JNB1-Initial_configrations.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "883181fd-a575-4b8e-ac97-9ee1b1024e98",
   "metadata": {},
   "source": [
    "# JNB1 - Build initial configuration for DFT simulations"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "5f3fa402",
   "metadata": {},
   "source": [
    "Welcome to this Jupyter Notebook, where we will explore crystal units stored in the [Material Project](https://materialsproject.org/) database. \n",
    "Our objective is to generate initial configurations for performing Density Functional Theory (DFT) simulations. To achieve this, we will be utilizing the `Pymatgen` Python library.\n",
    "\n",
    "This notebook represents the first step of the protocol discussed in the associated article (see the README.md file) as illustrated in the Figure 1 below.\n",
    "\n",
    "The notebook is organized into sections based on the type of manipulations applied to the crystals. Here is an overview of the sections:\n",
    "\n",
    "* [Get *Material Project* structures](#get_mp_structure)\n",
    "* [Pure species system](#pure_species)\n",
    "* [Supercells](#supercells)\n",
    "* [Elastic strain](#elastic_strain)\n",
    "* [Vacancies](#vacancies)\n",
    "* [Substitution](#substitution)\n",
    "* [Interstitial](#interstitial)\n",
    "* [Slabs](#slabs)\n",
    "* [Initial configuration for *ab initio* MD (DFTB) simulations](#ic_ab_initio)\n",
    "    * [Amorphous bulk](#amorphous_bulk)\n",
    "    * [Bulk at 500 K](#bulk_500)\n",
    "\n",
    "Each section focuses on specific manipulations and explores different aspects of the crystals. Our main goal is to generate initial configurations that can be used in subsequent Jupyter Notebooks for performing DFT simulations."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9880ac7b",
   "metadata": {},
   "source": [
    "<figure>\n",
    "    <img src=\"assets/img/wf-JNB1.png\" alt=\"JNB1 workflow\">\n",
    "    <figcaption><strong>Figure 1</strong>: Protocol workflow and part covered by this Jupyter Notebook highlighted in red.<figcaption>\n",
    "<figure> "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "2ccb7fd9",
   "metadata": {},
   "source": [
    "## Settings"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "86f609fe",
   "metadata": {},
   "source": [
    "In this section, we define the constants and generators that will be used throughout the notebook:\n",
    "\n",
    "* `MAIN_DIR_JNB1`: The main folder where the results of this Jupyter Notebook will be stored.\n",
    "* `CHEM_FORMULA`: The chemical formula of the SEI component.\n",
    "* `SEED`: The seed number used for the random number generator.\n",
    "* `pcg64`: The random number generator (PCG64).\n",
    "\n",
    "These constants and generators play a crucial role in ensuring reproducibility and providing the necessary inputs for various computations and simulations in this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "f139a15f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tools.misc import MAIN_DIR_JNB1\n",
    "\n",
    "CHEM_FORMULA = \"LiF\"\n",
    "SEED = 2718"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0352be92",
   "metadata": {},
   "source": [
    "## Load libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "0260fcc7-8434-408b-9a86-5bef09b6e1e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from ase.visualize import view\n",
    "from numpy.random import PCG64, Generator\n",
    "from pymatgen.analysis.defects.generators import (\n",
    "    SubstitutionGenerator,\n",
    "    VacancyGenerator,\n",
    "    VoronoiInterstitialGenerator,\n",
    ")\n",
    "from pymatgen.analysis.elasticity.strain import Strain\n",
    "from pymatgen.core.surface import SlabGenerator\n",
    "from pymatgen.io.ase import AseAtomsAdaptor\n",
    "\n",
    "from tools import make_dir\n",
    "from tools.interstitials import get_all_sites\n",
    "from tools.mp import get_crystals, load_cifs\n",
    "\n",
    "matget2ase = AseAtomsAdaptor()\n",
    "plt.style.use(\"default\")\n",
    "pcg64 = PCG64(seed=SEED)\n",
    "rg = Generator(pcg64)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "7bc61c0b",
   "metadata": {},
   "source": [
    "Making the output folder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "id": "caa7d2a0-b27b-4426-9f17-74db7581d3a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "make_dir(os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "27803353",
   "metadata": {},
   "source": [
    "## Get *Material Project* structures <a id='get_mp_structure'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "id": "a991a987",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"0-materialproject\"\n",
    "MP_CIF_DIR = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2a781d05-fc09-41c5-a0a7-1b1148a05cbf",
   "metadata": {},
   "source": [
    "Load possible structure from Material Project"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "id": "e9f1cc10",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N=3 crsital units found for the chemical formula `LiF`\n"
     ]
    }
   ],
   "source": [
    "info_df, systems, systems_ase = get_crystals(\"LiF\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "id": "68db6a8e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mp-1138</th>\n",
       "      <td>-4.845146</td>\n",
       "      <td>-3.180409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mp-1185301</th>\n",
       "      <td>-4.838474</td>\n",
       "      <td>-3.173737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mp-1009009</th>\n",
       "      <td>-4.557603</td>\n",
       "      <td>-2.892867</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            energy_per_atom  formation_energy_per_atom\n",
       "mp-1138           -4.845146                  -3.180409\n",
       "mp-1185301        -4.838474                  -3.173737\n",
       "mp-1009009        -4.557603                  -2.892867"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "info_df"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3a98db79-d14a-4b37-86d4-784e90242298",
   "metadata": {},
   "source": [
    "Store the retrieved system and info"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "id": "2e243adc-1c13-4511-b82d-f845dd5d3b9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name = []\n",
    "for i, (system, Ef_) in enumerate(zip(systems, info_df[\"formation_energy_per_atom\"])):\n",
    "    file_name.append(f\"{i:d}-{CHEM_FORMULA}_{system.get_space_group_info()[0]}_{Ef_:1.2f}.cif\".replace(\"/\", \"\"))\n",
    "    system.to(fmt=\"cif\", filename=os.path.join(MP_CIF_DIR, file_name[-1]))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "36797769-3111-4323-9993-b562c97eb4fc",
   "metadata": {},
   "source": [
    "In addition we store the info as `json` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "id": "1644170c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>file_name</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0-LiF_Fm-3m_-3.18.cif</th>\n",
       "      <td>-4.845146</td>\n",
       "      <td>-3.180409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1-LiF_P6_3mc_-3.17.cif</th>\n",
       "      <td>-4.838474</td>\n",
       "      <td>-3.173737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-LiF_Pm-3m_-2.89.cif</th>\n",
       "      <td>-4.557603</td>\n",
       "      <td>-2.892867</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        energy_per_atom  formation_energy_per_atom\n",
       "file_name                                                         \n",
       "0-LiF_Fm-3m_-3.18.cif         -4.845146                  -3.180409\n",
       "1-LiF_P6_3mc_-3.17.cif        -4.838474                  -3.173737\n",
       "2-LiF_Pm-3m_-2.89.cif         -4.557603                  -2.892867"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "info_df.loc[:, \"file_name\"] = file_name\n",
    "info_df.set_index(\"file_name\", inplace=True)\n",
    "info_df.to_json(os.path.join(MP_CIF_DIR, \"info.json\"))\n",
    "\n",
    "info_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "id": "ced12d7f-4b4b-4cd0-8a93-d867807c0e9d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 3], dpi=100, facecolor=\"w\")\n",
    "\n",
    "with plt.style.context(\"seaborn\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    Ef_np = info_df.energy_per_atom.values\n",
    "    ax.bar(info_df.index.values, info_df.energy_per_atom.values, width=0.20)\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks([i for i in range(len(info_df.index))])\n",
    "    ax.set_xticklabels(info_df.index.values, rotation=25, ha=\"right\")\n",
    "    delta_Ef = Ef_np.max() - Ef_np.min()\n",
    "    mean_Ef = (Ef_np.max() + Ef_np.min()) / 2\n",
    "    ax.set_ylim(mean_Ef - delta_Ef, mean_Ef + delta_Ef)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "1d1abcae",
   "metadata": {},
   "source": [
    "3D view of the system"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "id": "1c4e371f-8a26-451e-b87c-28bcd3f3ac4a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a095745da0854e61a62ace6aec91f2d2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=2), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li'),…"
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "fb18f5e9",
   "metadata": {},
   "source": [
    "## Pure Species system <a id='pure_species'></a>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "25a85fbc",
   "metadata": {},
   "source": [
    "When modifying a system by adding or removing atoms for the purpose of ReaxFF training, it is beneficial to compare it with a stable system. For instance, instead of relying on the absolute energy of a Li atom added to a 2x2x2 supercell, it is more effective to use the energy difference relative to a stable crystal:\n",
    "\n",
    "$\\Delta E = E(Li_9F_8) - ( 8\\times E(LiF) + 0.5 E(Li_2) )$\n",
    "\n",
    "Therefore, in this section of the Jupyter Notebook, we will query the Materials Project database to obtain stable unit crystals that consist of only a single element."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "id": "6efca06a",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"0.1-pure_species\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df9a6eaa-9f6f-4ae3-b270-76673d96af7f",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "id": "a282b6fa-827b-4d62-a08b-35cf53a11f17",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ed29e07f",
   "metadata": {},
   "source": [
    "Get unique elements in the system"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "id": "febc3f73",
   "metadata": {},
   "outputs": [],
   "source": [
    "elements = np.unique([i.name for i in systems[0][1].species])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5c92462",
   "metadata": {},
   "source": [
    "Load possible structure from Material Project"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "id": "62fabd45",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N=5 crsital units found for the chemical formula `F`\n",
      "N=8 crsital units found for the chemical formula `Li`\n"
     ]
    }
   ],
   "source": [
    "info_df = []\n",
    "systems = []\n",
    "systems_ase = []\n",
    "\n",
    "for el in elements:\n",
    "    info_df_, systems_, systems_ase_ = get_crystals(el)\n",
    "    indx = np.where(info_df_[\"energy_per_atom\"] == info_df_[\"energy_per_atom\"].min())[0][0]\n",
    "    systems.append(systems_[indx])\n",
    "    systems_ase.append(systems_ase_[indx])\n",
    "    info_df__ = info_df_[info_df_[\"energy_per_atom\"] == info_df_[\"energy_per_atom\"].min()]\n",
    "    temp_ = pd.DataFrame(columns=[\"element\", \"N\"], index=info_df__.index)\n",
    "    temp_[\"element\"] = el\n",
    "    temp_[\"N\"] = len(systems_ase_[indx])\n",
    "    info_df__ = info_df__.join(temp_)\n",
    "    info_df.append(info_df__)\n",
    "info_df = pd.concat(info_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "id": "14f90158",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "      <th>element</th>\n",
       "      <th>N</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mp-561203</th>\n",
       "      <td>-1.911456</td>\n",
       "      <td>0.0</td>\n",
       "      <td>F</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mp-1018134</th>\n",
       "      <td>-1.908923</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Li</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            energy_per_atom  formation_energy_per_atom element  N\n",
       "mp-561203         -1.911456                        0.0       F  4\n",
       "mp-1018134        -1.908923                        0.0      Li  3"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "info_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7644f7a",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "id": "5a4cf9f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "file_name = []\n",
    "for i, (system, el, nel) in enumerate(zip(systems, info_df[\"element\"], info_df[\"N\"])):\n",
    "    file_name.append(f\"{i:d}-{el}_{system.get_space_group_info()[0].replace('/','')}_na{nel:d}.cif\")\n",
    "    system.to(fmt=\"cif\", filename=os.path.join(out_dir, file_name[-1]))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4fe69e66",
   "metadata": {},
   "source": [
    "In addition we store the info as `json` file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "id": "0ca61b1d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>energy_per_atom</th>\n",
       "      <th>formation_energy_per_atom</th>\n",
       "      <th>element</th>\n",
       "      <th>N</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>file_name</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0-F_C2c_na4.cif</th>\n",
       "      <td>-1.911456</td>\n",
       "      <td>0.0</td>\n",
       "      <td>F</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1-Li_R-3m_na3.cif</th>\n",
       "      <td>-1.908923</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Li</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   energy_per_atom  formation_energy_per_atom element  N\n",
       "file_name                                                               \n",
       "0-F_C2c_na4.cif          -1.911456                        0.0       F  4\n",
       "1-Li_R-3m_na3.cif        -1.908923                        0.0      Li  3"
      ]
     },
     "execution_count": 164,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "info_df.loc[:, \"file_name\"] = file_name\n",
    "info_df.set_index(\"file_name\", inplace=True)\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "\n",
    "info_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "id": "cf24e2eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[5, 3], dpi=100, facecolor=\"w\")\n",
    "\n",
    "with plt.style.context(\"seaborn\"):\n",
    "    ax = fig.add_subplot(111)\n",
    "    Ef_np = info_df[\"energy_per_atom\"].values\n",
    "    ax.bar(file_name, Ef_np, width=0.20)\n",
    "    ax.set_ylabel(\"Energy per atoms [eV]\")\n",
    "    ax.set_xticks([i for i in range(len(Ef_np))])\n",
    "    ax.set_xticklabels(file_name, rotation=25, ha=\"right\")\n",
    "    delta_Ef = Ef_np.max() - Ef_np.min()\n",
    "    mean_Ef = (Ef_np.max() + Ef_np.min()) / 2\n",
    "    ax.set_ylim(mean_Ef - delta_Ef, mean_Ef + delta_Ef)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "id": "0409ddf1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d78b747522a84d908945d02044ffd39c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=1), VBox(children=(Dropdown(description='Show', options=('All', 'F'), value…"
      ]
     },
     "execution_count": 166,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9d02c16c-16e4-4583-b9db-5a97a1473abc",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Supercells <a id='supercells'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "id": "16859404-0834-478f-bc7b-98d16ba9723d",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"1-supercell\"\n",
    "SUB_TRAINSET_ID = 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4cad4318",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "id": "a282b6fa-827b-4d62-a08b-35cf53a11f17",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "id": "9f56c26c-c654-4974-bcd2-1ba55e3b94e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [[2, 1, 1], [2, 2, 1], [2, 2, 2], [3, 2, 2], [3, 3, 2], [3, 3, 3]]\n",
    "supercells = []\n",
    "supercells_ase = []\n",
    "supercell_info = {}\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale in supercell_scale:\n",
    "        filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_{scale[0]}x{scale[1]}x{scale[2]}.cif\"\n",
    "        supercell_info[filename] = scale[0] * scale[1] * scale[2]\n",
    "        supercell = syst_.copy()\n",
    "        supercell.make_supercell([scale[0], scale[1], scale[2]])\n",
    "        supercells.append((filename, supercell.copy()))\n",
    "        supercells_ase.append(matget2ase.get_atoms(supercell))\n",
    "        # print(filename)\n",
    "        n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "122de0ba-2f9b-4879-9c8d-c9868a918644",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "id": "8f2edf07-238e-4808-8538-150fa0d01081",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, system in supercells:\n",
    "    system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7495ffb6-3a5c-41ed-8573-8c3d702c9df4",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "id": "2eaffaf3-8032-4397-8c35-1356423c9073",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.2-0-LiF_Fm-3m_-3.18_2x2x2.cif</th>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.3-0-LiF_Fm-3m_-3.18_3x2x2.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.4-0-LiF_Fm-3m_-3.18_3x3x2.cif</th>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                unitcell_repetition\n",
       "1.0-0-LiF_Fm-3m_-3.18_2x1x1.cif                   2\n",
       "1.1-0-LiF_Fm-3m_-3.18_2x2x1.cif                   4\n",
       "1.2-0-LiF_Fm-3m_-3.18_2x2x2.cif                   8\n",
       "1.3-0-LiF_Fm-3m_-3.18_3x2x2.cif                  12\n",
       "1.4-0-LiF_Fm-3m_-3.18_3x3x2.cif                  18"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(supercell_info.keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"unitcell_repetition\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = supercell_info[file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "id": "3f3bcf83-1545-4afa-8ff7-55e8064c5824",
   "metadata": {},
   "outputs": [],
   "source": [
    "v = view(supercells_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "01f7f10b-1a40-4f15-8114-aaf34040bd6c",
   "metadata": {},
   "source": [
    "## Elastic strain <a id='elastic_strain'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "id": "ae3c2861",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"2-strain\"\n",
    "SUB_TRAINSET_ID = 2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad2a01c5",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "id": "4cf87f24",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "id": "c2601077-3297-4d50-8c18-99cc671598a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [(2, 2, 2)]\n",
    "strain_amount = np.linspace(-0.125, 0.235, 13)\n",
    "# defining the Green-Lagrange strain matrix\n",
    "# in voigt notation [σ_xx, σ_yy, σ_zz, σ_yz, σ_xz, σ_xy]\n",
    "strains = {\n",
    "    \"xx\": np.array([[1.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0.0, 0.0]]),\n",
    "    \"xy\": np.array([[0.0, 0.7, 0.0], [0.7, 0.0, 0.0], [0.0, 0.0, 0.0]]),\n",
    "    \"xxyyzz\": np.array([[1.0, 0.0, 0.0], [0.0, 1.0, 0.0], [0.0, 0.0, 1.0]]),\n",
    "}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "strain_info = {}\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale_x, scale_y, scale_z in supercell_scale:\n",
    "        supercell = syst_.copy()\n",
    "        supercell.make_supercell([scale_x, scale_y, scale_z])\n",
    "        volume = supercell.volume\n",
    "        for strain_type, strain_mtx in strains.items():\n",
    "            for amount in strain_amount:\n",
    "                filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_{strain_type}_{amount:1.3f}.cif\"\n",
    "                obj_strain = Strain(strain_mtx * amount)\n",
    "                strain_info[filename] = strain_mtx * amount\n",
    "                defomation_mtx = obj_strain.get_deformation_matrix()\n",
    "                obj_deformed = supercell.copy()\n",
    "                obj_deformed = defomation_mtx.apply_to_structure(obj_deformed)\n",
    "                new_systems.append((filename, obj_deformed.copy()))\n",
    "                new_systems_ase.append(matget2ase.get_atoms(obj_deformed))\n",
    "                n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "adbe7cf3-fc8e-4aa0-aeab-c5981f165661",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "id": "fbdbd830-e223-42da-84e0-3f3e5caf698c",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dbb7939b-13d1-4b10-9bec-9d5b933fabed",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "id": "32ab53d1-a1dc-47f5-ba8c-319d54f26291",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>strain_matrix</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2.0-0-LiF_Fm-3m_-3.18_xx_-0.125.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.125, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.1-0-LiF_Fm-3m_-3.18_xx_-0.095.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.095, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.2-0-LiF_Fm-3m_-3.18_xx_-0.065.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.065, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.3-0-LiF_Fm-3m_-3.18_xx_-0.035.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.035, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.4-0-LiF_Fm-3m_-3.18_xx_-0.005.cif</th>\n",
       "      <td>8</td>\n",
       "      <td>[[-0.0050000000000000044, -0.0, -0.0], [-0.0, ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    unitcell_repetition  \\\n",
       "2.0-0-LiF_Fm-3m_-3.18_xx_-0.125.cif                   8   \n",
       "2.1-0-LiF_Fm-3m_-3.18_xx_-0.095.cif                   8   \n",
       "2.2-0-LiF_Fm-3m_-3.18_xx_-0.065.cif                   8   \n",
       "2.3-0-LiF_Fm-3m_-3.18_xx_-0.035.cif                   8   \n",
       "2.4-0-LiF_Fm-3m_-3.18_xx_-0.005.cif                   8   \n",
       "\n",
       "                                                                         strain_matrix  \n",
       "2.0-0-LiF_Fm-3m_-3.18_xx_-0.125.cif  [[-0.125, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...  \n",
       "2.1-0-LiF_Fm-3m_-3.18_xx_-0.095.cif  [[-0.095, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...  \n",
       "2.2-0-LiF_Fm-3m_-3.18_xx_-0.065.cif  [[-0.065, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...  \n",
       "2.3-0-LiF_Fm-3m_-3.18_xx_-0.035.cif  [[-0.035, -0.0, -0.0], [-0.0, -0.0, -0.0], [-0...  \n",
       "2.4-0-LiF_Fm-3m_-3.18_xx_-0.005.cif  [[-0.0050000000000000044, -0.0, -0.0], [-0.0, ...  "
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(strain_info.keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"unitcell_repetition\", \"strain_matrix\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"strain_matrix\"] = strain_info[file_name_]\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = np.prod(supercell_scale)\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "id": "f88da728-ffd7-407c-b68b-b887b27d70ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "386d4844510f4834995e830476bc8219",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=116), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li'…"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "399ae74a-de54-4847-88b7-38b3dcb98f11",
   "metadata": {},
   "source": [
    "## Vacancies <a id='vacancies'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "id": "fa245f92-5d8d-4285-8c6f-a1e056e236a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"3-vacancies\"\n",
    "SUB_TRAINSET_ID = 3"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82ab7b4b-d5be-45c0-b3f0-5751e0f233de",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "id": "a41a66a5-c8a2-4e32-9ad8-1fc1f53c7097",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0d62eb70-69ed-481b-bb42-371b742d5aaf",
   "metadata": {},
   "source": [
    "Random removing Li atoms from in 2x2x2 supercell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "id": "2b2878da-33f8-42e0-a86d-07764ed1bf9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [[2, 2, 1]]\n",
    "max_vacacy = 5\n",
    "vacancy_info = {\"site\": {}, \"element\": {}, \"unitcell_repetition\": {}}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale in supercell_scale:\n",
    "        supercell = syst_.copy()\n",
    "        supercell.make_supercell([scale[0], scale[1], scale[2]])\n",
    "        temporary = supercell.copy()\n",
    "        prev_vacancy_el = []\n",
    "        prev_vacancy_site = []\n",
    "        for nv in range(max_vacacy):\n",
    "            primitive_elements = [e.species_string for e in supercell.get_primitive_structure().sites]\n",
    "            vacancy_gen = VacancyGenerator().generate(temporary)\n",
    "            temporary = []\n",
    "            for v_site in list(vacancy_gen):\n",
    "                site_el = v_site.site.species_string\n",
    "                if site_el in primitive_elements:\n",
    "                    primitive_elements.remove(site_el)\n",
    "                elif len(primitive_elements) == 0:\n",
    "                    break\n",
    "                else:\n",
    "                    continue\n",
    "                filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_nv_{nv+1}.cif\"\n",
    "                vacancy_info[\"element\"][filename] = prev_vacancy_el + [v_site.site.species_string]\n",
    "                vacancy_info[\"site\"][filename] = prev_vacancy_site + [v_site.site.coords]\n",
    "                vacancy_info[\"unitcell_repetition\"][filename] = np.prod(scale)\n",
    "                obj_with_vacancy = v_site.defect_structure.copy()\n",
    "                new_systems.append((filename, obj_with_vacancy.copy()))\n",
    "                new_systems_ase.append(matget2ase.get_atoms(obj_with_vacancy))\n",
    "                n += 1\n",
    "                temporary.append(obj_with_vacancy.copy())\n",
    "            # Store some data\n",
    "            Nidx = len(temporary)\n",
    "            keys = list(vacancy_info[\"element\"].keys())\n",
    "            idx = rg.choice(Nidx)\n",
    "            temporary = temporary[idx]\n",
    "            key = keys[-Nidx:][idx]\n",
    "            prev_vacancy_el = vacancy_info[\"element\"][key]\n",
    "            prev_vacancy_site = vacancy_info[\"site\"][key]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "255adf23-bfb2-4b06-8d07-6a46f641d899",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "id": "0f931d13-988c-409c-b3b5-ec068caac9f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3eb62bda-577d-4736-bdd1-905d6f52e949",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "id": "2d1ef04c-5223-47ba-8050-10e58f391186",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vacancy_site</th>\n",
       "      <th>vacancy_element</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3.0-0-LiF_Fm-3m_-3.18_nv_1.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0]]</td>\n",
       "      <td>[Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.1-0-LiF_Fm-3m_-3.18_nv_1.cif</th>\n",
       "      <td>[[1.6670525006439094, 1.1787841277992999, 2.88...</td>\n",
       "      <td>[F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.2-0-LiF_Fm-3m_-3.18_nv_2.cif</th>\n",
       "      <td>[[1.6670525006439094, 1.1787841277992999, 2.88...</td>\n",
       "      <td>[F, Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.3-0-LiF_Fm-3m_-3.18_nv_2.cif</th>\n",
       "      <td>[[1.6670525006439094, 1.1787841277992999, 2.88...</td>\n",
       "      <td>[F, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.4-0-LiF_Fm-3m_-3.18_nv_3.cif</th>\n",
       "      <td>[[1.6670525006439094, 1.1787841277992999, 2.88...</td>\n",
       "      <td>[F, Li, Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                     vacancy_site  \\\n",
       "3.0-0-LiF_Fm-3m_-3.18_nv_1.cif                                  [[0.0, 0.0, 0.0]]   \n",
       "3.1-0-LiF_Fm-3m_-3.18_nv_1.cif  [[1.6670525006439094, 1.1787841277992999, 2.88...   \n",
       "3.2-0-LiF_Fm-3m_-3.18_nv_2.cif  [[1.6670525006439094, 1.1787841277992999, 2.88...   \n",
       "3.3-0-LiF_Fm-3m_-3.18_nv_2.cif  [[1.6670525006439094, 1.1787841277992999, 2.88...   \n",
       "3.4-0-LiF_Fm-3m_-3.18_nv_3.cif  [[1.6670525006439094, 1.1787841277992999, 2.88...   \n",
       "\n",
       "                               vacancy_element unitcell_repetition  \n",
       "3.0-0-LiF_Fm-3m_-3.18_nv_1.cif            [Li]                   4  \n",
       "3.1-0-LiF_Fm-3m_-3.18_nv_1.cif             [F]                   4  \n",
       "3.2-0-LiF_Fm-3m_-3.18_nv_2.cif         [F, Li]                   4  \n",
       "3.3-0-LiF_Fm-3m_-3.18_nv_2.cif          [F, F]                   4  \n",
       "3.4-0-LiF_Fm-3m_-3.18_nv_3.cif     [F, Li, Li]                   4  "
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(vacancy_info[\"element\"].keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"vacancy_site\", \"vacancy_element\", \"unitcell_repetition\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = vacancy_info[\"unitcell_repetition\"][file_name_]\n",
    "    info_df.loc[file_name_, \"vacancy_site\"] = vacancy_info[\"site\"][file_name_]\n",
    "    info_df.loc[file_name_, \"vacancy_element\"] = vacancy_info[\"element\"][file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "id": "7243e27e-38bd-41f9-a716-8b9a1389d69e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "403d328f1f784000aafeffc4cdffbbe4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=37), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li')…"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d3bd7234-5e8f-4cc1-a305-59b029bba6bb",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Substitution <a id='substitution'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "id": "5734765d-9851-4102-9f79-6d1c204bf608",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"4-substitutions\"\n",
    "SUB_TRAINSET_ID = 4"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42015aac-488c-43be-bc20-02086f1d01b8",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "id": "b2092822-e075-4f4c-9e5a-1244d8e9feaf",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5379149b-2b32-4ac5-bab8-d99f4f325d98",
   "metadata": {},
   "source": [
    "Random substitue a Atom in 2x2x2 supercell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "id": "38db4431-ba64-473a-bc5e-ef8a338d474f",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [[2, 2, 1]]\n",
    "max_substitution = 8\n",
    "substitution_info = {\"site\": {}, \"element\": {}, \"unitcell_repetition\": {}}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale in supercell_scale:\n",
    "        supercell = syst_.copy()\n",
    "        supercell.make_supercell([scale[0], scale[1], scale[2]])\n",
    "        temporary = supercell.copy()\n",
    "        prev_substitution_el = []\n",
    "        prev_substitution_site = []\n",
    "        for ns in range(max_substitution):\n",
    "            substitution_gen = SubstitutionGenerator().generate(temporary, substitution={\"Li\": \"F\", \"F\": \"Li\"})\n",
    "            filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_ns_{ns+1}.cif\"\n",
    "            sub_site = list(substitution_gen)\n",
    "            n_sub_site = len(sub_site)\n",
    "            idx = rg.choice(n_sub_site)\n",
    "            sub_site = sub_site[idx]\n",
    "            substitution_info[\"element\"][filename] = prev_substitution_el + [sub_site.site.species_string]\n",
    "            substitution_info[\"site\"][filename] = prev_substitution_site + [sub_site.site.coords]\n",
    "            substitution_info[\"unitcell_repetition\"][filename] = np.prod(scale)\n",
    "            obj_with_substitution = sub_site.defect_structure.copy()\n",
    "            new_systems.append((filename, obj_with_substitution.copy()))\n",
    "            new_systems_ase.append(matget2ase.get_atoms(obj_with_substitution))\n",
    "            prev_substitution_el = substitution_info[\"element\"][filename]\n",
    "            prev_substitution_site = substitution_info[\"site\"][filename]\n",
    "            temporary = obj_with_substitution.copy()\n",
    "            n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "592e53d3-5546-4071-92d6-f7863179543f",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "id": "6bfe1df5-138f-48d6-ae8c-7872a440fd8c",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "12fec1cf-f09d-4562-b3f0-f56bdfaf611f",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "id": "14d8bac0-2e0f-40ea-a825-5759be2a06c4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>substitution_site</th>\n",
       "      <th>substitution_element</th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4.0-0-LiF_Fm-3m_-3.18_ns_1.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0]]</td>\n",
       "      <td>[F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.1-0-LiF_Fm-3m_-3.18_ns_2.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...</td>\n",
       "      <td>[F, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.2-0-LiF_Fm-3m_-3.18_ns_3.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...</td>\n",
       "      <td>[F, F, Li]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.3-0-LiF_Fm-3m_-3.18_ns_4.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...</td>\n",
       "      <td>[F, F, Li, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.4-0-LiF_Fm-3m_-3.18_ns_5.cif</th>\n",
       "      <td>[[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...</td>\n",
       "      <td>[F, F, Li, F, F]</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                substitution_site  \\\n",
       "4.0-0-LiF_Fm-3m_-3.18_ns_1.cif                                  [[0.0, 0.0, 0.0]]   \n",
       "4.1-0-LiF_Fm-3m_-3.18_ns_2.cif  [[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...   \n",
       "4.2-0-LiF_Fm-3m_-3.18_ns_3.cif  [[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...   \n",
       "4.3-0-LiF_Fm-3m_-3.18_ns_4.cif  [[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...   \n",
       "4.4-0-LiF_Fm-3m_-3.18_ns_5.cif  [[0.0, 0.0, 0.0], [0.833526250321955, 2.357568...   \n",
       "\n",
       "                               substitution_element unitcell_repetition  \n",
       "4.0-0-LiF_Fm-3m_-3.18_ns_1.cif                  [F]                   4  \n",
       "4.1-0-LiF_Fm-3m_-3.18_ns_2.cif               [F, F]                   4  \n",
       "4.2-0-LiF_Fm-3m_-3.18_ns_3.cif           [F, F, Li]                   4  \n",
       "4.3-0-LiF_Fm-3m_-3.18_ns_4.cif        [F, F, Li, F]                   4  \n",
       "4.4-0-LiF_Fm-3m_-3.18_ns_5.cif     [F, F, Li, F, F]                   4  "
      ]
     },
     "execution_count": 189,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(substitution_info[\"element\"].keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"substitution_site\", \"substitution_element\", \"unitcell_repetition\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = substitution_info[\"unitcell_repetition\"][file_name_]\n",
    "    info_df.loc[file_name_, \"substitution_site\"] = substitution_info[\"site\"][file_name_]\n",
    "    info_df.loc[file_name_, \"substitution_element\"] = substitution_info[\"element\"][file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "id": "ca1b4bac-4ffe-4c33-9d2b-948b8c589ec2",
   "metadata": {},
   "outputs": [],
   "source": [
    "v = view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "913a4404-084d-4ca4-9bac-04a6330bd15f",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Interstitial <a id='interstitial'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "id": "8f1ef7b9-807c-4f65-b815-c3e813b8fd6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"5-interstitial\"\n",
    "SUB_TRAINSET_ID = 5"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e9a00c2e-f34f-4414-aa8d-189a6e3d9bf5",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "id": "f6cbf7a9-1d03-44a5-ba44-5a0f41329409",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "be1278f8-c189-45b5-9ddf-cad1be347de9",
   "metadata": {},
   "source": [
    "Random substitue a Atom in 2x2x2 supercell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "id": "ae33cb84-f08d-4cc8-86de-2e0f80938bd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "pcg64 = PCG64(seed=SEED + SUB_TRAINSET_ID)\n",
    "rg = Generator(pcg64)\n",
    "\n",
    "supercell_scale = [[2, 2, 1]]\n",
    "N_interstitial = np.arange(7)\n",
    "interstitial_info = {\"site\": {}, \"element\": {}, \"unitcell_repetition\": {}}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale in supercell_scale:\n",
    "        lattice = syst_.lattice\n",
    "        supercell = syst_.copy()\n",
    "        interstitial_gen = VoronoiInterstitialGenerator().generate(syst_, insert_species=[\"Li\"])\n",
    "        sites = list(interstitial_gen)\n",
    "        r_sites = get_all_sites(sites, lattice, supercell=scale)\n",
    "        rg.shuffle(r_sites, axis=0)\n",
    "        supercell.make_supercell(scale)\n",
    "        temporary = supercell.copy()\n",
    "        prev_interstitial_el = []\n",
    "        prev_interstitial_site = []\n",
    "        for ni in N_interstitial:\n",
    "            filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_ni_{ni}.cif\"\n",
    "            if ni == 0:\n",
    "                interstitial_info[\"element\"][filename] = []\n",
    "                interstitial_info[\"site\"][filename] = []\n",
    "                interstitial_info[\"unitcell_repetition\"][filename] = np.prod(scale)\n",
    "            else:\n",
    "                n_sites = len(r_sites)\n",
    "                if n_sites > 0:\n",
    "                    idx = rg.choice(n_sites)\n",
    "                    intestitial_r = r_sites.pop(idx)\n",
    "                    interstitial_info[\"element\"][filename] = prev_interstitial_el + [\"Li\"]\n",
    "                    interstitial_info[\"site\"][filename] = prev_interstitial_site + [intestitial_r]\n",
    "                    interstitial_info[\"unitcell_repetition\"][filename] = np.prod(scale)\n",
    "                    temporary.append(\"Li\", intestitial_r, coords_are_cartesian=True)\n",
    "                    temporary.set_charge(0.0)\n",
    "                else:\n",
    "                    continue\n",
    "            new_systems.append((filename, temporary.copy()))\n",
    "            new_systems_ase.append(matget2ase.get_atoms(temporary))\n",
    "            prev_interstitial_el = interstitial_info[\"element\"][filename]\n",
    "            prev_interstitial_site = interstitial_info[\"site\"][filename]\n",
    "            n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ddf1f49f-396a-42fc-af70-f418e8250337",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "id": "af3e5b59-4cca-48b3-a921-0fdb3aad6539",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c893a8e2-68f3-4f0e-8a4b-d481a120f8b5",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "id": "aeb5932f-aa93-46f8-9460-7a396f1eea43",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "      <th>interstitial_site</th>\n",
       "      <th>interstitial_element</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5.0-0-LiF_Fm-3m_-3.18_ni_0.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[]</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.1-0-LiF_Fm-3m_-3.18_ni_1.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.833526250321955, 0.5893920638996494, 1.443...</td>\n",
       "      <td>[Li]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.2-0-LiF_Fm-3m_-3.18_ni_2.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.833526250321955, 0.5893920638996494, 1.443...</td>\n",
       "      <td>[Li, Li]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.3-0-LiF_Fm-3m_-3.18_ni_3.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.833526250321955, 0.5893920638996494, 1.443...</td>\n",
       "      <td>[Li, Li, Li]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.4-0-LiF_Fm-3m_-3.18_ni_4.cif</th>\n",
       "      <td>4</td>\n",
       "      <td>[[0.833526250321955, 0.5893920638996494, 1.443...</td>\n",
       "      <td>[Li, Li, Li, Li]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               unitcell_repetition  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                   4   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                   4   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif                   4   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif                   4   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif                   4   \n",
       "\n",
       "                                                                interstitial_site  \\\n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                                                 []   \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif  [[0.833526250321955, 0.5893920638996494, 1.443...   \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif  [[0.833526250321955, 0.5893920638996494, 1.443...   \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif  [[0.833526250321955, 0.5893920638996494, 1.443...   \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif  [[0.833526250321955, 0.5893920638996494, 1.443...   \n",
       "\n",
       "                               interstitial_element  \n",
       "5.0-0-LiF_Fm-3m_-3.18_ni_0.cif                   []  \n",
       "5.1-0-LiF_Fm-3m_-3.18_ni_1.cif                 [Li]  \n",
       "5.2-0-LiF_Fm-3m_-3.18_ni_2.cif             [Li, Li]  \n",
       "5.3-0-LiF_Fm-3m_-3.18_ni_3.cif         [Li, Li, Li]  \n",
       "5.4-0-LiF_Fm-3m_-3.18_ni_4.cif     [Li, Li, Li, Li]  "
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(interstitial_info[\"element\"].keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"unitcell_repetition\", \"interstitial_site\", \"interstitial_element\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"interstitial_site\"] = interstitial_info[\"site\"][file_name_]\n",
    "    info_df.loc[file_name_, \"interstitial_element\"] = interstitial_info[\"element\"][file_name_]\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = interstitial_info[\"unitcell_repetition\"][file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "id": "b1090240-676e-4982-beb0-a83ed51df616",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0c8e5cd0b6aa49029983ed30af62554a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=16), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li')…"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "573974bb-eff9-468e-86b8-134cea420972",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Slabs <a id='slabs'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "id": "0b3de943-57e7-409a-bde0-ac8055f597a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"6-slabs\"\n",
    "SUB_TRAINSET_ID = 6"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0666f782-0d5e-48e8-9305-9756da13ff92",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "id": "8301a81e-2102-4568-ac37-06020ce8d8c8",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "id": "c360d2f3-8b90-48bb-8d1a-29340dd787f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [1, 1, 1]\n",
    "Nslabs = [1, 2, 3]\n",
    "millers = [(0, 0, 1), (0, 1, 0), (1, 0, 0), (0, 1, 1), (1, 0, 1), (1, 1, 0), (1, 1, 1), (1, 1, 2), (1, 2, 1), (2, 1, 1)]\n",
    "min_vacuum_size = 10  # [Ang]\n",
    "slab_info = {\"millers\": {}, \"slabs\": {}}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for hkl in millers:\n",
    "        for n_slabs in Nslabs:\n",
    "            hkl_str = \"\".join([str(index) for index in hkl])\n",
    "            filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_{hkl_str}x{n_slabs}.cif\"\n",
    "            lc = syst_.lattice.parameters[3]\n",
    "            vacuum_size = max([int(np.ceil(min_vacuum_size / lc) + 1), 2])\n",
    "            slabgen = SlabGenerator(\n",
    "                syst_,\n",
    "                miller_index=hkl,\n",
    "                min_slab_size=n_slabs,\n",
    "                in_unit_planes=True,\n",
    "                min_vacuum_size=vacuum_size,\n",
    "                lll_reduce=False,\n",
    "                center_slab=True,\n",
    "            )\n",
    "            slab = slabgen.get_slab()\n",
    "            slab.make_supercell(supercell_scale)\n",
    "            slab_info[\"millers\"][filename] = hkl\n",
    "            slab_info[\"slabs\"][filename] = n_slabs\n",
    "            new_systems.append((filename, slab.copy()))\n",
    "            new_systems_ase.append(matget2ase.get_atoms(slab.copy()))\n",
    "            n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2283d13f-42c3-40db-b450-92014ee8f06b",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "id": "07e8640b-35c6-473f-9fc3-3382c043163f",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab82683a-7d33-4599-8842-e4124ad3bb40",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "id": "50bb7571-ed7e-4f35-89cb-335f2d03ec9a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>millers</th>\n",
       "      <th>slabs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6.0-0-LiF_Fm-3m_-3.18_001x1.cif</th>\n",
       "      <td>(0, 0, 1)</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.1-0-LiF_Fm-3m_-3.18_001x2.cif</th>\n",
       "      <td>(0, 0, 1)</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.2-0-LiF_Fm-3m_-3.18_001x3.cif</th>\n",
       "      <td>(0, 0, 1)</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.3-0-LiF_Fm-3m_-3.18_010x1.cif</th>\n",
       "      <td>(0, 1, 0)</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.4-0-LiF_Fm-3m_-3.18_010x2.cif</th>\n",
       "      <td>(0, 1, 0)</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   millers slabs\n",
       "6.0-0-LiF_Fm-3m_-3.18_001x1.cif  (0, 0, 1)     1\n",
       "6.1-0-LiF_Fm-3m_-3.18_001x2.cif  (0, 0, 1)     2\n",
       "6.2-0-LiF_Fm-3m_-3.18_001x3.cif  (0, 0, 1)     3\n",
       "6.3-0-LiF_Fm-3m_-3.18_010x1.cif  (0, 1, 0)     1\n",
       "6.4-0-LiF_Fm-3m_-3.18_010x2.cif  (0, 1, 0)     2"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(slab_info[\"millers\"].keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"millers\", \"slabs\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"millers\"] = slab_info[\"millers\"][file_name_]\n",
    "    info_df.loc[file_name_, \"slabs\"] = slab_info[\"slabs\"][file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "id": "ac3cab2a-d99e-4d2c-825f-469493fb1be2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8a1754a77e8a400caa9ae3cc33da7739",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=89), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li')…"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "efa1182b",
   "metadata": {},
   "source": [
    "## Initial configurationf for *ab intio* MD (DFTB) simulations <a id='ic_ab_initio'></a>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d4a04ded",
   "metadata": {},
   "source": [
    "In this section, we initialize a 2x2x2 supercell, which will be used for *ab initio* Density Functional Tight Binding (DFTB) simulations. Frames will be extracted from this supercell and included in the database."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d9335f0c-1c66-4e0c-b86f-5a3995429be1",
   "metadata": {},
   "source": [
    "### Amorphous bulk  <a id='amorphous_bulk '></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "id": "1df844b9-a789-49ae-811d-1727092fe4ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"7-amorphous\"\n",
    "SUB_TRAINSET_ID = 7"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e455add-b6f6-48f8-a6b2-6eba262bbe34",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "id": "839580af-3221-4c38-8e86-326358f3faee",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "id": "5540e656-e92f-4a02-b1bc-fa02827f24c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [[3, 3, 3]]\n",
    "min_vacuum_size = 10  # [Ang]\n",
    "supercell_info = {}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale in supercell_scale:\n",
    "        supercell = syst_.copy()\n",
    "        filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_{scale[0]}x{scale[1]}x{scale[2]}.cif\"\n",
    "        supercell_info[filename] = scale[0] * scale[1] * scale[2]\n",
    "        supercell.make_supercell(scale)\n",
    "        new_systems.append((filename, supercell.copy()))\n",
    "        new_systems_ase.append(matget2ase.get_atoms(supercell))\n",
    "        n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df3d259c-b94d-4b29-b87c-b4c6d8be49b6",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "id": "9e445c13-dd39-43c9-a4cd-eda1c43fae09",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "626fdf69-a00e-4f17-bc8a-f5b83e671f92",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "id": "e514e6b1-e76d-48a2-b2b7-2b20694b2f9e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif</th>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 unitcell_repetition\n",
       "7.0-0-LiF_Fm-3m_-3.18_3x3x3.cif                   27\n",
       "7.0-1-LiF_P6_3mc_-3.17_3x3x3.cif                  27\n",
       "7.0-2-LiF_Pm-3m_-2.89_3x3x3.cif                   27"
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(supercell_info.keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"unitcell_repetition\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = supercell_info[file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "id": "e9bcef57",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "70375b7d0ee24918942aaa8e4b72f5f6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=2), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li'),…"
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "78031031-5392-42c5-8b88-4dda4fed3bea",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Bulk at 300 K  <a id='bulk_300'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "id": "cf5d850f-4f65-4b62-8ba6-fc04722b1e68",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"8-bulk_300K\"\n",
    "SUB_TRAINSET_ID = 8"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "022668d7-d425-4573-9143-b96f3cb6dc33",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "id": "f31f3622-ddcc-417e-af4a-90649ec12f7a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "id": "03bdc465-0013-440a-b8ac-d057099a699e",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [[2, 2, 3]]\n",
    "min_vacuum_size = 10  # [Ang]\n",
    "supercell_info = {}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale in supercell_scale:\n",
    "        supercell = syst_.copy()\n",
    "        filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_{scale[0]}x{scale[1]}x{scale[2]}.cif\"\n",
    "        supercell_info[filename] = scale[0] * scale[1] * scale[2]\n",
    "        supercell.make_supercell([scale[0], scale[1], scale[2]])\n",
    "        new_systems.append((filename, supercell.copy()))\n",
    "        new_systems_ase.append(matget2ase.get_atoms(supercell))\n",
    "        n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3db8f80d-da91-44b8-9edd-c311c7f1323c",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "id": "718c6def-02a4-485b-9a9e-5409f6dd8b02",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08bd1d5b-1855-4037-8933-360c260349a1",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "id": "fe7b240e-3b85-4a73-94bf-84aa7bf3dd8a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 unitcell_repetition\n",
       "8.0-0-LiF_Fm-3m_-3.18_2x2x3.cif                   12\n",
       "8.0-1-LiF_P6_3mc_-3.17_2x2x3.cif                  12\n",
       "8.0-2-LiF_Pm-3m_-2.89_2x2x3.cif                   12"
      ]
     },
     "execution_count": 213,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(supercell_info.keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"unitcell_repetition\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = supercell_info[file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "id": "cc73a6a1-a7dd-4cea-9cca-33dbf21c4d78",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "145a3ad499d34cfa9e2796e4c1e11146",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=2), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li'),…"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "3519be27",
   "metadata": {},
   "source": [
    "### Bulk at 500 K  <a id='bulk_500'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "id": "ade2f021",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUBDIR = \"9-bulk_500K\"\n",
    "SUB_TRAINSET_ID = 9"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "355d6b58",
   "metadata": {},
   "source": [
    "Load structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "id": "8b796e4c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paolo/Enhancing_ReaxFF/venv/lib/python3.8/site-packages/pymatgen/io/cif.py:1134: UserWarning: Issues encountered while parsing CIF: Some fractional coordinates rounded to ideal values to avoid issues with finite precision.\n",
      "  warnings.warn(\"Issues encountered while parsing CIF: \" + \"\\n\".join(self.warnings))\n"
     ]
    }
   ],
   "source": [
    "systems = load_cifs(MP_CIF_DIR)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5cde082d",
   "metadata": {},
   "source": [
    "Random substitue a Atom in 2x2x2 supercell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "id": "d7971ada",
   "metadata": {},
   "outputs": [],
   "source": [
    "supercell_scale = [[2, 2, 3]]\n",
    "min_vacuum_size = 10  # [Ang]\n",
    "supercell_info = {}\n",
    "new_systems = []\n",
    "new_systems_ase = []\n",
    "\n",
    "for i, (name, syst_, syst_ase_) in enumerate(systems):\n",
    "    n = 0\n",
    "    for scale in supercell_scale:\n",
    "        supercell = syst_.copy()\n",
    "        filename = f\"{SUB_TRAINSET_ID}.{n}-{i}-{name}_{scale[0]}x{scale[1]}x{scale[2]}.cif\"\n",
    "        supercell_info[filename] = scale[0] * scale[1] * scale[2]\n",
    "        supercell.make_supercell([scale[0], scale[1], scale[2]])\n",
    "        new_systems.append((filename, supercell.copy()))\n",
    "        new_systems_ase.append(matget2ase.get_atoms(supercell))\n",
    "        n += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd7b571a",
   "metadata": {},
   "source": [
    "Store it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "af636a3e",
   "metadata": {},
   "outputs": [],
   "source": [
    "out_dir = os.path.join(MAIN_DIR_JNB1, CHEM_FORMULA, SUBDIR)\n",
    "make_dir(out_dir)\n",
    "for filename, new_system in new_systems:\n",
    "    new_system.to(fmt=\"cif\", filename=os.path.join(out_dir, filename))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e49ce79",
   "metadata": {},
   "source": [
    "Save properties"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "id": "3a403787",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>unitcell_repetition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 unitcell_repetition\n",
       "9.0-0-LiF_Fm-3m_-3.18_2x2x3.cif                   12\n",
       "9.0-1-LiF_P6_3mc_-3.17_2x2x3.cif                  12\n",
       "9.0-2-LiF_Pm-3m_-2.89_2x2x3.cif                   12"
      ]
     },
     "execution_count": 219,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_name = list(supercell_info.keys())\n",
    "info_df = pd.DataFrame(index=file_name, columns=[\"unitcell_repetition\"])\n",
    "for file_name_ in file_name:\n",
    "    info_df.loc[file_name_, \"unitcell_repetition\"] = supercell_info[file_name_]\n",
    "info_df.to_json(os.path.join(out_dir, \"info.json\"))\n",
    "info_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "id": "60d3f8c8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f38551247a8b4dc3953372dcbcd39f4c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(NGLWidget(max_frame=2), VBox(children=(Dropdown(description='Show', options=('All', 'F', 'Li'),…"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view(new_systems_ase, viewer=\"ngl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8300ed04",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.12"
  },
  "toc-autonumbering": true,
  "vscode": {
   "interpreter": {
    "hash": "6b8d2ee35f28cb0939199389d8f28e1089d9032c18d891ecd9d96baa0be783b2"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}