CVD_prediction/CVD_model_thickness.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Modelling and prediction of a CVD process\n",
    "\n",
    "## Problem\n",
    "- Model layer thickness resulting from a CVD process, based on set of DOE data\n",
    "- Thickness was measured at 49 sites across a wafer\n",
    "\n",
    "## Goals\n",
    "1. Use regression model to predict thickness for every single site\n",
    "2. Adapt model to also consider X and Y coordinates of the site to form a single model\n",
    "3. Implement model to include temperature as process parameter\n",
    "\n",
    "## Possible improvements\n",
    "- Hyperparameter tuning for ANN, improve stability of model quality, show learning rate\n",
    "- Encode TOOL into dummy variables instead of scalar\n",
    "- Encode X, Y coordinates into dummy variables / allow for prediction based on any X and Y coordinates, not only fixed sites"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Preprocessing\n",
    "### Import data\n",
    "- Import raw data from csv files\n",
    "- Set all independant variables as X as array\n",
    "- Set dependant variable Layer thickness as y\n",
    "\n",
    "### Split data & Feature scaling\n",
    "- Split into training and test data (75:25)\n",
    "- Fix seed for randomness\n",
    "- Scale all features using StandardScaler\n",
    "\n",
    "### Inlcude temperature in model\n",
    "- Thickness factors from literature:\n",
    "  - 500 °C = 0.94\n",
    "  - 480 °C = 1.0 (*default*)\n",
    "  - 460 °C = 1.04\n",
    "- Calculated scaling formula for CVD thickness d depending on temperature:\n",
    "  - d = d_0 * (-0.0025 * temperature + 2.1933)\n",
    "  - temperature_factor(480) = 1 (except rounding error)\n",
    "- Temperature is not needed as feature in training data since linear model is assumed  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from tkinter import *\n",
    "from tkinter.ttk import Combobox\n",
    "from matplotlib.figure import Figure\n",
    "from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg\n",
    "from numpy.random import seed\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.model_selection import cross_val_score, train_test_split\n",
    "from sklearn.preprocessing import StandardScaler, OneHotEncoder, PolynomialFeatures\n",
    "from keras.layers import Dense\n",
    "from keras.models import Sequential\n",
    "from keras.optimizers import RMSprop\n",
    "from tensorflow import random\n",
    "\n",
    "%matplotlib inline\n",
    "from matplotlib import pyplot as plt\n",
    "plt.rcParams['figure.figsize'] = [10, 8]\n",
    "\n",
    "seed(0)\n",
    "random.set_seed(0)\n",
    "\n",
    "DIR = './'\n",
    "SITES = 'site_coordinates.csv'\n",
    "DATA = 'test_assignment_sim.csv'\n",
    "\n",
    "raw_sites = pd.read_csv(DIR+SITES)\n",
    "raw_data = pd.read_csv(DIR+DATA)\n",
    "\n",
    "X = raw_data.iloc[:,0:4]\n",
    "y = raw_data.iloc[:,4:53]\n",
    "\n",
    "def preprocess(X, y):\n",
    "    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25, random_state = 0)\n",
    "    sc_X = StandardScaler()\n",
    "    X_train = sc_X.fit_transform(X_train)\n",
    "    X_test = sc_X.transform(X_test)\n",
    "    sc_y = StandardScaler()\n",
    "    y_train = sc_y.fit_transform(y_train)\n",
    "    return X_train, X_test, y_train, y_test, sc_X, sc_y\n",
    "\n",
    "def temperature_factor(temperature):\n",
    "    '''Returns CVD layer thickness scaling factor for a given temperature.'''\n",
    "    return -0.0025 * float(temperature) + 2.1933"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Train model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "r^2 score:  0.981680264282583\n"
     ]
    }
   ],
   "source": [
    "# 4 input nodes\n",
    "# 2 layers of 27 nodes\n",
    "# 49 output nodes\n",
    "\n",
    "X_train, X_test, y_train, y_test, sc_X, sc_y = preprocess(X, y)\n",
    "regressor = Sequential()\n",
    "regressor.add(Dense(units=27, activation='relu', input_dim=X_train.shape[1]))\n",
    "regressor.add(Dense(units=27, activation='relu'))\n",
    "regressor.add(Dense(units=49))\n",
    "regressor.compile(optimizer=RMSprop(0.001), loss='mse', metrics=['mae','mse'])\n",
    "regressor.fit(X_train, y_train, batch_size=10, epochs=200, verbose=0, use_multiprocessing=True)\n",
    "y_pred = sc_y.inverse_transform(regressor.predict(X_test))\n",
    "r2_ann_multi = r2_score(y_test, y_pred)\n",
    "print('r^2 score: ', r2_ann_multi);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Predictions\n",
    "- Defines prediction function\n",
    "- Creates GUI window to allow easy editing of process parameters or use prediction in notebook\n",
    "- Shows predicted layer thickness for all sites in window\n",
    "- Shows wafer map of thickness distribution in Jupyter notebook after GUI is closed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict(notebook=False):\n",
    "    if notebook:\n",
    "        flow = 0.92\n",
    "        spacing = 0.35\n",
    "        deptime = 70\n",
    "        temperature = 480\n",
    "        tool = 1\n",
    "    else:\n",
    "        flow = in_flow.get()\n",
    "        spacing = in_spacing.get()\n",
    "        deptime = in_deptime.get()\n",
    "        temperature = in_temperature.get()\n",
    "        tool = cb_tool.get()\n",
    "    \n",
    "    parameters = np.array([flow, spacing, deptime, tool]).reshape(1,-1)\n",
    "    parameters = sc_X.transform(parameters)\n",
    "    print('Process parameters:\\nFLOWFACTOR:\\t{}\\nSPACING:\\t{}\\nDEP TIME:\\t{} s\\nTOOL:\\t\\t#{}\\nTemperature:\\t{} °C\\n'.format(\n",
    "    flow, spacing, deptime, tool, temperature))\n",
    "    y_pred = sc_y.inverse_transform(regressor.predict(parameters)) * temperature_factor(temperature)\n",
    "    \n",
    "    if notebook:\n",
    "        print('Predicted SiO2 layer thickness for all sites (1 - 49):\\n\\n', y_pred)\n",
    "    else: \n",
    "        lbl_result = Label(window, text=\"Thickness prediction: \", font=(\"Helvetica\", 11))\n",
    "        lbl_result.place(x=pos_x_labels, y=y_tool+70)\n",
    "        result_box = Listbox(window, height=20)\n",
    "        for _ in range(0, len(y_pred[0])):\n",
    "            entry = 'Site ' + str(_) + ': ' + str(y_pred[0][_])\n",
    "            result_box.insert(_+1, entry)\n",
    "        result_box.place(x=pos_x_labels+3, y=y_tool+95)\n",
    "    \n",
    "    df_pred = pd.DataFrame()\n",
    "    df_pred['Prediction'] = y_pred.flatten()\n",
    "    df_pred['X'] = raw_sites.SITE_X.values.round(0)\n",
    "    df_pred['Y'] = raw_sites.SITE_Y.values.round(0)\n",
    "    fig, ax1 = plt.subplots()\n",
    "    fig = Figure(figsize = (9, 6), facecolor = \"white\")\n",
    "    ax1.set_title('Predicted SiO2 layer thickness')\n",
    "\n",
    "    df_pred.plot.scatter('X', 'Y', c='Prediction', colormap='Blues', ax=ax1)\n",
    "    return y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Process parameters:\n",
      "FLOWFACTOR:\t0.92\n",
      "SPACING:\t0.35\n",
      "DEP TIME:\t70 s\n",
      "TOOL:\t\t#1\n",
      "Temperature:\t480 °C\n",
      "\n",
      "Predicted SiO2 layer thickness for all sites (1 - 49):\n",
      "\n",
      " [[1566.6394 1516.5896 1569.493  1568.1709 1566.4678 1565.4136 1570.3058\n",
      "  1570.922  1570.2164 1622.5685 1622.0509 1621.2283 1620.6143 1620.0347\n",
      "  1619.7888 1618.7257 1619.5199 1618.6635 1622.9792 1623.565  1623.6831\n",
      "  1623.9048 1624.4808 1623.137  1623.0607 1599.2683 1598.7465 1598.635\n",
      "  1598.5309 1597.9188 1596.967  1596.46   1596.0399 1594.5374 1594.0562\n",
      "  1593.6328 1592.9905 1593.3906 1596.6774 1596.9202 1598.4688 1598.4441\n",
      "  1598.2537 1598.4424 1599.0908 1599.4822 1599.5094 1599.303  1599.316 ]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "predict(notebook=True);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [],
   "source": [
    "window = Tk()\n",
    "width = 450\n",
    "height = 600\n",
    "pos_x = '+0' # Specify +/- for position from left or right side\n",
    "pos_y = '+0' # Specify +/- for position from left or right side\n",
    "window.title('CVD process predicter')\n",
    "window.geometry('{}x{}{}{}'.format(width, height, pos_x, pos_y))\n",
    "#window.config()\n",
    "\n",
    "lbl_intro = Label(window, text=\"Define process parameters\", font=(\"Helvetica\", 10))\n",
    "lbl_intro.place(x=5, y=5)\n",
    "\n",
    "pos_x_labels = 5\n",
    "pos_y_labels = 35\n",
    "pos_x_in = pos_x_labels + 150\n",
    "\n",
    "lbl_flow = Label(window, text=\"Flow factor: \", font=(\"Helvetica\", 11))\n",
    "lbl_flow.place(x=pos_x_labels, y=pos_y_labels)\n",
    "in_flow = Entry(window, bd=1)\n",
    "in_flow.insert(0, '0.93')\n",
    "in_flow.place(x=pos_x_in, y=pos_y_labels+2)\n",
    "\n",
    "lbl_spacing = Label(window, text=\"Spacing: \", font=(\"Helvetica\", 11))\n",
    "lbl_spacing.place(x=pos_x_labels, y=pos_y_labels+5*6)\n",
    "in_spacing = Entry(window, bd=1)\n",
    "in_spacing.insert(0, '0.35')\n",
    "in_spacing.place(x=pos_x_in, y=pos_y_labels+5*6+2)\n",
    "\n",
    "lbl_deptime = Label(window, text=\"Deposition time [s]: \", font=(\"Helvetica\", 11))\n",
    "lbl_deptime.place(x=pos_x_labels, y=pos_y_labels+5*11)\n",
    "in_deptime = Entry(window, bd=1)\n",
    "in_deptime.insert(0, '68')\n",
    "in_deptime.place(x=pos_x_in, y=pos_y_labels+5*11+2)\n",
    "\n",
    "lbl_temperature = Label(window, text=\"Temperature [°C]: \", font=(\"Helvetica\", 11))\n",
    "lbl_temperature.place(x=pos_x_labels, y=pos_y_labels+5*16)\n",
    "in_temperature = Entry(window, bd=1)\n",
    "in_temperature.insert(0, '480')\n",
    "in_temperature.place(x=pos_x_in, y=pos_y_labels+5*16+2)\n",
    "\n",
    "y_tool = pos_y_labels+5*21\n",
    "lbl_tool = Label(window, text=\"Tool: \", font=(\"Helvetica\", 11))\n",
    "lbl_tool.place(x=pos_x_labels, y=y_tool)\n",
    "cb_tool = Combobox(window, values=['1', '2', '3', '4'])\n",
    "cb_tool.place(x=pos_x_in, y=y_tool)\n",
    "cb_tool.current(0)\n",
    "\n",
    "\n",
    "btn = Button(window, text=\"Calculate\", fg='blue', command=predict)\n",
    "btn.place(x=pos_x_labels, y=y_tool+35);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Process parameters:\n",
      "FLOWFACTOR:\t0.93\n",
      "SPACING:\t0.35\n",
      "DEP TIME:\t68 s\n",
      "TOOL:\t\t#1\n",
      "Temperature:\t480 °C\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "window.mainloop();"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}