others/Performance on the test set with large rotations.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "f31b7a6f",
   "metadata": {},
   "source": [
    "# Performance of SigmaCCS on the test set with different coordinates generated by ETKDG and MMFF94"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "93cee49a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append(\"..\")\n",
    "from sigma.GraphData import *\n",
    "from sigma.model import *\n",
    "import pandas as pd\n",
    "import random\n",
    "from pandas import Series,DataFrame\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "import statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2ce40f9c",
   "metadata": {},
   "outputs": [],
   "source": [
    "ifile = '../data/TestData.csv'\n",
    "ParameterPath = '../parameter/parameter.pkl'\n",
    "mfileh5 = '../model/model_new.h5'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f15423f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "res = random.sample(range(1, 100), 30)\n",
    "def Generating_coordinates_different_seeds(smiles, adduct, ccs, a, All_Atoms, ps = AllChem.ETKDGv3(),):\n",
    "    '''\n",
    "    * Using ETKDG to generate 3D coordinates of molecules\n",
    "    *\n",
    "    * Attributes\n",
    "    * ----------\n",
    "    * smiles    : The SMILES string of the molecule\n",
    "    * adduct    : Adduct of molecules\n",
    "    * ccs       : CCS of molecules\n",
    "    * All_Atoms : Element set (The type of element provided must cover all elements contained in the molecule)\n",
    "    * ps        : ETKDG algorithm provided by RDkit\n",
    "    *\n",
    "    * Returns\n",
    "    * -------\n",
    "    * succ_smiles : SMILES of The molecules with 3D conformation can be successfully generated\n",
    "    * succ_adduct : Adduct of The molecules with 3D conformation can be successfully generated\n",
    "    * succ_ccs    : CCS of The molecules with 3D conformation can be successfully generated\n",
    "    * Coordinate  : 3D coordinates of molecules\n",
    "    '''\n",
    "    succ_smiles = []\n",
    "    succ_adduct = []\n",
    "    succ_ccs    = []\n",
    "    Coordinate  = []\n",
    "    \n",
    "    INDEX = -1\n",
    "    for smi in smiles:\n",
    "        INDEX += 1\n",
    "        try:\n",
    "            iMol = Chem.MolFromSmiles(smi)\n",
    "            iMol = Chem.RemoveHs(iMol)\n",
    "        except:\n",
    "            continue;\n",
    "        atoms = [atom.GetSymbol() for atom in iMol.GetAtoms()]\n",
    "        bonds = [bond for bond in iMol.GetBonds()]\n",
    "        # Is the number of atoms greater than 1\n",
    "        if len(atoms) == 1 and len(bonds) <= 1:\n",
    "            continue;\n",
    "        # Determine whether the element is in all_ In atoms\n",
    "        Elements_not_included = 0\n",
    "        for atom in atoms:\n",
    "            if atom not in All_Atoms:\n",
    "                Elements_not_included = 1\n",
    "        if Elements_not_included == 1:\n",
    "            continue;\n",
    "        # Adding H to a molecular object\n",
    "        iMol3D = Chem.AddHs(iMol)\n",
    "        \n",
    "        # The 3D conformation of the generating molecule\n",
    "        \n",
    "        ps.randomSeed = a   #!!!!!!!!\n",
    "        ps.maxAttempts = 1\n",
    "        ps.numThreads = 0\n",
    "        ps.useRandomCoords = True\n",
    "        re = AllChem.EmbedMultipleConfs(iMol3D, numConfs = 1, params = ps)\n",
    "        # Whether the conformation is successful or not\n",
    "        if len(re) == 0:\n",
    "            print('conformation is error')\n",
    "            continue;\n",
    "        # MMFF94\n",
    "        re = AllChem.MMFFOptimizeMoleculeConfs(iMol3D,  numThreads = 0)\n",
    "\n",
    "        This_mol_Coordinate = []\n",
    "        for atom in iMol3D.GetAtoms():\n",
    "            Coord = list(iMol3D.GetConformer().GetAtomPosition(atom.GetIdx()))\n",
    "            This_mol_Coordinate.append(Coord)\n",
    "        Coordinate.append(This_mol_Coordinate)\n",
    "        \n",
    "        succ_smiles.append(smi)\n",
    "        succ_adduct.append(adduct[INDEX])\n",
    "        succ_ccs.append(ccs[INDEX])\n",
    "            \n",
    "    return succ_smiles, succ_adduct, succ_ccs,  Coordinate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "47c56d2d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:27<00:00, 20.37it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 45\n",
      "R2 Score : 0.9937630410609587\n",
      "Median Relative Error : 1.2043710031130037 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:26<00:00, 20.76it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 34\n",
      "R2 Score : 0.9938415869677824\n",
      "Median Relative Error : 1.2026298794199095 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:17<00:00, 31.85it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 43\n",
      "R2 Score : 0.9937822358466177\n",
      "Median Relative Error : 1.2783695420865757 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 23.18it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 80\n",
      "R2 Score : 0.9936336531518316\n",
      "Median Relative Error : 1.231228652878192 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 23.75it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 68\n",
      "R2 Score : 0.9936368929409544\n",
      "Median Relative Error : 1.2201551784600997 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 23.44it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 92\n",
      "R2 Score : 0.9938535524303723\n",
      "Median Relative Error : 1.18266739425306 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:22<00:00, 25.33it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 88\n",
      "R2 Score : 0.9937978154275091\n",
      "Median Relative Error : 1.2124823867651036 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 23.68it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 41\n",
      "R2 Score : 0.9937839445751436\n",
      "Median Relative Error : 1.2473667834330997 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 23.98it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 86\n",
      "R2 Score : 0.9936868037264808\n",
      "Median Relative Error : 1.2224954272073412 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 23.00it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 98\n",
      "R2 Score : 0.9937185596444922\n",
      "Median Relative Error : 1.2547213135645647 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:17<00:00, 32.23it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 87\n",
      "R2 Score : 0.9936724794038739\n",
      "Median Relative Error : 1.2183028382140344 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:25<00:00, 22.20it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 82\n",
      "R2 Score : 0.9937338418070045\n",
      "Median Relative Error : 1.2724800411095738 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 23.13it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 21\n",
      "R2 Score : 0.993776196783835\n",
      "Median Relative Error : 1.2548317472298371 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 23.43it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 83\n",
      "R2 Score : 0.993827077655933\n",
      "Median Relative Error : 1.2295660324829787 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:17<00:00, 31.49it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 31\n",
      "R2 Score : 0.9939187213489945\n",
      "Median Relative Error : 1.206380066555967 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:26<00:00, 20.76it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 66\n",
      "R2 Score : 0.9937850018976171\n",
      "Median Relative Error : 1.2385847406944615 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:25<00:00, 22.29it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 65\n",
      "R2 Score : 0.9937679194406991\n",
      "Median Relative Error : 1.239196193275269 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:25<00:00, 21.49it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 90\n",
      "R2 Score : 0.9936766507091022\n",
      "Median Relative Error : 1.2592734691079204 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 23.58it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 60\n",
      "R2 Score : 0.9937601156479846\n",
      "Median Relative Error : 1.2455721079192623 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 24.08it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 32\n",
      "R2 Score : 0.9936606459112262\n",
      "Median Relative Error : 1.2222484323981095 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 23.05it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 67\n",
      "R2 Score : 0.9937357415420917\n",
      "Median Relative Error : 1.2411803751089934 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:17<00:00, 32.64it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 49\n",
      "R2 Score : 0.9938092300936576\n",
      "Median Relative Error : 1.2320123638785607 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:23<00:00, 23.57it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 46\n",
      "R2 Score : 0.9937681117449658\n",
      "Median Relative Error : 1.2091947637048417 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 23.02it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 13\n",
      "R2 Score : 0.9938110605411836\n",
      "Median Relative Error : 1.2039685732496903 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 22.35it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 36\n",
      "R2 Score : 0.9937978573201155\n",
      "Median Relative Error : 1.1657171031450995 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:17<00:00, 31.31it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 69\n",
      "R2 Score : 0.9936400851120422\n",
      "Median Relative Error : 1.2538308028427638 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 22.51it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 52\n",
      "R2 Score : 0.9938545988573427\n",
      "Median Relative Error : 1.2199781385712083 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:24<00:00, 22.92it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 37\n",
      "R2 Score : 0.9938509774176151\n",
      "Median Relative Error : 1.1873798209848483 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:10<00:00, 54.86it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 35\n",
      "R2 Score : 0.9938795558837294\n",
      "Median Relative Error : 1.2301139283978768 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 558/559 [00:09<00:00, 55.86it/s]\n",
      "The number of molecules in the test set : 559 \n",
      "\n",
      "randomSeed : 91\n",
      "R2 Score : 0.9938147113891018\n",
      "Median Relative Error : 1.1872586468987925 %\n"
     ]
    }
   ],
   "source": [
    "R2 = []\n",
    "MeRE = []\n",
    "randomSeeds = []\n",
    "\n",
    "for ii in range(30):\n",
    "    smiles, adduct, ccs = read_data(ifile)\n",
    "    param = parameter.Parameter()\n",
    "    with open(ParameterPath,'rb') as file:\n",
    "        param  = pickle.loads(file.read())  \n",
    "    a = res[ii]\n",
    "    smiles, adduct, ccs, Coordinate = Generating_coordinates_different_seeds(smiles, adduct, ccs, a, param.All_Atoms)   ###\n",
    "    \n",
    "    for i in range(len(Coordinate)):\n",
    "        Coordinate[i] = (np.array(Coordinate[i]) - param.Min_Coor) / (param.Max_Coor - param.Min_Coor)\n",
    "    \n",
    "    adj, features, edge_features = convertToGraph(smiles, Coordinate, param.All_Atoms)\n",
    "    DataSet = MyDataset(features, adj, edge_features, ccs)\n",
    "    \n",
    "    ECC_Model = load_Model_from_file(mfileh5)\n",
    "        \n",
    "    re = predict(ECC_Model,param.adduct_SET,DataSet,adduct,)\n",
    "    print(\"The number of molecules in the test set :\", len(re), '\\n')\n",
    "\n",
    "    Sigma   = re\n",
    "    CCS     = ccs\n",
    "    SigmaPer = Metrics(CCS, Sigma)\n",
    "    print(\"randomSeed :\", a)\n",
    "    print(\"R2 Score :\", SigmaPer[0])\n",
    "    print(\"Median Relative Error :\", SigmaPer[1], '%')\n",
    "    \n",
    "    R2.append(SigmaPer[0])\n",
    "    MeRE.append(SigmaPer[1])\n",
    "    randomSeeds.append(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2f71441f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "the mean value of R2 : 0.9938\n",
      "the standard deviation of R2 : 8e-05\n",
      "the mean value of Median Relative Error : 1.226 %\n",
      "the standard deviation of Median Relative Error : 0.0269 %\n"
     ]
    }
   ],
   "source": [
    "r2_mean = statistics.mean(R2)\n",
    "r2_std = statistics.stdev(R2)\n",
    "\n",
    "MeRE_mean = statistics.mean(MeRE)\n",
    "MeRE_std = statistics.stdev(MeRE)\n",
    "\n",
    "print(\"the mean value of R2 :\", round(r2_mean, 4))\n",
    "print(\"the standard deviation of R2 :\", round(r2_std, 5))\n",
    "\n",
    "print(\"the mean value of Median Relative Error :\", round(MeRE_mean, 3), '%')\n",
    "print(\"the standard deviation of Median Relative Error :\", round(MeRE_std, 4), '%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "251fab97",
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "randomSeed : 45\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.204 %\n",
      "randomSeed : 34\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.203 %\n",
      "randomSeed : 43\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.278 %\n",
      "randomSeed : 80\n",
      "R2 Score : 0.9936\n",
      "Median Relative Error : 1.231 %\n",
      "randomSeed : 68\n",
      "R2 Score : 0.9936\n",
      "Median Relative Error : 1.22 %\n",
      "randomSeed : 92\n",
      "R2 Score : 0.9939\n",
      "Median Relative Error : 1.183 %\n",
      "randomSeed : 88\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.212 %\n",
      "randomSeed : 41\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.247 %\n",
      "randomSeed : 86\n",
      "R2 Score : 0.9937\n",
      "Median Relative Error : 1.222 %\n",
      "randomSeed : 98\n",
      "R2 Score : 0.9937\n",
      "Median Relative Error : 1.255 %\n",
      "randomSeed : 87\n",
      "R2 Score : 0.9937\n",
      "Median Relative Error : 1.218 %\n",
      "randomSeed : 82\n",
      "R2 Score : 0.9937\n",
      "Median Relative Error : 1.272 %\n",
      "randomSeed : 21\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.255 %\n",
      "randomSeed : 83\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.23 %\n",
      "randomSeed : 31\n",
      "R2 Score : 0.9939\n",
      "Median Relative Error : 1.206 %\n",
      "randomSeed : 66\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.239 %\n",
      "randomSeed : 65\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.239 %\n",
      "randomSeed : 90\n",
      "R2 Score : 0.9937\n",
      "Median Relative Error : 1.259 %\n",
      "randomSeed : 60\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.246 %\n",
      "randomSeed : 32\n",
      "R2 Score : 0.9937\n",
      "Median Relative Error : 1.222 %\n",
      "randomSeed : 67\n",
      "R2 Score : 0.9937\n",
      "Median Relative Error : 1.241 %\n",
      "randomSeed : 49\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.232 %\n",
      "randomSeed : 46\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.209 %\n",
      "randomSeed : 13\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.204 %\n",
      "randomSeed : 36\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.166 %\n",
      "randomSeed : 69\n",
      "R2 Score : 0.9936\n",
      "Median Relative Error : 1.254 %\n",
      "randomSeed : 52\n",
      "R2 Score : 0.9939\n",
      "Median Relative Error : 1.22 %\n",
      "randomSeed : 37\n",
      "R2 Score : 0.9939\n",
      "Median Relative Error : 1.187 %\n",
      "randomSeed : 35\n",
      "R2 Score : 0.9939\n",
      "Median Relative Error : 1.23 %\n",
      "randomSeed : 91\n",
      "R2 Score : 0.9938\n",
      "Median Relative Error : 1.187 %\n"
     ]
    }
   ],
   "source": [
    "for j in range(len(randomSeeds)):\n",
    "    seed = randomSeeds[j]\n",
    "    r = R2[j]\n",
    "    re = MeRE[j]\n",
    "    print(\"randomSeed :\", seed)\n",
    "    print(\"R2 Score :\", round(r, 4))\n",
    "    print(\"Median Relative Error :\", round(re, 3), '%')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73572b4e",
   "metadata": {},
   "source": [
    "# Performance of SigmaCCS on the external test set with different coordinates generated by ETKDG and MMFF94"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "0ba1ce96",
   "metadata": {},
   "outputs": [],
   "source": [
    "ifile = '../data/ExternalTestData.csv'\n",
    "ParameterPath = '../parameter/parameter.pkl'\n",
    "mfileh5 = '../model/model_new.h5'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "405d7b58",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 62.32it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 97\n",
      "R2 Score : 0.9792576132103041\n",
      "Median Relative Error : 1.871191665827953 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 65.45it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 22\n",
      "R2 Score : 0.9793375253036047\n",
      "Median Relative Error : 1.8811994244720718 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 65.14it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 45\n",
      "R2 Score : 0.9798463606673514\n",
      "Median Relative Error : 1.8922467640096974 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.00it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 91\n",
      "R2 Score : 0.9795149057932717\n",
      "Median Relative Error : 1.891354330104663 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.39it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 78\n",
      "R2 Score : 0.9798731884971258\n",
      "Median Relative Error : 1.8710280832061295 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.57it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 80\n",
      "R2 Score : 0.9796024313458452\n",
      "Median Relative Error : 1.954111069712016 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.37it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 93\n",
      "R2 Score : 0.9797033567297349\n",
      "Median Relative Error : 1.953890904108012 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.06it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 3\n",
      "R2 Score : 0.9798650384961817\n",
      "Median Relative Error : 1.8781685159614745 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.06it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 50\n",
      "R2 Score : 0.9792594361555591\n",
      "Median Relative Error : 2.00598148678338 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.07it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 28\n",
      "R2 Score : 0.979158296432702\n",
      "Median Relative Error : 1.9685103195683284 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 62.10it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 76\n",
      "R2 Score : 0.9802473035458075\n",
      "Median Relative Error : 1.8707653587724158 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 62.59it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 6\n",
      "R2 Score : 0.9797443899891197\n",
      "Median Relative Error : 1.9761153965484195 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.64it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 15\n",
      "R2 Score : 0.9792707291134716\n",
      "Median Relative Error : 2.007216772223062 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 62.82it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 51\n",
      "R2 Score : 0.9796338821313076\n",
      "Median Relative Error : 1.9392233003012453 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 62.89it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 66\n",
      "R2 Score : 0.9797011101616764\n",
      "Median Relative Error : 1.9284416454511164 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.92it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 79\n",
      "R2 Score : 0.9801762051108515\n",
      "Median Relative Error : 1.900089657848237 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 65.04it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 12\n",
      "R2 Score : 0.9799317991273397\n",
      "Median Relative Error : 1.9175556064946535 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 65.32it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 49\n",
      "R2 Score : 0.9798158159939795\n",
      "Median Relative Error : 1.9511861130801142 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 65.88it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 2\n",
      "R2 Score : 0.9794762637337676\n",
      "Median Relative Error : 1.9043251765587683 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 66.15it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 21\n",
      "R2 Score : 0.9795188216508944\n",
      "Median Relative Error : 1.866909841923007 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.21it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 33\n",
      "R2 Score : 0.9794073082441926\n",
      "Median Relative Error : 1.9309082635004235 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.87it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 16\n",
      "R2 Score : 0.9795971815747104\n",
      "Median Relative Error : 1.8672512268421717 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 62.92it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 7\n",
      "R2 Score : 0.9800257435596267\n",
      "Median Relative Error : 1.9620538572171269 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.80it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 69\n",
      "R2 Score : 0.9793057629911172\n",
      "Median Relative Error : 1.9798182953168395 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.59it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 14\n",
      "R2 Score : 0.979975310607948\n",
      "Median Relative Error : 1.9945601826554902 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 63.20it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 60\n",
      "R2 Score : 0.9795083090177947\n",
      "Median Relative Error : 1.9292675343913497 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 62.54it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 87\n",
      "R2 Score : 0.9795895438255904\n",
      "Median Relative Error : 1.9651177686386447 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.42it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 13\n",
      "R2 Score : 0.9797374382133336\n",
      "Median Relative Error : 1.9189949113280895 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 64.69it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 65\n",
      "R2 Score : 0.9795750322466021\n",
      "Median Relative Error : 1.9842256685621396 %\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "..\\sigma\\GraphData.py:187: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  features = np.asarray(features)\n",
      "..\\sigma\\GraphData.py:188: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
      "  edge_features = np.asarray(edge_features)\n",
      "C:\\Users\\yxliao\\anaconda3\\envs\\sig\\lib\\site-packages\\keras\\initializers\\initializers_v2.py:121: UserWarning: The initializer GlorotUniform is unseeded and being called multiple times, which will return identical values  each time (even if the initializer is unseeded). Please update your code to provide a seed to the initializer, or avoid using the same initalizer instance more than once.\n",
      "  f\"The initializer {self.__class__.__name__} is unseeded \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions: 100%|█████████████████████████████████████▉| 343/344 [00:05<00:00, 65.08it/s]\n",
      "The number of molecules in the test set : 344 \n",
      "\n",
      "randomSeed : 62\n",
      "R2 Score : 0.9793332338379265\n",
      "Median Relative Error : 1.950404333516613 %\n"
     ]
    }
   ],
   "source": [
    "res = random.sample(range(1, 100), 30)\n",
    "R2 = []\n",
    "MeRE = []\n",
    "randomSeeds = []\n",
    "\n",
    "for ii in range(30):\n",
    "    smiles, adduct, ccs = read_data(ifile)\n",
    "    param = parameter.Parameter()\n",
    "    with open(ParameterPath,'rb') as file:\n",
    "        param  = pickle.loads(file.read())  \n",
    "    a = res[ii]\n",
    "    smiles, adduct, ccs, Coordinate = Generating_coordinates_different_seeds(smiles, adduct, ccs, a, param.All_Atoms)   ###\n",
    "    \n",
    "    for i in range(len(Coordinate)):\n",
    "        Coordinate[i] = (np.array(Coordinate[i]) - param.Min_Coor) / (param.Max_Coor - param.Min_Coor)\n",
    "    \n",
    "    adj, features, edge_features = convertToGraph(smiles, Coordinate, param.All_Atoms)\n",
    "    DataSet = MyDataset(features, adj, edge_features, ccs)\n",
    "    \n",
    "    ECC_Model = load_Model_from_file(mfileh5)\n",
    "        \n",
    "    re = predict(ECC_Model,param.adduct_SET,DataSet,adduct,)\n",
    "    print(\"The number of molecules in the test set :\", len(re), '\\n')\n",
    "\n",
    "    Sigma   = re\n",
    "    CCS     = ccs\n",
    "    SigmaPer = Metrics(CCS, Sigma)\n",
    "    print(\"randomSeed :\", a)\n",
    "    print(\"R2 Score :\", SigmaPer[0])\n",
    "    print(\"Median Relative Error :\", SigmaPer[1], '%')\n",
    "    \n",
    "    R2.append(SigmaPer[0])\n",
    "    MeRE.append(SigmaPer[1])\n",
    "    randomSeeds.append(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ddea9f86",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "the mean value of R2 : 0.9796\n",
      "the standard deviation of R2 : 0.00028\n",
      "the mean value of Median Relative Error : 1.93 %\n",
      "the standard deviation of Median Relative Error : 0.0445 %\n"
     ]
    }
   ],
   "source": [
    "r2_mean = statistics.mean(R2)\n",
    "r2_std = statistics.stdev(R2)\n",
    "\n",
    "MeRE_mean = statistics.mean(MeRE)\n",
    "MeRE_std = statistics.stdev(MeRE)\n",
    "\n",
    "print(\"the mean value of R2 :\", round(r2_mean, 4))\n",
    "print(\"the standard deviation of R2 :\", round(r2_std, 5))\n",
    "\n",
    "print(\"the mean value of Median Relative Error :\", round(MeRE_mean, 3), '%')\n",
    "print(\"the standard deviation of Median Relative Error :\", round(MeRE_std, 4), '%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d0aa98ba",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "randomSeed : 97\n",
      "R2 Score : 0.9793\n",
      "Median Relative Error : 1.871 % \n",
      "\n",
      "randomSeed : 22\n",
      "R2 Score : 0.9793\n",
      "Median Relative Error : 1.881 % \n",
      "\n",
      "randomSeed : 45\n",
      "R2 Score : 0.9798\n",
      "Median Relative Error : 1.892 % \n",
      "\n",
      "randomSeed : 91\n",
      "R2 Score : 0.9795\n",
      "Median Relative Error : 1.891 % \n",
      "\n",
      "randomSeed : 78\n",
      "R2 Score : 0.9799\n",
      "Median Relative Error : 1.871 % \n",
      "\n",
      "randomSeed : 80\n",
      "R2 Score : 0.9796\n",
      "Median Relative Error : 1.954 % \n",
      "\n",
      "randomSeed : 93\n",
      "R2 Score : 0.9797\n",
      "Median Relative Error : 1.954 % \n",
      "\n",
      "randomSeed : 3\n",
      "R2 Score : 0.9799\n",
      "Median Relative Error : 1.878 % \n",
      "\n",
      "randomSeed : 50\n",
      "R2 Score : 0.9793\n",
      "Median Relative Error : 2.006 % \n",
      "\n",
      "randomSeed : 28\n",
      "R2 Score : 0.9792\n",
      "Median Relative Error : 1.969 % \n",
      "\n",
      "randomSeed : 76\n",
      "R2 Score : 0.9802\n",
      "Median Relative Error : 1.871 % \n",
      "\n",
      "randomSeed : 6\n",
      "R2 Score : 0.9797\n",
      "Median Relative Error : 1.976 % \n",
      "\n",
      "randomSeed : 15\n",
      "R2 Score : 0.9793\n",
      "Median Relative Error : 2.007 % \n",
      "\n",
      "randomSeed : 51\n",
      "R2 Score : 0.9796\n",
      "Median Relative Error : 1.939 % \n",
      "\n",
      "randomSeed : 66\n",
      "R2 Score : 0.9797\n",
      "Median Relative Error : 1.928 % \n",
      "\n",
      "randomSeed : 79\n",
      "R2 Score : 0.9802\n",
      "Median Relative Error : 1.9 % \n",
      "\n",
      "randomSeed : 12\n",
      "R2 Score : 0.9799\n",
      "Median Relative Error : 1.918 % \n",
      "\n",
      "randomSeed : 49\n",
      "R2 Score : 0.9798\n",
      "Median Relative Error : 1.951 % \n",
      "\n",
      "randomSeed : 2\n",
      "R2 Score : 0.9795\n",
      "Median Relative Error : 1.904 % \n",
      "\n",
      "randomSeed : 21\n",
      "R2 Score : 0.9795\n",
      "Median Relative Error : 1.867 % \n",
      "\n",
      "randomSeed : 33\n",
      "R2 Score : 0.9794\n",
      "Median Relative Error : 1.931 % \n",
      "\n",
      "randomSeed : 16\n",
      "R2 Score : 0.9796\n",
      "Median Relative Error : 1.867 % \n",
      "\n",
      "randomSeed : 7\n",
      "R2 Score : 0.98\n",
      "Median Relative Error : 1.962 % \n",
      "\n",
      "randomSeed : 69\n",
      "R2 Score : 0.9793\n",
      "Median Relative Error : 1.98 % \n",
      "\n",
      "randomSeed : 14\n",
      "R2 Score : 0.98\n",
      "Median Relative Error : 1.995 % \n",
      "\n",
      "randomSeed : 60\n",
      "R2 Score : 0.9795\n",
      "Median Relative Error : 1.929 % \n",
      "\n",
      "randomSeed : 87\n",
      "R2 Score : 0.9796\n",
      "Median Relative Error : 1.965 % \n",
      "\n",
      "randomSeed : 13\n",
      "R2 Score : 0.9797\n",
      "Median Relative Error : 1.919 % \n",
      "\n",
      "randomSeed : 65\n",
      "R2 Score : 0.9796\n",
      "Median Relative Error : 1.984 % \n",
      "\n",
      "randomSeed : 62\n",
      "R2 Score : 0.9793\n",
      "Median Relative Error : 1.95 % \n",
      "\n"
     ]
    }
   ],
   "source": [
    "for j in range(len(randomSeeds)):\n",
    "    seed = randomSeeds[j]\n",
    "    r = R2[j]\n",
    "    re = MeRE[j]\n",
    "    print(\"randomSeed :\", seed)\n",
    "    print(\"R2 Score :\", round(r, 4))\n",
    "    print(\"Median Relative Error :\", round(re, 3), '%', '\\n')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}