test_dataset.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Library"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/envs/moleculegraph38/lib/python3.8/site-packages/openbabel/__init__.py:14: UserWarning: \"import openbabel\" is deprecated, instead use \"from openbabel import openbabel\"\n",
      "  warnings.warn('\"import openbabel\" is deprecated, instead use \"from openbabel import openbabel\"')\n"
     ]
    }
   ],
   "source": [
    "# My library\n",
    "from molgraph.dataset import *\n",
    "from molgraph.experiment import *\n",
    "# General library\n",
    "import argparse\n",
    "import os as os\n",
    "import numpy as np\n",
    "import pickle as pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Molecular Graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "reducedgraph = ['atom']\n",
    "reducedgraph = ['atom', 'pharmacophore', 'junctiontree', 'functional']\n",
    "\n",
    "parser = ArgumentParser()\n",
    "args = parser.getArgument('''\n",
    "--file bbbp\n",
    "--model GIN\n",
    "--schema A\n",
    "--reduced \n",
    "--mol_embedding 256\n",
    "--batch_normalize\n",
    "--fold 5\n",
    "--seed 42\n",
    "'''.split())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "file = args.file\n",
    "smiles = args.smiles\n",
    "task = args.task\n",
    "splitting = args.splitting\n",
    "splitting_fold = args.fold\n",
    "splitting_seed = args.splitting_seed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Function: getDataset()\n",
      "number of valid smiles: 2019\n"
     ]
    }
   ],
   "source": [
    "datasets = getDataset(file, smiles, task, splitting)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "scaffold ScaffoldSplitter\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:05] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:06] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:07] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n",
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Function: generateDatasetSplitting()\n",
      "Fold: 5\n",
      "Fold Number: 0\n",
      "-- Datasets Train: 1292\n",
      "-- Datasets Val: 323\n",
      "-- Datasets Test:  404\n",
      "-- Total: 2019\n",
      "Fold Number: 1\n",
      "-- Datasets Train: 1292\n",
      "-- Datasets Val: 323\n",
      "-- Datasets Test:  404\n",
      "-- Total: 2019\n",
      "Fold Number: 2\n",
      "-- Datasets Train: 1292\n",
      "-- Datasets Val: 323\n",
      "-- Datasets Test:  404\n",
      "-- Total: 2019\n",
      "Fold Number: 3\n",
      "-- Datasets Train: 1292\n",
      "-- Datasets Val: 323\n",
      "-- Datasets Test:  404\n",
      "-- Total: 2019\n",
      "Fold Number: 4\n",
      "-- Datasets Train: 1292\n",
      "-- Datasets Val: 323\n",
      "-- Datasets Test:  404\n",
      "-- Total: 2019\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[12:28:08] WARNING: not removing hydrogen atom without neighbors\n"
     ]
    }
   ],
   "source": [
    "datasets_splitted = generateDatasetSplitting(file, splitting, splitting_fold, splitting_seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Function: generateGraphDataset()\n",
      "Datasets graph:  1965\n"
     ]
    }
   ],
   "source": [
    "datasets_graph = generateGraphDataset(file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# generateVocabTrain(file, splitting_fold, vocab_len=len(datasets_graph)//10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== atom ===\n",
      "=== pharmacophore ===\n",
      "=== junctiontree ===\n",
      "=== functional ===\n"
     ]
    }
   ],
   "source": [
    "dict_reducedgraph_all = dict()\n",
    "for r in reducedgraph:\n",
    "    print('===',r,'===')\n",
    "    dict_reducedgraph = generateReducedGraphDict(file, r, vocab_file=file+'_0')\n",
    "    dict_reducedgraph_all[r] = dict_reducedgraph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "atom_list = list()\n",
    "for r in dict_reducedgraph_all:\n",
    "    substructure_atom_list = list()\n",
    "    for d in dict_reducedgraph_all[r]:\n",
    "        substructure_atom_list.append(dict_reducedgraph_all[r][d][0].x_r.shape[0])\n",
    "    atom_list.append(substructure_atom_list)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "atom\n",
      "- mean 24.223409669211197 10.665320919128412\n",
      "- max 132\n",
      "- min 2\n",
      "pharmacophore\n",
      "- mean 20.003562340966923 10.574273236743004\n",
      "- max 128\n",
      "- min 1\n",
      "junctiontree\n",
      "- mean 14.911450381679389 8.074985480931723\n",
      "- max 85\n",
      "- min 1\n",
      "functional\n",
      "- mean 11.914503816793893 6.252760678484563\n",
      "- max 64\n",
      "- min 1\n"
     ]
    }
   ],
   "source": [
    "max_node = 0\n",
    "min_node = 999\n",
    "for r, a in zip(reducedgraph, atom_list):\n",
    "    print(r)\n",
    "    print('- mean', np.mean(a), np.std(a))\n",
    "    print('- max', np.max(a))\n",
    "    print('- min', np.min(a))\n",
    "    max_node = np.max(a) if np.max(a) > max_node else max_node\n",
    "    min_node = np.min(a) if np.min(a) < min_node else min_node"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x1000 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams.update({'font.size': 12})\n",
    "\n",
    "colors = {'atom': '#104C96',\n",
    "          'junctiontree': '#C60505',\n",
    "          'cluster': '#E26AB0',\n",
    "          'functional': '#FD6B12',\n",
    "          'pharmacophore': '#088D36',\n",
    "          'substructure': '#8B50C9'}\n",
    "          #F4D53B yellow\n",
    "\n",
    "fig, ax = plt.subplots(4, 1, sharex=True, figsize=(15,10))\n",
    "for i, a in enumerate(atom_list):\n",
    "    sns.histplot(a, bins=range(1, 110, 1), ax=ax[i], kde=False, legend=False, discrete=True,\n",
    "                 label=reducedgraph[i], color=colors[reducedgraph[i]])\n",
    "    ax[i].axvline(x=np.mean(a), linewidth=1, color='black')\n",
    "    ax[i].set_ylabel('#compounds')\n",
    "    ax[i].set_xlim([0, np.ceil(max_node / 10) * 10])\n",
    "    ax[i].set_ylim([0, 800])\n",
    "    ax[i].legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 265\n",
    "fold_number = 0\n",
    "train_loader, val_loader, test_loader, datasets_train, datasets_val, datasets_test = generateDataLoader(file, batch_size, seed=42, fold_number=fold_number)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "PairDataBatch(smiles=[265], x_g=[6814, 79], x_g_batch=[6814], x_g_ptr=[266], edge_index_g=[2, 14816], edge_attr_g=[14816, 10], y=[265, 1])\n",
      "265\n",
      "1\n",
      "PairDataBatch(smiles=[265], x_g=[6582, 79], x_g_batch=[6582], x_g_ptr=[266], edge_index_g=[2, 14326], edge_attr_g=[14326, 10], y=[265, 1])\n",
      "265\n",
      "2\n",
      "PairDataBatch(smiles=[265], x_g=[6627, 79], x_g_batch=[6627], x_g_ptr=[266], edge_index_g=[2, 14356], edge_attr_g=[14356, 10], y=[265, 1])\n",
      "265\n",
      "3\n",
      "PairDataBatch(smiles=[265], x_g=[6480, 79], x_g_batch=[6480], x_g_ptr=[266], edge_index_g=[2, 14006], edge_attr_g=[14006, 10], y=[265, 1])\n",
      "265\n",
      "4\n",
      "PairDataBatch(smiles=[232], x_g=[5761, 79], x_g_batch=[5761], x_g_ptr=[233], edge_index_g=[2, 12484], edge_attr_g=[12484, 10], y=[232, 1])\n",
      "232\n"
     ]
    }
   ],
   "source": [
    "# visualize dataloader\n",
    "for _, item in enumerate(train_loader):\n",
    "    print(_)\n",
    "    print(item)\n",
    "    print(getNumberofSmiles(item))\n",
    "    # print(item.x_g)\n",
    "    # print(item.edge_index_g)\n",
    "    # print(item.x_g_batch)\n",
    "    # print(item.y)\n",
    "    # break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pos_weight: 0.3025806451612903\n"
     ]
    }
   ],
   "source": [
    "if args.graphtask == 'classification':\n",
    "    args.pos_weight = getPosWeight(datasets)\n",
    "    print('pos_weight:', args.pos_weight)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Positive y1 1500\n",
      "Negative y0 465\n"
     ]
    }
   ],
   "source": [
    "y0 = 0\n",
    "y1 = 0\n",
    "for g in datasets_graph:\n",
    "    # print(datasets_graph[g])\n",
    "    if datasets_graph[g].y == 0:\n",
    "        y0+=1\n",
    "    else:\n",
    "        y1+=1\n",
    "    # break\n",
    "\n",
    "print('Positive y1', y1)\n",
    "print('Negative y0', y0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.8.13 ('myenv38')",
   "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": "3.8.13"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "39c542c5c02b7a511e1ae79f0bb478cef0c12733a4ae7993722600208abb2b65"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}