{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ba904ebb-c42c-4469-872f-dbf0f8cef3c8",
   "metadata": {},
   "source": [
    "Making Predictions with GlassPy\n",
    "===============================\n",
    "\n",
    "**Author:** Daniel R. Cassar\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ccfd5b68-f4dd-45e4-a546-b82f25695a68",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f10f97ca-d23f-4d9a-8b60-72043f1c1764",
   "metadata": {},
   "source": [
    "GlassPy provides three prediction models: `GlassNet`, `VITRIFY`, and `ViscNet`.\n",
    "\n",
    "-   `GlassNet` is a multitask deep neural network that predicts 85 glass properties and the temperature dependence of viscosity.\n",
    "-   `VITRIFY` is a classification model that estimates the probability of a given composition forming a glass.\n",
    "-   `ViscNet` is a physics-informed deep neural network designed specifically to predict the temperature dependence of viscosity.\n",
    "\n",
    "All three models were trained on data from the SciGlass database.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "af4db9cb-1101-490b-87ae-434940df1459",
   "metadata": {},
   "source": [
    "## GlassNet basic usage\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "908d76f4-c680-4fb1-9dd8-ff5504a59dc2",
   "metadata": {},
   "source": [
    "Below is a minimal example of how to load and use the GlassNet model.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e7647470-9e5b-4a63-9113-3c5408301cb0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>T0</th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "      <th>T6</th>\n",
       "      <th>T7</th>\n",
       "      <th>T8</th>\n",
       "      <th>T9</th>\n",
       "      <th>...</th>\n",
       "      <th>Cp1673K</th>\n",
       "      <th>TMaxGrowthVelocity</th>\n",
       "      <th>MaxGrowthVelocity</th>\n",
       "      <th>CrystallizationPeak</th>\n",
       "      <th>CrystallizationOnset</th>\n",
       "      <th>SurfaceTensionAboveTg</th>\n",
       "      <th>SurfaceTension1173K</th>\n",
       "      <th>SurfaceTension1473K</th>\n",
       "      <th>SurfaceTension1573K</th>\n",
       "      <th>SurfaceTension1673K</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1612.586818</td>\n",
       "      <td>1286.966695</td>\n",
       "      <td>1187.265614</td>\n",
       "      <td>1070.224145</td>\n",
       "      <td>968.06795</td>\n",
       "      <td>901.811624</td>\n",
       "      <td>859.579252</td>\n",
       "      <td>825.204682</td>\n",
       "      <td>782.490883</td>\n",
       "      <td>759.993951</td>\n",
       "      <td>...</td>\n",
       "      <td>1717.653045</td>\n",
       "      <td>1040.284178</td>\n",
       "      <td>-6.241866</td>\n",
       "      <td>937.262437</td>\n",
       "      <td>788.40385</td>\n",
       "      <td>0.316586</td>\n",
       "      <td>0.3044</td>\n",
       "      <td>0.321539</td>\n",
       "      <td>0.318579</td>\n",
       "      <td>0.319566</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 85 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            T0           T1           T2           T3         T4          T5  \\\n",
       "0  1612.586818  1286.966695  1187.265614  1070.224145  968.06795  901.811624   \n",
       "\n",
       "           T6          T7          T8          T9  ...      Cp1673K  \\\n",
       "0  859.579252  825.204682  782.490883  759.993951  ...  1717.653045   \n",
       "\n",
       "   TMaxGrowthVelocity  MaxGrowthVelocity  CrystallizationPeak  \\\n",
       "0         1040.284178          -6.241866           937.262437   \n",
       "\n",
       "   CrystallizationOnset  SurfaceTensionAboveTg  SurfaceTension1173K  \\\n",
       "0             788.40385               0.316586               0.3044   \n",
       "\n",
       "   SurfaceTension1473K  SurfaceTension1573K  SurfaceTension1673K  \n",
       "0             0.321539             0.318579             0.319566  \n",
       "\n",
       "[1 rows x 85 columns]"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from glasspy.predict import GlassNet\n",
    "\n",
    "model = GlassNet()\n",
    "composition = \"Li2O(SiO2)2\"\n",
    "predictions = model.predict(composition)\n",
    "\n",
    "predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf662bde-e8ff-4a62-8a51-04e77a51872c",
   "metadata": {},
   "source": [
    "A composition can also be defined using a dictionary.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4c41d59c-1161-49fd-bbac-8da67fe05619",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>T0</th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "      <th>T6</th>\n",
       "      <th>T7</th>\n",
       "      <th>T8</th>\n",
       "      <th>T9</th>\n",
       "      <th>...</th>\n",
       "      <th>Cp1673K</th>\n",
       "      <th>TMaxGrowthVelocity</th>\n",
       "      <th>MaxGrowthVelocity</th>\n",
       "      <th>CrystallizationPeak</th>\n",
       "      <th>CrystallizationOnset</th>\n",
       "      <th>SurfaceTensionAboveTg</th>\n",
       "      <th>SurfaceTension1173K</th>\n",
       "      <th>SurfaceTension1473K</th>\n",
       "      <th>SurfaceTension1573K</th>\n",
       "      <th>SurfaceTension1673K</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1612.586818</td>\n",
       "      <td>1286.966695</td>\n",
       "      <td>1187.265614</td>\n",
       "      <td>1070.224145</td>\n",
       "      <td>968.06795</td>\n",
       "      <td>901.811624</td>\n",
       "      <td>859.579252</td>\n",
       "      <td>825.204682</td>\n",
       "      <td>782.490883</td>\n",
       "      <td>759.993951</td>\n",
       "      <td>...</td>\n",
       "      <td>1717.653045</td>\n",
       "      <td>1040.284178</td>\n",
       "      <td>-6.241866</td>\n",
       "      <td>937.262437</td>\n",
       "      <td>788.40385</td>\n",
       "      <td>0.316586</td>\n",
       "      <td>0.3044</td>\n",
       "      <td>0.321539</td>\n",
       "      <td>0.318579</td>\n",
       "      <td>0.319566</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 85 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            T0           T1           T2           T3         T4          T5  \\\n",
       "0  1612.586818  1286.966695  1187.265614  1070.224145  968.06795  901.811624   \n",
       "\n",
       "           T6          T7          T8          T9  ...      Cp1673K  \\\n",
       "0  859.579252  825.204682  782.490883  759.993951  ...  1717.653045   \n",
       "\n",
       "   TMaxGrowthVelocity  MaxGrowthVelocity  CrystallizationPeak  \\\n",
       "0         1040.284178          -6.241866           937.262437   \n",
       "\n",
       "   CrystallizationOnset  SurfaceTensionAboveTg  SurfaceTension1173K  \\\n",
       "0             788.40385               0.316586               0.3044   \n",
       "\n",
       "   SurfaceTension1473K  SurfaceTension1573K  SurfaceTension1673K  \n",
       "0             0.321539             0.318579             0.319566  \n",
       "\n",
       "[1 rows x 85 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "composition = {\n",
    "    \"SiO2\": 2,\n",
    "    \"Li2O\": 1,\n",
    "}\n",
    "predictions = model.predict(composition)\n",
    "\n",
    "predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ecc9e9d-214a-4fab-b7ff-78bba64f3004",
   "metadata": {},
   "source": [
    "GlassNet also accepts `pandas` DataFrames as input. Note that each row represents a material and that only columns related to chemical composition can be present in the DataFrame.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "49cb9218-6913-4dfb-9b3e-28e970b7bce0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Li2O</th>\n",
       "      <th>Na2O</th>\n",
       "      <th>SiO2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Li2O  Na2O  SiO2\n",
       "0     1     0     2\n",
       "1     0     1     2\n",
       "2     1     1     2"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "data = [\n",
    "    [1, 0, 2],\n",
    "    [0, 1, 2],\n",
    "    [1, 1, 2],\n",
    "]\n",
    "\n",
    "df = pd.DataFrame(data, columns=[\"Li2O\", \"Na2O\", \"SiO2\"])\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e0b62145-9078-4685-b371-e6f18f99dde6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>T0</th>\n",
       "      <th>T1</th>\n",
       "      <th>T2</th>\n",
       "      <th>T3</th>\n",
       "      <th>T4</th>\n",
       "      <th>T5</th>\n",
       "      <th>T6</th>\n",
       "      <th>T7</th>\n",
       "      <th>T8</th>\n",
       "      <th>T9</th>\n",
       "      <th>...</th>\n",
       "      <th>Cp1673K</th>\n",
       "      <th>TMaxGrowthVelocity</th>\n",
       "      <th>MaxGrowthVelocity</th>\n",
       "      <th>CrystallizationPeak</th>\n",
       "      <th>CrystallizationOnset</th>\n",
       "      <th>SurfaceTensionAboveTg</th>\n",
       "      <th>SurfaceTension1173K</th>\n",
       "      <th>SurfaceTension1473K</th>\n",
       "      <th>SurfaceTension1573K</th>\n",
       "      <th>SurfaceTension1673K</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1612.586818</td>\n",
       "      <td>1286.966759</td>\n",
       "      <td>1187.265614</td>\n",
       "      <td>1070.224201</td>\n",
       "      <td>968.067993</td>\n",
       "      <td>901.811569</td>\n",
       "      <td>859.579252</td>\n",
       "      <td>825.204682</td>\n",
       "      <td>782.490883</td>\n",
       "      <td>759.993951</td>\n",
       "      <td>...</td>\n",
       "      <td>1717.653120</td>\n",
       "      <td>1040.284178</td>\n",
       "      <td>-6.241865</td>\n",
       "      <td>937.262560</td>\n",
       "      <td>788.403850</td>\n",
       "      <td>0.316586</td>\n",
       "      <td>0.304400</td>\n",
       "      <td>0.321539</td>\n",
       "      <td>0.318579</td>\n",
       "      <td>0.319566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1700.171170</td>\n",
       "      <td>1521.813466</td>\n",
       "      <td>1287.576204</td>\n",
       "      <td>1130.354623</td>\n",
       "      <td>1036.780176</td>\n",
       "      <td>955.844950</td>\n",
       "      <td>911.027803</td>\n",
       "      <td>848.791035</td>\n",
       "      <td>810.406750</td>\n",
       "      <td>782.243953</td>\n",
       "      <td>...</td>\n",
       "      <td>1482.022135</td>\n",
       "      <td>1058.443194</td>\n",
       "      <td>-6.394961</td>\n",
       "      <td>963.434464</td>\n",
       "      <td>905.734703</td>\n",
       "      <td>0.279031</td>\n",
       "      <td>0.310337</td>\n",
       "      <td>0.290839</td>\n",
       "      <td>0.287974</td>\n",
       "      <td>0.272271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1457.627094</td>\n",
       "      <td>1246.016165</td>\n",
       "      <td>1114.455164</td>\n",
       "      <td>987.722016</td>\n",
       "      <td>934.740619</td>\n",
       "      <td>861.280868</td>\n",
       "      <td>794.972760</td>\n",
       "      <td>769.729834</td>\n",
       "      <td>730.195119</td>\n",
       "      <td>700.667308</td>\n",
       "      <td>...</td>\n",
       "      <td>1751.036502</td>\n",
       "      <td>1082.912718</td>\n",
       "      <td>-5.101020</td>\n",
       "      <td>875.732689</td>\n",
       "      <td>820.388535</td>\n",
       "      <td>0.299141</td>\n",
       "      <td>0.308964</td>\n",
       "      <td>0.319907</td>\n",
       "      <td>0.321466</td>\n",
       "      <td>0.294990</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 85 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            T0           T1           T2           T3           T4  \\\n",
       "0  1612.586818  1286.966759  1187.265614  1070.224201   968.067993   \n",
       "1  1700.171170  1521.813466  1287.576204  1130.354623  1036.780176   \n",
       "2  1457.627094  1246.016165  1114.455164   987.722016   934.740619   \n",
       "\n",
       "           T5          T6          T7          T8          T9  ...  \\\n",
       "0  901.811569  859.579252  825.204682  782.490883  759.993951  ...   \n",
       "1  955.844950  911.027803  848.791035  810.406750  782.243953  ...   \n",
       "2  861.280868  794.972760  769.729834  730.195119  700.667308  ...   \n",
       "\n",
       "       Cp1673K  TMaxGrowthVelocity  MaxGrowthVelocity  CrystallizationPeak  \\\n",
       "0  1717.653120         1040.284178          -6.241865           937.262560   \n",
       "1  1482.022135         1058.443194          -6.394961           963.434464   \n",
       "2  1751.036502         1082.912718          -5.101020           875.732689   \n",
       "\n",
       "   CrystallizationOnset  SurfaceTensionAboveTg  SurfaceTension1173K  \\\n",
       "0            788.403850               0.316586             0.304400   \n",
       "1            905.734703               0.279031             0.310337   \n",
       "2            820.388535               0.299141             0.308964   \n",
       "\n",
       "   SurfaceTension1473K  SurfaceTension1573K  SurfaceTension1673K  \n",
       "0             0.321539             0.318579             0.319566  \n",
       "1             0.290839             0.287974             0.272271  \n",
       "2             0.319907             0.321466             0.294990  \n",
       "\n",
       "[3 rows x 85 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions = model.predict(df)\n",
    "\n",
    "predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "93f9f6e2-5beb-4ac7-98c4-30f684bbf63b",
   "metadata": {},
   "source": [
    "GlassNet can also predict viscosity and the MYEGA viscosity equation parameters.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9021f4d9-47c6-4df5-9a57-81091a27d41c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3.53501871, 4.40572086, 2.83984945])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions = model.predict_log10_viscosity(\n",
    "    T=1000,\n",
    "    composition=df,\n",
    ")\n",
    "\n",
    "predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "13515495-1066-4765-912a-6b5774a408a1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>log10_eta_infinity (Pa.s)</th>\n",
       "      <th>Tg_MYEGA (K)</th>\n",
       "      <th>fragility</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.222603</td>\n",
       "      <td>701.790333</td>\n",
       "      <td>42.844319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.660096</td>\n",
       "      <td>710.677354</td>\n",
       "      <td>35.863589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.494931</td>\n",
       "      <td>647.762409</td>\n",
       "      <td>40.367431</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   log10_eta_infinity (Pa.s)  Tg_MYEGA (K)  fragility\n",
       "0                  -1.222603    701.790333  42.844319\n",
       "1                  -1.660096    710.677354  35.863589\n",
       "2                  -1.494931    647.762409  40.367431"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "viscosity_parameters = model.viscosity_parameters(df)\n",
    "\n",
    "viscosity_parameters"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bac99c66-8ae5-42c2-aab0-86f8747312e1",
   "metadata": {},
   "source": [
    "## VITRIFY base usage\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "374cf723-3818-4ebc-99c5-50a2597d91b8",
   "metadata": {},
   "source": [
    "Below is a minimal example of how to load and use the VITRIFY model. Like GlassNet, VITRIFY accepts strings, dictionaries, and `pandas` DataFrames as input. It returns `1` if the composition is predicted to form a glass under typical laboratory conditions, or `0` otherwise.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f2ed18c2-54d3-4afd-ab57-03f7346e3380",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 1, 1], dtype=int64)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from glasspy.predict import VITRIFY\n",
    "\n",
    "model = VITRIFY()\n",
    "predictions = model.predict(df)\n",
    "\n",
    "predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7635f095-4b3d-484d-bf77-8485f119c7d4",
   "metadata": {},
   "source": [
    "You can also retrieve the probability that the composition will form a glass, as estimated by the model.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "4fcadcb0-8502-4389-86a5-c36c44fdc3cb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.96919906, 0.99467885, 0.73446614])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "probabilities = model.predict_proba_glass(df)\n",
    "\n",
    "probabilities"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2134f33a-bb00-4875-96ac-db609d2d5be3",
   "metadata": {},
   "source": [
    "## ViscNet basic usage\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fc05e13-01d3-4e99-a167-8cd6e05cdb90",
   "metadata": {},
   "source": [
    "ViscNet is used similarly to GlassNet. However, since GlassNet outperforms ViscNet on viscosity prediction, using GlassNet is generally recommended for that task. The minimal example below demonstrates how to load and use ViscNet.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ead9eec3-c10e-489b-ba4c-c665d2be37fc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5.197888 , 5.6581345, 6.0721517], dtype=float32)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from glasspy.predict import ViscNet\n",
    "\n",
    "model = ViscNet()\n",
    "log10_viscosity = model.predict(T=1000, composition=df)\n",
    "\n",
    "log10_viscosity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "0954d492-b012-4052-9589-f1a4b2b7ad35",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[34.436737 29.760372 34.16933 ]\n"
     ]
    }
   ],
   "source": [
    "fragility = model.predict_fragility(df)\n",
    "\n",
    "print(fragility)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "13d3a8e7-443c-452b-8798-071b752ce0c3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[748.3823  739.4691  788.51385]\n"
     ]
    }
   ],
   "source": [
    "Tg = model.predict_Tg(df)\n",
    "\n",
    "print(Tg)"
   ]
  }
 ],
 "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": "3.14.3"
  },
  "org": null
 },
 "nbformat": 4,
 "nbformat_minor": 5
}