc764e79",
   "metadata": {},
   "outputs": [],
   "source": [
    "xx = enumerate(df_new.columns.get_level_values(\"lv\").unique()[::-1])\n",
    "yy = enumerate(df_new.columns.get_level_values(\"h2\").unique())\n",
    "\n",
    "fig, axs = plt.subplots(2, 2, figsize=(4, 6), sharey=True)\n",
    "\n",
    "plt.subplots_adjust(hspace=0.5, wspace=1)\n",
    "\n",
    "kwargs = dict(stacked=True, color=tech_colors, ylim=(0, 900), legend=False)\n",
    "\n",
    "for x, y in product(xx, yy):\n",
    "    ax = axs[x[0], y[0]]\n",
    "\n",
    "    toprow_kwargs = (\n",
    "        dict(\n",
    "            title=\"with\\nhydrogen grid\\n\"\n",
    "            if y[1] == \"H2 grd\"\n",
    "            else \"without\\nhydrogen grid\\n\"\n",
    "        )\n",
    "        if x[0] == 0\n",
    "        else {}\n",
    "    )\n",
    "\n",
    "    ylabel = (\n",
    "        \"with power\\ngrid expansion\\n\\nbn€/a\"\n",
    "        if x[1] == \"opt\"\n",
    "        else \"without power\\ngrid expansion\\n\\nbn€/a\"\n",
    "    )\n",
    "\n",
    "    to_plot = df_new.xs((x[1], y[1]), axis=1, level=[\"lv\", \"h2\"]).T\n",
    "\n",
    "    to_plot.plot.bar(\n",
    "        ax=ax,\n",
    "        ylabel=ylabel,\n",
    "        **kwargs,\n",
    "        **toprow_kwargs,\n",
    "    )\n",
    "\n",
    "    ax.set_xlabel(\"\", rotation=0)\n",
    "    ax.set_xticks([], [])\n",
    "\n",
    "    ax.tick_params(labelrotation=0)\n",
    "\n",
    "    ax.grid(axis=\"y\")\n",
    "    ax.title.set_size(11)\n",
    "\n",
    "    ax.set_yticks(np.arange(0, 901, 100))\n",
    "\n",
    "    print()\n",
    "\n",
    "    ax.text(-0.3, 825, f\"{to_plot.sum().sum():.0f} bn€/a\", color=\"grey\", fontsize=9.5)\n",
    "\n",
    "    for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "        ax.spines[i].set_visible(False)\n",
    "\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "fig.legend(handles, labels, bbox_to_anchor=(1.55, 0.9))\n",
    "\n",
    "fig.text(\n",
    "    0.41,\n",
    "    0.3,\n",
    "    f\"+ {h2_rel_benefit[0]:.1f}%\\n+ {h2_abs_benefit[0]:.0f} bn€/a\",\n",
    "    fontsize=11,\n",
    ")\n",
    "fig.text(\n",
    "    0.41,\n",
    "    0.76,\n",
    "    f\"+ {h2_rel_benefit[1]:.1f}%\\n+ {h2_abs_benefit[1]:.0f} bn€/a\",\n",
    "    fontsize=11,\n",
    ")\n",
    "fig.text(\n",
    "    0.11,\n",
    "    0.47,\n",
    "    f\"+ {ac_rel_benefit[0]:.1f}%\\n+ {ac_abs_benefit[0]:.0f}\\nbn€/a\",\n",
    "    fontsize=11,\n",
    ")\n",
    "fig.text(\n",
    "    0.8,\n",
    "    0.47,\n",
    "    f\"+ {ac_rel_benefit[1]:.1f}%\\n+ {ac_abs_benefit[1]:.0f}\\nbn€/a\",\n",
    "    fontsize=11,\n",
    ")\n",
    "fig.text(\n",
    "    0.33,\n",
    "    0.47,\n",
    "    f\"+ {max_rel_benefit:.1f}%\\n+ {max_abs_benefit} bn€/a\",\n",
    "    fontsize=11,\n",
    "    color=\"grey\",\n",
    ")\n",
    "\n",
    "\n",
    "def add_arrow(ax0, ax1, pos0, pos1, **arrow_kwargs):\n",
    "    ax0tr = ax0.transData  # Axis 0 -> Display\n",
    "    ax1tr = ax1.transData  # Axis 1 -> Display\n",
    "    figtr = fig.transFigure.inverted()  # Display -> Figure\n",
    "    ptB = figtr.transform(ax0tr.transform(pos0))\n",
    "    ptE = figtr.transform(ax1tr.transform(pos1))\n",
    "    arrow = FancyArrowPatch(\n",
    "        ptB,\n",
    "        ptE,\n",
    "        transform=fig.transFigure,  # Place arrow in figure coord system\n",
    "        **arrow_kwargs,\n",
    "    )\n",
    "    fig.patches.append(arrow)\n",
    "\n",
    "\n",
    "norm = mpl.colors.Normalize(vmin=0, vmax=10)\n",
    "m = cm.ScalarMappable(norm=norm, cmap=cm.cividis)\n",
    "\n",
    "arrow_style = dict(arrowstyle=\"simple\", mutation_scale=22, ec=None)\n",
    "\n",
    "add_arrow(\n",
    "    axs[0, 0],\n",
    "    axs[0, 1],\n",
    "    (0.5, 500),\n",
    "    (-0.5, 500),\n",
    "    fc=m.to_rgba(h2_rel_benefit[1]),\n",
    "    **arrow_style,\n",
    ")\n",
    "\n",
    "add_arrow(\n",
    "    axs[1, 0],\n",
    "    axs[1, 1],\n",
    "    (0.5, 500),\n",
    "    (-0.5, 500),\n",
    "    fc=m.to_rgba(h2_rel_benefit[0]),\n",
    "    **arrow_style,\n",
    ")\n",
    "\n",
    "add_arrow(\n",
    "    axs[0, 0],\n",
    "    axs[1, 0],\n",
    "    (0, 0),\n",
    "    (0, 900),\n",
    "    fc=m.to_rgba(ac_rel_benefit[0]),\n",
    "    **arrow_style,\n",
    ")\n",
    "\n",
    "add_arrow(\n",
    "    axs[0, 1],\n",
    "    axs[1, 1],\n",
    "    (0, 0),\n",
    "    (0, 900),\n",
    "    fc=m.to_rgba(ac_rel_benefit[1]),\n",
    "    **arrow_style,\n",
    ")\n",
    "\n",
    "add_arrow(\n",
    "    axs[0, 0],\n",
    "    axs[1, 1],\n",
    "    (0.5, 0),\n",
    "    (-0.5, 900),\n",
    "    fc=m.to_rgba(max_rel_benefit),\n",
    "    **arrow_style,\n",
    ")\n",
    "\n",
    "# plt.tight_layout()\n",
    "\n",
    "plt.savefig(\n",
    "    f\"{OUTPUT}sensitivity-h2-new{suffix}{heavy_grouping}.pdf\", bbox_inches=\"tight\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8b329788-f2c1-4fab-bba1-626f283b4c6f",
   "metadata": {},
   "outputs": [],
   "source": [
    "dff.round(0).astype(int).loc[\"opt\", \"H2 grid\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dec7590a-fd4d-4d0a-b1a1-a271a193e120",
   "metadata": {},
   "outputs": [],
   "source": [
    "latex[\"gridbenefitabs\"] = max_abs_benefit\n",
    "latex[\"gridbenefitrel\"] = np.round(max_rel_benefit, 1)\n",
    "latex[\"minacbenefitabs\"] = ac_abs_benefit.round().astype(int).min()\n",
    "latex[\"maxacbenefitabs\"] = ac_abs_benefit.round().astype(int).max()\n",
    "latex[\"minhybenefitabs\"] = h2_abs_benefit.round().astype(int).min()\n",
    "latex[\"maxhybenefitabs\"] = h2_abs_benefit.round().astype(int).max()\n",
    "latex[\"minacbenefitrel\"] = ac_rel_benefit.round(1).min()\n",
    "latex[\"maxacbenefitrel\"] = ac_rel_benefit.round(1).max()\n",
    "latex[\"minhybenefitrel\"] = h2_rel_benefit.round(1).min()\n",
    "latex[\"maxhybenefitrel\"] = h2_rel_benefit.round(1).max()\n",
    "latex[\"minsystemcost\"] = dff.round(0).astype(int).min().min()\n",
    "latex[\"maxsystemcost\"] = dff.round(0).astype(int).max().max()\n",
    "\n",
    "tsc = dff.round(0).astype(int)\n",
    "latex[\"acvshycost\"] = np.round(\n",
    "    tsc.loc[\"1.0\", \"H2 grid\"] / tsc.loc[\"opt\", \"no H2 grid\"] * 100 - 100, 1\n",
    ")\n",
    "\n",
    "latex[\"minaccost\"] = ac_grid_cost.min()\n",
    "latex[\"maxaccost\"] = ac_grid_cost.max()\n",
    "latex[\"minhycost\"] = h2_grid_cost.xs(\"H2 grid\", level=\"h2\").min()\n",
    "latex[\"maxhycost\"] = h2_grid_cost.xs(\"H2 grid\", level=\"h2\").max()\n",
    "latex[\"benefithyofac\"] = np.round(\n",
    "    latex[\"maxhybenefitabs\"] / latex[\"maxacbenefitabs\"] * 100, 1\n",
    ")\n",
    "latex[\"additivebenefitabs\"] = latex[\"maxhybenefitabs\"] + latex[\"maxacbenefitabs\"]\n",
    "latex[\"additivebenefitrel\"] = np.round(\n",
    "    (latex[\"maxhybenefitabs\"] + latex[\"maxacbenefitabs\"])\n",
    "    / latex[\"gridbenefitabs\"]\n",
    "    * 100\n",
    "    - 100,\n",
    "    1,\n",
    ")\n",
    "\n",
    "nohydro = (\n",
    "    df[~df.index.str.contains(\"hydroelectricity\")]\n",
    "    .sum()\n",
    "    .xs(100.0, level=\"onw\")\n",
    "    .astype(int)\n",
    ")\n",
    "latex[\"maxsystemcostnohydro\"] = nohydro.max()\n",
    "latex[\"minsystemcostnohydro\"] = nohydro.min()\n",
    "\n",
    "gw, twh = load_main_capacities(SCENARIO)\n",
    "\n",
    "gw = gw.xs(100, level=\"onw\", axis=1).round().astype(int)\n",
    "twh = twh.xs(100, level=\"onw\", axis=1).round().astype(int)\n",
    "\n",
    "max_gw = gw.max(axis=1)\n",
    "\n",
    "min_gw = gw.min(axis=1)\n",
    "\n",
    "latex[\"minoffwind\"] = min_gw[\"offshore wind\"]\n",
    "latex[\"maxoffwind\"] = max_gw[\"offshore wind\"]\n",
    "latex[\"minonwind\"] = min_gw[\"onshore wind\"]\n",
    "latex[\"maxonwind\"] = max_gw[\"onshore wind\"]\n",
    "latex[\"minsolar\"] = min_gw[\"solar PV\"] + min_gw[\"solar rooftop\"]\n",
    "latex[\"maxsolar\"] = max_gw[\"solar PV\"] + max_gw[\"solar rooftop\"]\n",
    "\n",
    "rooftop_share = (\n",
    "    gw.loc[\"solar rooftop\"] / (gw.loc[\"solar PV\"] + gw.loc[\"solar rooftop\"]) * 100\n",
    ")\n",
    "\n",
    "latex[\"meanrooftopshare\"] = int(rooftop_share.mean())\n",
    "latex[\"meanutilityshare\"] = 100 - latex[\"meanrooftopshare\"]\n",
    "\n",
    "offshore_share = (\n",
    "    gw.loc[\"offshore wind\"] / (gw.loc[\"offshore wind\"] + gw.loc[\"onshore wind\"]) * 100\n",
    ")\n",
    "\n",
    "latex[\"minoffshoreshare\"] = int(offshore_share.min())\n",
    "latex[\"maxoffshoreshare\"] = int(offshore_share.max())\n",
    "\n",
    "latex[\"minelectrolysis\"] = gw.loc[\"power-to-hydrogen\"].min()\n",
    "latex[\"maxelectrolysis\"] = gw.loc[\"power-to-hydrogen\"].max()\n",
    "\n",
    "cfs = load_main_cfs(SCENARIO).xs(100, level=\"onw\", axis=1).mul(100).astype(int)\n",
    "\n",
    "latex[\"mincfFT\"] = cfs.loc[\"Fischer-Tropsch\"].min()\n",
    "latex[\"maxcfFT\"] = cfs.loc[\"Fischer-Tropsch\"].max()\n",
    "latex[\"mincfelectrolysis\"] = cfs.loc[\"H2 Electrolysis\"].min()\n",
    "latex[\"maxcfelectrolysis\"] = cfs.loc[\"H2 Electrolysis\"].max()\n",
    "\n",
    "sel = (\"opt\", \"H2 grid\")\n",
    "\n",
    "latex[\"utilisationAC\"] = int(cfs.loc[\"AC\", sel])\n",
    "\n",
    "latex[\"utilisationHy\"] = int(\n",
    "    cfs.loc[[\"H2 pipeline retrofitted\", \"H2 pipeline\"], sel].mean()\n",
    ")\n",
    "\n",
    "latex[\"hydrogenstorageacyhyy\"] = twh.loc[\"H2 storage\"][(\"opt\", \"H2 grid\")]\n",
    "latex[\"hydrogenstorageacyhyn\"] = twh.loc[\"H2 storage\"][(\"opt\", \"no H2 grid\")]\n",
    "latex[\"hydrogenstorageacnhyy\"] = twh.loc[\"H2 storage\"][(\"1.0\", \"H2 grid\")]\n",
    "latex[\"hydrogenstorageacnhyn\"] = twh.loc[\"H2 storage\"][(\"1.0\", \"no H2 grid\")]\n",
    "\n",
    "latex[\"thermalstoragemin\"] = twh.loc[\"hot water storage\"].min()\n",
    "latex[\"thermalstoragemax\"] = twh.loc[\"hot water storage\"].max()\n",
    "\n",
    "sankey = pd.read_csv(\n",
    "    \"../results/graphics-20221227/20221227-main/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/sankey.csv\",\n",
    "    index_col=0,\n",
    ")\n",
    "sankey.value = sankey.value.astype(int)\n",
    "\n",
    "latex[\"hydrogenproduction\"] = sankey.query(\"target == 'H2'\").value.sum()\n",
    "\n",
    "ptl = [\"Fischer-Tropsch\", \"methanolisation\"]\n",
    "latex[\"ptlhydrogenusage\"] = sankey.query(\"source == 'H2' and label in @ptl\").value.sum()\n",
    "\n",
    "latex[\"ptlwasteheat\"] = sankey.query(\n",
    "    \"source == 'H2' and label in @ptl and target == 'heat'\"\n",
    ").value.sum()\n",
    "\n",
    "latex[\"hydrogenindustrydemand\"] = sankey.query(\n",
    "    \"source == 'H2' and target == 'H2 for industry demand'\"\n",
    ").value.sum()\n",
    "latex[\"hydrogentransportdemand\"] = sankey.query(\n",
    "    \"source == 'H2' and target == 'land transport fuel cell demand'\"\n",
    ").value.sum()\n",
    "\n",
    "latex[\"hydrogenlosses\"] = sankey.query(\n",
    "    \"source == 'H2' and target == 'losses'\"\n",
    ").value.sum()\n",
    "\n",
    "sankey = pd.read_csv(\n",
    "    \"../results/graphics-20221227/20221227-main/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/sankey.csv\",\n",
    "    index_col=0,\n",
    ")\n",
    "sankey.value = sankey.value.astype(int)\n",
    "\n",
    "latex[\"hydrogenfuelcell\"] = sankey.query(\n",
    "    \"source == 'H2' and label == 'H2 Fuel Cell'\"\n",
    ").value.sum()\n",
    "\n",
    "sankey = pd.read_csv(\n",
    "    \"../results/graphics-20221227/20221227-main/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/sankey.csv\",\n",
    "    index_col=0,\n",
    ")\n",
    "sankey.value = sankey.value.astype(int)\n",
    "\n",
    "latex[\"hydrogenmethanation\"] = sankey.query(\n",
    "    \"source == 'H2' and label == 'Sabatier'\"\n",
    ").value.sum()\n",
    "\n",
    "latex[\"fossilgas\"] = sankey.query(\"source == 'fossil gas'\").value.sum()\n",
    "\n",
    "latex[\"biogas\"] = sankey.query(\"source == 'biogas'\").value.sum()\n",
    "\n",
    "smr = [\"SMR CC\", \"SMR\"]\n",
    "latex[\"bluehydrogen\"] = sankey.query(\"target == 'H2' and label in @smr\").value.sum()\n",
    "\n",
    "carbon = pd.read_csv(\n",
    "    \"../results/graphics-20221227/20221227-main/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/sankey-carbon.csv\",\n",
    "    index_col=0,\n",
    ")\n",
    "carbon.value = carbon.value.astype(int)\n",
    "\n",
    "latex[\"mindac\"] = carbon.query(\"label == 'DAC'\").value.sum()\n",
    "\n",
    "carbon = pd.read_csv(\n",
    "    \"../results/graphics-20221227/20221227-main/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/sankey-carbon.csv\",\n",
    "    index_col=0,\n",
    ")\n",
    "carbon.value = carbon.value.astype(int)\n",
    "\n",
    "latex[\"maxdac\"] = carbon.query(\"label == 'DAC'\").value.sum()\n",
    "\n",
    "latex[\"maxvres\"] = (\n",
    "    load_main_energy()\n",
    "    .xs(100, level=\"onw\", axis=1)\n",
    "    .loc[[\"onshore wind\", \"offshore wind\", \"solar PV\", \"solar rooftop\"]]\n",
    "    .sum()\n",
    "    .max()\n",
    ")\n",
    "\n",
    "latex[\"acoftotalbenefit\"] = int(\n",
    "    latex[\"maxacbenefitabs\"] / latex[\"gridbenefitabs\"] * 100\n",
    ")\n",
    "latex[\"hyoftotalbenefit\"] = int(\n",
    "    latex[\"maxhybenefitabs\"] / latex[\"gridbenefitabs\"] * 100\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44b13092-421e-4a87-8cf9-a4907751a309",
   "metadata": {},
   "source": [
    "## H2 vs Electricity Grid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6da2824-2d76-4b10-8342-3d83c6c11e9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "levels = [\"lv\", \"onw\", \"h2\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8f1b7027-d171-45dd-b752-91649c68649f",
   "metadata": {},
   "outputs": [],
   "source": [
    "scenA = df.xs((\"1.0\", 100.0, \"H2 grid\"), level=levels, axis=1)\n",
    "scenB = df.xs((\"opt\", 100.0, \"no H2 grid\"), level=levels, axis=1)\n",
    "scenC = df.xs((\"1.0\", 100.0, \"no H2 grid\"), level=levels, axis=1)\n",
    "scenD = df.xs((\"opt\", 100.0, \"H2 grid\"), level=levels, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "20dff7d2",
   "metadata": {},
   "outputs": [],
   "source": [
    "scenA.columns = [\"grid expansion\\nelectricity: no\\nhydrogen: yes\"]\n",
    "scenB.columns = [\"grid expansion\\nelectricity: yes\\nhydrogen: no\"]\n",
    "scenC.columns = [\"grid expansion\\nelectricity: no\\nhydrogen: no\"]\n",
    "scenD.columns = [\"grid expansion\\nelectricity: yes\\nhydrogen: yes\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0699e3b8-dc09-44b7-8cb9-9b5279d8f929",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_plot = pd.concat([scenD, scenB, scenA, scenC], axis=1).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0348f162-abce-46c2-8d4b-dcbbf358ce65",
   "metadata": {},
   "outputs": [],
   "source": [
    "tsc = to_plot.sum(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2bebafb0-0661-40ab-a865-25f894b2c7b8",
   "metadata": {},
   "outputs": [],
   "source": [
    "base = tsc[\"grid expansion\\nelectricity: yes\\nhydrogen: no\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d5aed672",
   "metadata": {},
   "outputs": [],
   "source": [
    "diff_rel = (100 * tsc / base - 100).round(1)[\n",
    "    \"grid expansion\\nelectricity: no\\nhydrogen: yes\"\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2fa23c9b-4583-4435-a3d6-49e136949029",
   "metadata": {},
   "outputs": [],
   "source": [
    "diff_abs = (tsc - base).round(1)[\"grid expansion\\nelectricity: no\\nhydrogen: yes\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f518979-7816-4cf0-850e-211630113e10",
   "metadata": {},
   "outputs": [],
   "source": [
    "diff_abs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2dd3bd45-5a03-41d3-8cc4-9d451c884075",
   "metadata": {},
   "outputs": [],
   "source": [
    "tsc.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8ffe6a1f-5362-4f78-8d0d-0780d3d5a770",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=(9, 5))\n",
    "\n",
    "to_plot.plot.bar(\n",
    "    ax=ax,\n",
    "    stacked=True,\n",
    "    color=colors,\n",
    "    ylim=(0, 900),\n",
    "    ylabel=\"total system cost [bn€/a]\",\n",
    ")\n",
    "\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "plt.legend(handles, labels, bbox_to_anchor=(1, 1.02))\n",
    "\n",
    "label = f\"+ {diff_rel}%\"\n",
    "ax.text(1.28, 820, label, size=11, color=\"#444444\")\n",
    "\n",
    "label = f\"+ {diff_abs} bn€/a\"\n",
    "ax.text(1.28, 860, label, size=11, color=\"k\")\n",
    "\n",
    "for i, v in enumerate(tsc.astype(int).values):\n",
    "    ax.text(i - 0.07, v + 12, v, size=10, color=\"#444444\")\n",
    "\n",
    "ax.grid(axis=\"y\")\n",
    "plt.xticks(rotation=0, fontsize=11)\n",
    "plt.yticks(np.arange(0, 901, 100), fontsize=11)\n",
    "\n",
    "for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "    ax.spines[i].set_visible(False)\n",
    "\n",
    "ax.axvline(0.5, color=\"k\", linewidth=1.25, linestyle=\"--\")\n",
    "ax.axvline(2.5, color=\"k\", linewidth=1.25, linestyle=\"--\")\n",
    "\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig(OUTPUT + \"h2-vs-elec-grid.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "711f4a3f-c5fe-4a05-b571-7e5885663fc5",
   "metadata": {},
   "source": [
    "## Growing Exclusion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d72e86db-6eaf-47d4-8e19-332c61dc1d39",
   "metadata": {},
   "outputs": [],
   "source": [
    "kwargs = dict(level=[\"lv\", \"onw\", \"h2\"], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9d327c10",
   "metadata": {},
   "outputs": [],
   "source": [
    "costs = pd.read_csv(\n",
    "    DEC_SCENARIOS + \"/csvs/costs.csv\", header=[0, 1, 2, 3], index_col=[0, 1, 2]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cbff01ca-03dd-41c8-8339-81d1bc231a09",
   "metadata": {},
   "outputs": [],
   "source": [
    "costs = costs.xs((str(CLUSTERS), \"2050\"), level=[\"cluster\", \"planning_horizon\"], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bf4bf1fa-512d-4f15-ab9d-e4f6987bdeba",
   "metadata": {},
   "outputs": [],
   "source": [
    "scenA = df.xs((\"opt\", 100.0, \"H2 grid\"), **kwargs)\n",
    "scenB = df.xs((\"1.0\", 100.0, \"H2 grid\"), **kwargs)\n",
    "scenC = df.xs((\"1.0\", 0.0, \"H2 grid\"), **kwargs)\n",
    "scenD = df.xs((\"1.0\", 0.0, \"no H2 grid\"), **kwargs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0c7b3aaa-11cd-4590-8034-269ec1ae6c79",
   "metadata": {},
   "outputs": [],
   "source": [
    "scenA.columns = [\"least-cost\"]\n",
    "scenB.columns = [\"no grid expansion\"]\n",
    "scenC.columns = [\"no grid expansion\\nno onshore wind\"]\n",
    "scenD.columns = [\"no grid expansion\\nno onshore wind\\nno hydrogen grid\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4c4b4ae2-cdf1-4565-b5e8-eabf5dfae31b",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_plot = pd.concat([scenA, scenB, scenC, scenD], axis=1).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5ee7e384-2294-42db-aa78-957ff842f854",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_plot.loc[\"today\", \"today\"] = 700\n",
    "\n",
    "order = to_plot.index[:-1].insert(0, to_plot.index[-1])\n",
    "to_plot = to_plot.loc[order]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "38d8b9ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_plot.sum(axis=1) / to_plot.sum(axis=1)[\"least-cost\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a813e234-a71f-47ee-a5fb-305303138f08",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=(10, 5))\n",
    "\n",
    "to_plot.plot.bar(\n",
    "    ax=ax,\n",
    "    stacked=True,\n",
    "    color=colors,\n",
    "    ylim=(0, 1000),\n",
    "    ylabel=\"total system cost [bn€/a]\",\n",
    ")\n",
    "\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "plt.legend(handles, labels, bbox_to_anchor=(1, 1.05))\n",
    "\n",
    "ax.grid(axis=\"y\")\n",
    "plt.xticks(rotation=0, fontsize=11)\n",
    "plt.yticks(np.arange(0, 1001, 100), fontsize=11)\n",
    "\n",
    "for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "    ax.spines[i].set_visible(False)\n",
    "\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig(OUTPUT + \"growing-exclusion.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aa66bf56",
   "metadata": {},
   "source": [
    "## Decentral Scenarios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3ec7cd9c-4618-4349-837b-3b69b61ac674",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = load_decentral()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7c3ee199",
   "metadata": {},
   "outputs": [],
   "source": [
    "to_plot = df.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5fe2a30c",
   "metadata": {},
   "outputs": [],
   "source": [
    "tsc = to_plot.sum(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d671661",
   "metadata": {},
   "outputs": [],
   "source": [
    "diff_rel = (100 * tsc / tsc.min() - 100).round(1)[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "993059ab-7417-4dd6-883d-dfc1a74924d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "diff_abs = (tsc - tsc.min()).round(1)[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b9b1bad2-342b-4eba-b712-7ddc29eb26d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a6611371-bbd9-4612-bcdc-3ec92613f159",
   "metadata": {},
   "outputs": [],
   "source": [
    "tech_colors = config[\"plotting\"][\"tech_colors\"]\n",
    "colors = [tech_colors[i] for i in to_plot.columns]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "68530242",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=(6.5, 5))\n",
    "\n",
    "to_plot.plot.bar(\n",
    "    ax=ax,\n",
    "    stacked=True,\n",
    "    color=colors,\n",
    "    ylim=(0, 1000),\n",
    "    ylabel=\"total system cost [bn€/a]\",\n",
    ")\n",
    "\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "handles.reverse()\n",
    "labels.reverse()\n",
    "plt.legend(handles, labels, bbox_to_anchor=(1, 1.02))\n",
    "\n",
    "label = f\"+ {diff_rel}%\"\n",
    "ax.text(0.2, 920, label, size=11, color=\"#444444\")\n",
    "\n",
    "label = f\"+ {diff_abs} bn€/a\"\n",
    "ax.text(0.2, 960, label, size=11, color=\"k\")\n",
    "\n",
    "ax.grid(axis=\"y\")\n",
    "plt.xticks(rotation=0, fontsize=11)\n",
    "plt.yticks(np.arange(0, 1001, 100), fontsize=11)\n",
    "\n",
    "for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "    ax.spines[i].set_visible(False)\n",
    "\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig(OUTPUT + \"decentral.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fcb638c6-78a7-48f7-b917-78242314c227",
   "metadata": {
    "tags": []
   },
   "source": [
    "## TWkm & EWhkm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bd3d8b2a-2015-4a8f-b352-6b97ba22b259",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def plot_network_stats(data, kind):\n",
    "    fig, ax = plt.subplots(\n",
    "        1, 2, sharey=True, figsize=(4, 3.5), gridspec_kw={\"width_ratios\": [1.6, 1]}\n",
    "    )\n",
    "\n",
    "    kwargs = dict(stacked=True, edgecolor=\"k\", width=0.3)\n",
    "\n",
    "    nindex = [\"cost-optimal\\nexpansion\", \"no power grid\\nexpansion\"]\n",
    "\n",
    "    df = data.xs(\"yes\", level=\"h2\").filter(like=\"Electricity\")[::-1]\n",
    "    df.index = nindex\n",
    "    color = [\"#abd982\", \"#d2ebbc\"] if kind == \"ewhkm\" else [\"#6c9459\", \"#a6be9b\"]\n",
    "    df.plot.bar(ax=ax[0], color=color, position=1, **kwargs)\n",
    "\n",
    "    dfh = data.xs(\"yes\", level=\"h2\").filter(like=\"Hydrogen\")[::-1]\n",
    "    dfh.index = nindex\n",
    "    dfh.plot.bar(ax=ax[0], color=[\"#904d84\", \"#f081dc\"], position=0, **kwargs)\n",
    "\n",
    "    df2 = data.xs(\"no\", level=\"h2\")[::-1]\n",
    "    df2.index = nindex[:2]\n",
    "    df2.plot.bar(\n",
    "        ax=ax[1],\n",
    "        color=color,\n",
    "        legend=False,\n",
    "        edgecolor=\"k\",\n",
    "        width=0.5,\n",
    "        stacked=True,\n",
    "    )\n",
    "\n",
    "    if kind == \"twkm\":\n",
    "        ax[0].set_ylabel(\"TWkm\")\n",
    "        ax[0].set_ylim([0, 900])\n",
    "        threshold = 50\n",
    "\n",
    "    if kind == \"ewhkm\":\n",
    "        ax[0].set_ylabel(\"EWhkm\")\n",
    "        ax[0].set_ylim([0, 4])\n",
    "        threshold = 0.3\n",
    "\n",
    "    ax[0].set_xlim([-0.5, 1.5])\n",
    "    ax[0].set_xlabel(\"\")\n",
    "    ax[0].set_title(\n",
    "        \"H$_2$ network\",\n",
    "        color=\"#444444\",\n",
    "    )\n",
    "    ax[1].set_title(\n",
    "        \"no H$_2$ network\",\n",
    "        color=\"#444444\",\n",
    "    )\n",
    "    ax[1].set_xlabel(\"\")\n",
    "\n",
    "    ax[0].title.set_size(11)\n",
    "    ax[1].title.set_size(11)\n",
    "\n",
    "    ax[0].grid(axis=\"y\")\n",
    "    ax[1].grid(axis=\"y\")\n",
    "\n",
    "    for i in [\"top\", \"right\", \"left\", \"bottom\"]:\n",
    "        ax[0].spines[i].set_visible(False)\n",
    "        ax[1].spines[i].set_visible(False)\n",
    "\n",
    "    for a in ax:\n",
    "        for bars in a.containers:\n",
    "            if kind == \"ewhkm\":\n",
    "                labels = [f\"{v:.2f}\" if v > threshold else \"\" for v in bars.datavalues]\n",
    "            else:\n",
    "                labels = [f\"{v:.0f}\" if v > threshold else \"\" for v in bars.datavalues]\n",
    "            a.bar_label(\n",
    "                bars, labels=labels, label_type=\"center\", color=\"#444444\", fontsize=9\n",
    "            )\n",
    "\n",
    "    plt.tight_layout()\n",
    "\n",
    "    ax[0].legend(bbox_to_anchor=(1.45, 1.6), ncol=1, labelcolor=\"#444444\")\n",
    "\n",
    "    suffix = COMMON_PATH.split(\"-\")[-1]\n",
    "    plt.savefig(f\"{OUTPUT}/{kind}-{suffix}.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "738c8ef0-5e03-4a79-8667-e4afe1b3dc76",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "COMMON_PATH = \"../results/graphics-20221227/20221227-main\"\n",
    "\n",
    "twkm = (\n",
    "    pd.concat(\n",
    "        {\n",
    "            (\"opt\", \"yes\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/twkm.csv\",\n",
    "                index_col=[0, 1],\n",
    "            ),\n",
    "            (\"opt\", \"no\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/twkm.csv\",\n",
    "                index_col=[0, 1],\n",
    "            ),\n",
    "            (\"1.0\", \"yes\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/twkm.csv\",\n",
    "                index_col=[0, 1],\n",
    "            ),\n",
    "            (\"1.0\", \"no\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/twkm.csv\",\n",
    "                index_col=[0, 1],\n",
    "            ),\n",
    "        },\n",
    "        names=[\"lv\", \"h2\"],\n",
    "    )\n",
    "    .squeeze()\n",
    "    .unstack([2, 3])\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0752f76b-a486-4757-8010-2ab3a4763065",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "twkm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "99873cfb-3ad6-46e8-938e-9589c1e251b8",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "data = pd.DataFrame()\n",
    "\n",
    "data[\"Electricity network existing\"] = (\n",
    "    twkm[(\"DC\", \"existing\")] + twkm[(\"AC\", \"existing\")]\n",
    ")\n",
    "data[\"Electricity network new\"] = twkm[(\"DC\", \"added\")] + twkm[(\"AC\", \"added\")]\n",
    "data[\"Hydrogen network retrofitted\"] = twkm[(\"H2 pipeline retrofitted\", \"optimal\")]\n",
    "data[\"Hydrogen network new\"] = twkm[(\"H2 pipeline\", \"optimal\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d8543793-bf0a-4de0-999d-17c46a1b438a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "retro_share = (\n",
    "    data[\"Hydrogen network retrofitted\"]\n",
    "    / data.filter(like=\"Hydrogen\").sum(axis=1)\n",
    "    * 100\n",
    ").round(1)\n",
    "latex[\"maxretroshare\"] = retro_share.max()\n",
    "latex[\"minretroshare\"] = retro_share.min()\n",
    "\n",
    "latex[\"maxtwkmelectricity\"] = data.filter(like=\"Electricity\").sum(axis=1).max()\n",
    "latex[\"mintwkmelectricity\"] = data.filter(like=\"Electricity\").sum(axis=1).min()\n",
    "latex[\"mintwkmhydrogen\"] = data.filter(like=\"Hydrogen\").sum(axis=1).min()\n",
    "latex[\"maxtwkmhydrogen\"] = data.filter(like=\"Hydrogen\").sum(axis=1).max()\n",
    "\n",
    "latex[\"twkmhigher\"] = np.round(\n",
    "    data[\"Electricity network new\"].max()\n",
    "    / data[\"Electricity network new\"].replace(0, np.NaN).min()\n",
    "    * 100\n",
    "    - 100,\n",
    "    1,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a69c5791-c679-4733-aab0-c4835985a9f1",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "plot_network_stats(data, \"twkm\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9b20cf91-cdb0-4a2a-a9a1-d0ebb7fa4616",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "ewhkm = (\n",
    "    pd.concat(\n",
    "        {\n",
    "            (\"opt\", \"yes\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/ewhkm.csv\",\n",
    "                index_col=0,\n",
    "            ),\n",
    "            (\"opt\", \"no\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lvopt__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/ewhkm.csv\",\n",
    "                index_col=0,\n",
    "            ),\n",
    "            (\"1.0\", \"yes\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10_2050/ewhkm.csv\",\n",
    "                index_col=0,\n",
    "            ),\n",
    "            (\"1.0\", \"no\"): pd.read_csv(\n",
    "                f\"{COMMON_PATH}/elec_s_181_lv1.0__Co2L0-3H-T-H-B-I-A-solar+p3-linemaxext10-noH2network_2050/ewhkm.csv\",\n",
    "                index_col=0,\n",
    "            ),\n",
    "        },\n",
    "        names=[\"lv\", \"h2\"],\n",
    "    )\n",
    "    .squeeze()\n",
    "    .unstack(2)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8804a77d-ae35-45e1-9904-4cc076fd42b4",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "data = pd.DataFrame()\n",
    "\n",
    "data[\"Electricity network (HVAC)\"] = ewhkm[\"AC\"]\n",
    "data[\"Electricity network (HVDC)\"] = ewhkm[\"DC\"]\n",
    "data[\"Hydrogen network retrofitted\"] = ewhkm[\"H2 pipeline retrofitted\"]\n",
    "data[\"Hydrogen network new\"] = ewhkm[\"H2 pipeline\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86043540-cef5-4728-99b0-33c0a29425f6",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "latex[\"ewhkmelectricity\"] = data.loc[(\"1.0\", \"yes\")].filter(like=\"Electricity\").sum()\n",
    "latex[\"ewhkmhydrogen\"] = data.loc[(\"1.0\", \"yes\")].filter(like=\"Hydrogen\").sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "53d96e37-4286-48d9-81f0-1d1eb98c878b",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "latex[\"ewhkmdiff\"] = np.round(\n",
    "    data.loc[(\"1.0\", \"yes\")].sum() / data.loc[(\"opt\", \"no\")].sum() * 100 - 100, 1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5aa8a8fc-9505-45e5-924c-43c177e754c3",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "latex[\"ewhkmdiff\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "394ce1bc-15e6-4164-b12c-b00fe9552f38",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "plot_network_stats(data, \"ewhkm\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0fa63603-370a-4bbc-8546-268665a04b8e",
   "metadata": {},
   "source": [
    "## Capacities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8581b5b6-1f29-4c50-a152-63189c3d3047",
   "metadata": {},
   "outputs": [],
   "source": [
    "gw, twh = load_main_capacities(scenarios=MAIN_SCENARIOS, rename=False)\n",
    "suffix = \"-\" + SCENARIO.split(\"-\")[-1] if SCENARIO != MAIN_SCENARIOS else \"\"\n",
    "df = gw.groupby(rename_techs).sum().xs(100.0, axis=1, level=\"onw\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c99154ba-c396-4dba-ae23-1329776fc43b",
   "metadata": {},
   "outputs": [],
   "source": [
    "twh = twh.groupby(rename_techs).sum().xs(100.0, axis=1, level=\"onw\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9831947b-497d-4a57-a213-ea02fab8211e",
   "metadata": {},
   "outputs": [],
   "source": [
    "index_dict = {\n",
    "    \"offwind-ac\": \"offshore wind (AC)\",\n",
    "    \"offwind-dc\": \"offshore wind (DC)\",\n",
    "    \"onwind\": \"onshore wind\",\n",
    "    \"SMR\": \"steam methane reforming\",\n",
    "    \"SMR CC\": \"steam methane reforming CC\",\n",
    "    \"battery charger\": \"battery\",\n",
    "    \"Sabatier\": \"methanation\",\n",
    "    \"H2\": \"hydrogen storage\",\n",
    "}\n",
    "\n",
    "column_dict = {\n",
    "    \"1.0\": \"w/o power\",\n",
    "    \"opt\": \"w power\",\n",
    "    \"H2 grid\": \"w H$_2$\",\n",
    "    \"no H2 grid\": \"w/o H$_2$\",\n",
    "}\n",
    "\n",
    "df.rename(index=index_dict, columns=column_dict, inplace=True)\n",
    "twh.rename(index=index_dict, columns=column_dict, inplace=True)\n",
    "\n",
    "df.columns = [\"\\n\".join(col).strip() for col in df.columns.values]\n",
    "twh.columns = [\"\\n\".join(col).strip() for col in twh.columns.values]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "53441216-df7a-4adf-b886-59fc9836bc78",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sort_index(axis=1, inplace=True)\n",
    "twh.sort_index(axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e71dcc52-b93a-4b47-881f-245b21f28c2e",
   "metadata": {},
   "outputs": [],
   "source": [
    "tech_colors = config[\"plotting\"][\"tech_colors\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "508e51c9-fe48-4c35-953f-5ea585e484d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f710eef7-a5c0-4f2f-bf27-28511f788edc",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, axs = plt.subplots(2, 4, figsize=(12, 8))\n",
    "\n",
    "groups = [\n",
    "    [\"solar rooftop\", \"solar PV\"],\n",
    "    [\"onshore wind\", \"offshore wind (AC)\", \"offshore wind (DC)\"],\n",
    "    [\"H2 Electrolysis\", \"steam methane reforming\", \"steam methane reforming CC\"],\n",
    "    [\"Fischer-Tropsch\", \"methanolisation\", \"methanation\"],\n",
    "    [\"battery storage\"],\n",
    "    [\"hydrogen storage\", \"hot water storage\"],\n",
    "    [\"ground heat pump\", \"air heat pump\", \"resistive heater\"],\n",
    "    [\"CHP\", \"OCGT\", \"gas boiler\", \"H2 Fuel Cell\"],\n",
    "]\n",
    "\n",
    "ylims = [\n",
    "    [0, 4000],\n",
    "    [0, 2200],\n",
    "    [0, 1400],\n",
    "    [0, 500],\n",
    "    [0, 300],\n",
    "    [0, 150],\n",
    "    [0, 700],\n",
    "    [0, 800],\n",
    "]\n",
    "\n",
    "for ax, group, ylim in zip(axs.flatten(), groups, ylims):\n",
    "    div = 2 if \"battery storage\" in group else 1\n",
    "    to_plot = twh if \"hydrogen storage\" in group else df\n",
    "    unit = \"TWh\" if \"hydrogen storage\" in group else \"GW\"\n",
    "    to_plot.loc[group].T.div(div).plot.bar(ax=ax, stacked=True, color=tech_colors)\n",
    "\n",
    "    ax.set_ylabel(f\"capacity [{unit}]\")\n",
    "    ax.set_ylim(ylim)\n",
    "\n",
    "    handles, labels = ax.get_legend_handles_labels()\n",
    "\n",
    "    handles.reverse()\n",
    "    labels.reverse()\n",
    "\n",
    "    ax.legend(\n",
    "        handles, labels, loc=(-0.05, 1.05), labelcolor=\"#444444\", labelspacing=0.25\n",
    "    )\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(OUTPUT + f\"capacities{suffix}.pdf\", bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65bb0f8e-a397-4709-b7b6-ef1ad2fe6eaf",
   "metadata": {},
   "source": [
    "## Export Latex Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5495b79b-7e43-4195-bcb5-1bedd3ed614f",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"../paper/variables.tex\", \"w\") as f:\n",
    "    for k, v in latex.items():\n",
    "        f.write(\"\\\\newcommand{\\\\\" + k + \"}{\" + str(v) + \"}\\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ebdb43e5-1160-430e-8b08-de5b9a572508",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6fd6ec69-7c8f-4ddd-bae5-56df88f43c14",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cb1542e0-507d-4660-a9c0-eaa050d77d54",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d714f4b9-3ca1-4de5-aba2-0921b153f765",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "beb72cfb-6ab6-4122-bcde-0ce991d75c6a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7f0b7e29-75db-405a-a542-364c8102a58b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4ad794ee-fece-4eb2-93d6-67ae6b0baf26",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "494d35a45045e3fb9015f1d2cd094f8e1540ef3ea8b346f6f51a6ceb6be203c2"
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}