{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "1bc1e215f15013c69c7bfbd76f2a047f",
     "grade": false,
     "grade_id": "cell-23f9a13fcbe0410a",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "# Exercise Sheet No. 6\n",
    "\n",
    "---\n",
    "\n",
    "> Machine Learning for Natural Sciences, Summer 2024, Jun.-Prof. Pascal Friederich\n",
    "> \n",
    "> Deadline: Monday 03.06.2024, 8:00 am\n",
    ">\n",
    "> Tutor: houssam.metni@kit.edu\n",
    ">\n",
    "> **Please ask questions in the forum/discussion board and only contact the Tutor when there are issues with the grading**\n",
    "\n",
    "\n",
    "---\n",
    "\n",
    "**Topic**: This exercise sheet will focus on feed-forward neural networks, their implementation and training, as well as an application to materials science."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Please add here your group members' names and student IDs. \n",
    "\n",
    "You are encouraged to work in groups of a maximum of 3 people, however **each of you** has to submit a solution.\n",
    "\n",
    "Names: Nils Lennart Bruns\n",
    "\n",
    "IDs: 2460137"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "d57acb4fd0f979daecb1c62cf876a3d4",
     "grade": false,
     "grade_id": "cell-46bffdcf61984b78",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "# Application\n",
    "\n",
    "Next week you will start learning about applications of NNs in materials science. To give you a little insight, we will use an application in materials science already here:\n",
    "\n",
    "# Organic Solar Cells\n",
    "For organic materials to become semi-conducting, electrons must be delocalized in the molecule. For electrons to be delocalized, a high level of conjugation is necessary:  \n",
    "When single and double bonds are alternating in an organic molecule, electrons can move. When we think about an aromatic ring, like benzene, it is not defined where the double bonds would form, so they can move around the ring and are delocalized along the whole aromatic system:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "a7af2d10231edb2545565e09c316ccbc",
     "grade": false,
     "grade_id": "cell-81ee300cc1325d05",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1;31merror\u001b[0m: \u001b[1mexternally-managed-environment\u001b[0m\n",
      "\n",
      "\u001b[31m×\u001b[0m This environment is externally managed\n",
      "\u001b[31m╰─>\u001b[0m To install Python packages system-wide, try apt install\n",
      "\u001b[31m   \u001b[0m python3-xyz, where xyz is the package you are trying to\n",
      "\u001b[31m   \u001b[0m install.\n",
      "\u001b[31m   \u001b[0m \n",
      "\u001b[31m   \u001b[0m If you wish to install a non-Debian-packaged Python package,\n",
      "\u001b[31m   \u001b[0m create a virtual environment using python3 -m venv path/to/venv.\n",
      "\u001b[31m   \u001b[0m Then use path/to/venv/bin/python and path/to/venv/bin/pip. Make\n",
      "\u001b[31m   \u001b[0m sure you have python3-full installed.\n",
      "\u001b[31m   \u001b[0m \n",
      "\u001b[31m   \u001b[0m If you wish to install a non-Debian packaged Python application,\n",
      "\u001b[31m   \u001b[0m it may be easiest to use pipx install xyz, which will manage a\n",
      "\u001b[31m   \u001b[0m virtual environment for you. Make sure you have pipx installed.\n",
      "\u001b[31m   \u001b[0m \n",
      "\u001b[31m   \u001b[0m See /usr/share/doc/python3.11/README.venv for more information.\n",
      "\n",
      "\u001b[1;35mnote\u001b[0m: If you believe this is a mistake, please contact your Python installation or OS distribution provider. You can override this, at the risk of breaking your Python installation or OS, by passing --break-system-packages.\n",
      "\u001b[1;36mhint\u001b[0m: See PEP 668 for the detailed specification.\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<rdkit.Chem.rdchem.Mol at 0x7f61f17835a0>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "%pip install rdkit\n",
    "import rdkit\n",
    "from rdkit import Chem\n",
    "from rdkit.Chem import Draw\n",
    "from rdkit.Chem.Draw import IPythonConsole\n",
    "\n",
    "mol = Chem.MolFromSmiles('c1ccccc1')\n",
    "mol\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "2ce8ac009cd582e199ac642fb9f5fe34",
     "grade": false,
     "grade_id": "cell-16b46a9bf02c686c",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "These electrons have higher energies and are part of what is called the highest occupied molecular orbit (HOMO). This is basically the equivalent of the valence band in classical semiconductors.  \n",
    "\n",
    "The equivalent of the conduction band is the lowest level at higher energies that is unoccupied or the lowest unoccupied molecular orbit (LUMO). The gap between those two levels is the bandgap of organic semiconductors.\n",
    "\n",
    "For organic photovoltaic cells this bandgap needs to be small enough so that visible light can excite an electron from the HOMO to the LUMO. This requires a high level of conjugation and hence aromatic systems (Figure 1)\n",
    "\n",
    "<a title=\"Alevina89, CC BY-SA 3.0 &lt;https://creativecommons.org/licenses/by-sa/3.0&gt;, via Wikimedia Commons\" href=\"https://commons.wikimedia.org/wiki/File:Homo-lumo_gap.tif\"><img width=\"512\" alt=\"Homo-lumo gap\" src=\"https://upload.wikimedia.org/wikipedia/commons/thumb/5/58/Homo-lumo_gap.tif/lossless-page1-607px-Homo-lumo_gap.tif.png\"></a>\n",
    "<p style=\"text-align:center;font-size:80%;font-style:italic\">\n",
    "    Figure 1: Conjugation induced LUMO reduction.\n",
    "</p>\n",
    "\n",
    "For the development of organic semiconductors, HOMO and LUMO can be simulated by density functional theory (DFT) using different levels of theory. All you need to know now, is that depending on the level of theory  - or better the details of the approximations taken - the calculation of properties like HOMO and LUMO from the molecule can take hours.\n",
    "\n",
    "This is a problem for high-throughput screening if one wants to discover new materials. It is extremely costly to evaluate e.g. 100,000 molecules for their properties with methods that are precise enough. Hence, one usually tries to do a detailed simulation only on a subset of molecules and then train a ML-model to predict the properties of interest on the labeled data.\n",
    "\n",
    "## Dataset\n",
    "The dataset contains the simulated LUMO and HOMO values for 51,247 organic molecules. Additionally, it contains 63 molecule descriptors that were calculated using [rdkit](https://www.rdkit.org/docs/index.html):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "75f4304eac196581641748ac695fb946",
     "grade": false,
     "grade_id": "cell-fda560899aaf7a0a",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "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>LUMO</th>\n",
       "      <th>HOMO</th>\n",
       "      <th>NumHAcceptors</th>\n",
       "      <th>NumHDonors</th>\n",
       "      <th>NumHeteroatoms</th>\n",
       "      <th>NumValenceElectrons</th>\n",
       "      <th>fr_thiophene</th>\n",
       "      <th>PEOE_VSA5</th>\n",
       "      <th>PEOE_VSA6</th>\n",
       "      <th>PEOE_VSA7</th>\n",
       "      <th>...</th>\n",
       "      <th>HeavyAtomCount</th>\n",
       "      <th>HeavyAtomMolWt</th>\n",
       "      <th>fr_imide</th>\n",
       "      <th>LabuteASA</th>\n",
       "      <th>MaxAbsEStateIndex</th>\n",
       "      <th>MaxEStateIndex</th>\n",
       "      <th>fr_nitrile</th>\n",
       "      <th>MolWt</th>\n",
       "      <th>NHOHCount</th>\n",
       "      <th>NOCount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "      <td>51247.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>-3.469117</td>\n",
       "      <td>-6.278187</td>\n",
       "      <td>9.479892</td>\n",
       "      <td>0.434699</td>\n",
       "      <td>11.535895</td>\n",
       "      <td>176.291646</td>\n",
       "      <td>1.188674</td>\n",
       "      <td>18.012681</td>\n",
       "      <td>14.876386</td>\n",
       "      <td>29.051809</td>\n",
       "      <td>...</td>\n",
       "      <td>36.115675</td>\n",
       "      <td>502.741343</td>\n",
       "      <td>0.533202</td>\n",
       "      <td>211.089329</td>\n",
       "      <td>12.268723</td>\n",
       "      <td>12.268722</td>\n",
       "      <td>0.848284</td>\n",
       "      <td>516.483676</td>\n",
       "      <td>0.463617</td>\n",
       "      <td>8.701622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.556705</td>\n",
       "      <td>0.586477</td>\n",
       "      <td>2.439179</td>\n",
       "      <td>0.773003</td>\n",
       "      <td>2.727198</td>\n",
       "      <td>29.118399</td>\n",
       "      <td>1.147485</td>\n",
       "      <td>16.466421</td>\n",
       "      <td>21.106906</td>\n",
       "      <td>16.974289</td>\n",
       "      <td>...</td>\n",
       "      <td>6.068335</td>\n",
       "      <td>81.664870</td>\n",
       "      <td>0.882705</td>\n",
       "      <td>35.016824</td>\n",
       "      <td>2.787968</td>\n",
       "      <td>2.787973</td>\n",
       "      <td>1.286705</td>\n",
       "      <td>83.755283</td>\n",
       "      <td>0.865545</td>\n",
       "      <td>2.913345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-5.785082</td>\n",
       "      <td>-8.829189</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>54.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>146.085000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>61.168567</td>\n",
       "      <td>2.188704</td>\n",
       "      <td>2.188704</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>149.109000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-3.879740</td>\n",
       "      <td>-6.699339</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>160.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>17.696186</td>\n",
       "      <td>...</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>458.376000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>190.573070</td>\n",
       "      <td>11.637669</td>\n",
       "      <td>11.637669</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>470.446000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-3.480653</td>\n",
       "      <td>-6.319910</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>11.336786</td>\n",
       "      <td>6.076020</td>\n",
       "      <td>28.832742</td>\n",
       "      <td>...</td>\n",
       "      <td>37.000000</td>\n",
       "      <td>516.482000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>215.878722</td>\n",
       "      <td>13.047850</td>\n",
       "      <td>13.047850</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>530.613000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>-3.069627</td>\n",
       "      <td>-5.820485</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>196.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>23.098671</td>\n",
       "      <td>23.523770</td>\n",
       "      <td>40.041517</td>\n",
       "      <td>...</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>558.416000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>234.220092</td>\n",
       "      <td>13.919575</td>\n",
       "      <td>13.919575</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>573.685000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>11.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>-1.577372</td>\n",
       "      <td>-4.088010</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>318.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>124.704645</td>\n",
       "      <td>169.858278</td>\n",
       "      <td>136.508949</td>\n",
       "      <td>...</td>\n",
       "      <td>66.000000</td>\n",
       "      <td>902.998000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>380.808169</td>\n",
       "      <td>17.589617</td>\n",
       "      <td>17.589617</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>904.006000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>24.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 65 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               LUMO          HOMO  NumHAcceptors    NumHDonors  \\\n",
       "count  51247.000000  51247.000000   51247.000000  51247.000000   \n",
       "mean      -3.469117     -6.278187       9.479892      0.434699   \n",
       "std        0.556705      0.586477       2.439179      0.773003   \n",
       "min       -5.785082     -8.829189       0.000000      0.000000   \n",
       "25%       -3.879740     -6.699339       8.000000      0.000000   \n",
       "50%       -3.480653     -6.319910       9.000000      0.000000   \n",
       "75%       -3.069627     -5.820485      11.000000      1.000000   \n",
       "max       -1.577372     -4.088010      23.000000      6.000000   \n",
       "\n",
       "       NumHeteroatoms  NumValenceElectrons  fr_thiophene     PEOE_VSA5  \\\n",
       "count    51247.000000         51247.000000  51247.000000  51247.000000   \n",
       "mean        11.535895           176.291646      1.188674     18.012681   \n",
       "std          2.727198            29.118399      1.147485     16.466421   \n",
       "min          0.000000            54.000000      0.000000      0.000000   \n",
       "25%          9.000000           160.000000      0.000000      0.000000   \n",
       "50%         12.000000           180.000000      1.000000     11.336786   \n",
       "75%         13.000000           196.000000      2.000000     23.098671   \n",
       "max         28.000000           318.000000      9.000000    124.704645   \n",
       "\n",
       "          PEOE_VSA6     PEOE_VSA7  ...  HeavyAtomCount  HeavyAtomMolWt  \\\n",
       "count  51247.000000  51247.000000  ...    51247.000000    51247.000000   \n",
       "mean      14.876386     29.051809  ...       36.115675      502.741343   \n",
       "std       21.106906     16.974289  ...        6.068335       81.664870   \n",
       "min        0.000000      0.000000  ...       11.000000      146.085000   \n",
       "25%        0.000000     17.696186  ...       32.000000      458.376000   \n",
       "50%        6.076020     28.832742  ...       37.000000      516.482000   \n",
       "75%       23.523770     40.041517  ...       40.000000      558.416000   \n",
       "max      169.858278    136.508949  ...       66.000000      902.998000   \n",
       "\n",
       "           fr_imide     LabuteASA  MaxAbsEStateIndex  MaxEStateIndex  \\\n",
       "count  51247.000000  51247.000000       51247.000000    51247.000000   \n",
       "mean       0.533202    211.089329          12.268723       12.268722   \n",
       "std        0.882705     35.016824           2.787968        2.787973   \n",
       "min        0.000000     61.168567           2.188704        2.188704   \n",
       "25%        0.000000    190.573070          11.637669       11.637669   \n",
       "50%        0.000000    215.878722          13.047850       13.047850   \n",
       "75%        1.000000    234.220092          13.919575       13.919575   \n",
       "max        6.000000    380.808169          17.589617       17.589617   \n",
       "\n",
       "         fr_nitrile         MolWt     NHOHCount       NOCount  \n",
       "count  51247.000000  51247.000000  51247.000000  51247.000000  \n",
       "mean       0.848284    516.483676      0.463617      8.701622  \n",
       "std        1.286705     83.755283      0.865545      2.913345  \n",
       "min        0.000000    149.109000      0.000000      0.000000  \n",
       "25%        0.000000    470.446000      0.000000      6.000000  \n",
       "50%        0.000000    530.613000      0.000000      9.000000  \n",
       "75%        2.000000    573.685000      1.000000     11.000000  \n",
       "max       11.000000    904.006000      8.000000     24.000000  \n",
       "\n",
       "[8 rows x 65 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# get dataset\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "import requests\n",
    "\n",
    "url = 'https://bwsyncandshare.kit.edu/s/WcG2c96Xq82gB4E/download'\n",
    "response: str = requests.get(url).content.decode('utf-8')\n",
    "with open('OPV.csv','w') as file:\n",
    "    file.write(response)\n",
    "\n",
    "df = pd.read_csv('OPV.csv', header=1)\n",
    "df.describe()\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "afebed28e987e219d17a79b68be03bd5",
     "grade": false,
     "grade_id": "cell-f36353d854da7e0e",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# HOMO/LUMO visualization\n",
    "df['LUMO'].hist(bins=100, label='LUMO')\n",
    "df['HOMO'].hist(bins=100, label='HOMO')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "cbc553c4334a88a8744e0e219a2d50d2",
     "grade": false,
     "grade_id": "cell-4d50ab2729193a45",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "# Linear regression benchmark\n",
    "For a first shot we can train a ridge regression to predict the `labels` from the `mol_descriptors`.\n",
    "\n",
    "This time we will use `sklearn`. Additionally we will also pre-process the features with a standard scaler to shift them to zero mean and scale them to unit variance.\n",
    "\n",
    "As already mentioned, the calculation of DFT properties can be very costly. Hence, we want to have a model that extrapolates well to unseen data. Hence, we will use only 20% of the dataset for training and test on 80%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "46be2ca9e3ca32c4c1df4b2600f0c53d",
     "grade": false,
     "grade_id": "cell-54d8b4c6288e7337",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "from sklearn.linear_model import LinearRegression, Ridge\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "mol_descriptors = df.columns[2:]\n",
    "\n",
    "X = df[mol_descriptors].values\n",
    "y = df[['LUMO','HOMO']].values\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "6cdddd44300266f722e86661821e83ce",
     "grade": false,
     "grade_id": "cell-e7d2a2a8bde77257",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Next we use the [`train_test_split()`](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html?highlight=train_test_split#sklearn.model_selection.train_test_split) to split of 20% of `X` and `y` as train set and use the rest as test set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "e835d707374b1b1c94d73a24496a0eb8",
     "grade": false,
     "grade_id": "cell-7f158597a8c0d652",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.8, random_state=0)\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we fit a `StandardScaler` object to scale `X_train`, thus obtaining `X_train_scaled`, and use the parameters of the scaler to transform `X_test` to `X_test_scaled` (so using the same scaler from the train test). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "9a699faf0aec8a7c992a26799188a6ba",
     "grade": false,
     "grade_id": "Scaling_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# Assign a StandardScaler object to scaler and obtain the scaled fatures as X_train_scaled and X_test_scaled\n",
    "scaler = None\n",
    "X_train_scaled = None\n",
    "X_test_scaled = None\n",
    "\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_train)\n",
    "X_train_scaled = scaler.transform(X_train)\n",
    "X_test_scaled = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "d81dc3d73d22958f4965037bcfde49c7",
     "grade": true,
     "grade_id": "Scaling_Test",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: Scaling_Test - possible points: 1\n",
    "\n",
    "# Scaler - 1 point\n",
    "\n",
    "assert isinstance(scaler, StandardScaler), \"The scaler should be an instance of the sklearn StandardScaler\"\n",
    "assert X_train_scaled.shape == X_train.shape, \"The shape of the scaled array should be the same as the unscaled\"\n",
    "assert X_test_scaled.shape == X_test.shape, \"The shape of the scaled array should be the same as the unscaled\"\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "47d36403817e5c3db554ab3db412a45f",
     "grade": false,
     "grade_id": "cell-d8a9322b1f746727",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# renaming variables for simplicity\n",
    "\n",
    "X_train = X_train_scaled\n",
    "X_test = X_test_scaled\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "69121501e7e945fbe23241cab9c25129",
     "grade": false,
     "grade_id": "cell-861d7538c8eb14bd",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Now use the [`Ridge()`](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Ridge.html?highlight=ridge#sklearn.linear_model.Ridge) model to fit a ridge regression with standard parameters to the train data, and assign the predictions of the fitted model on the test data to `y_pred`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "6ee4cdc139df320687db5397d6c01f15",
     "grade": false,
     "grade_id": "Ridge_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# Instantiate a Ridge() model as model, fit it and assign the predictions to y_pred\n",
    "model = None\n",
    "y_pred = None\n",
    "\n",
    "model = Ridge(alpha=1.0)\n",
    "model.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "4495a4729984b39f7e35a4803d65c482",
     "grade": false,
     "grade_id": "cell-6cceb6120d7e8ba5",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R2 LUMO: 0.7165783593656454\n",
      "R2 HOMO: 0.7971706345668933\n"
     ]
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "r2_lumo_ridge = r2_score(y_test[:, 0], y_pred[:, 0])\n",
    "r2_homo_ridge = r2_score(y_test[:, 1], y_pred[:, 1])\n",
    "print(f'R2 LUMO: {r2_lumo_ridge}\\nR2 HOMO: {r2_homo_ridge}')\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "29ac1fbd4209a19536d4fd3f1c8ac4e1",
     "grade": true,
     "grade_id": "Ridge_Test",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: Ridge_Test - possible points: 1\n",
    "\n",
    "# Ridge Regression - 1 point\n",
    "\n",
    "assert y_pred.shape[0] == y_test.shape[0], 'assert that the shapes are the same'\n",
    "assert y_pred.shape[1] == y_test.shape[1], 'assert that the shapes are the same'\n",
    "assert r2_lumo_ridge > 0.70, 'the r2 score should be of at least around 0.7'\n",
    "assert r2_homo_ridge > 0.70, 'the r2 score should be of at least around 0.7'\n",
    "# Possible hidden tests\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "c8410820da60cf6b42090c4ab97714ff",
     "grade": false,
     "grade_id": "cell-d3d706cc5b4b7361",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Now we can plot the correlations of the the true and predicted values:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "015649585578733901a1ef86685f8165",
     "grade": false,
     "grade_id": "cell-4f27d57d7fb34459",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "fig, axs = plt.subplots(1, 2, figsize=(12, 6))\n",
    "\n",
    "axs[0].scatter(y_test[:, 0], y_pred[:, 0], alpha=0.2, s=5)\n",
    "axs[0].plot([-6, -1], [-6, -1], 'k')\n",
    "axs[0].set_title(f'R2 LUMO: {r2_lumo_ridge}')\n",
    "axs[0].set_xlabel('true LUMO')\n",
    "axs[0].set_ylabel('predicted LUMO')\n",
    "axs[0].set_xlim([-6, -1])\n",
    "axs[0].set_ylim([-6, -1])\n",
    "\n",
    "axs[1].scatter(y_test[:, 1], y_pred[:, 1], alpha=0.2, s=5)\n",
    "axs[1].plot([-9, -3], [-9, -3], 'k')\n",
    "axs[1].set_title(f'R2 HOMO: {r2_homo_ridge}')\n",
    "axs[1].set_xlabel('true HOMO')\n",
    "axs[1].set_ylabel('predicted HOMO')\n",
    "axs[1].set_xlim([-9, -3])\n",
    "axs[1].set_ylim([-9, -3])\n",
    "\n",
    "plt.show()\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "52e8ecf636b85cc39788f21b6f9b6977",
     "grade": false,
     "grade_id": "cell-3b3e939a2ea02702",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "As you can see, the ridge regression learns something, but the predictions are still way off from the true values.\n",
    "\n",
    "Hence, we will apply a non-linear model, namely a:\n",
    "\n",
    "# Neural Network\n",
    "\n",
    "## Inspiration from Neuroscience\n",
    "In this week's lecture you saw that feedforward neural networks can be thought of as 'function approximation machines' [Goodfellow et al. 2016](https://www.deeplearningbook.org/contents/mlp.html).\n",
    "Some ideas of todays artificial neural networks are loosely inspired by neuroscientific models of biological networks.\n",
    "For a different perspective we can try to make a connection to models of biological neural networks (while remembering the goal of todays neural networks is not to get a model of the brain but rather to approximate some function).\n",
    "\n",
    "The fundamental units of the brain are neurons which can also be thought of as information messengers. Neurons receive, transmit and transform information in the form of electrical impulses.\n",
    "A neuron integrates information from its upstream neurons at the dendrites, creating a post-synaptic potential. \n",
    "At the axon hillock this potential is encoded into a spike train of action potentials. These are sent down the axon until they reach a synapse of an axon terminal. At the synapse the spike train is again decoded into a pre-synaptic potential and depending on the incoming spike frequency more or less neurotransmitters are released to pass the signal on to the next neuron (Figure 2).\n",
    "<p style=\"text-align:center;font-size:80%;font-style:italic\">\n",
    "    <img src=\"https://images.topperlearning.com/topper/tinymce/imagemanager/files/Synapse_between_Neurons.jpg\", width=\"70%\">\n",
    "    <br>\n",
    "    Figure 2: Simplified sketch of a neuron.\n",
    "</p>\n",
    "\n",
    "\n",
    "## Forward Pass\n",
    "\n",
    "For todays artificial neural networks this was simplified in the following way:  \n",
    "Instead of simulating spikes, we only simulate a spike frequency as a continuous value. Depending on the post-synaptic potential, this frequency can be higher or lower. E.g. neurons do not respond at all until a specific post-synaptic potential is reached and also have a maximum spike-frequency. This implies two things:  \n",
    "**1. The transformation from post-synaptic potential to spike-frequency is non-linear. (activation function)**  \n",
    "**2. Each neuron has a different base activity. (bias)**\n",
    "\n",
    "In the following mathematical definitions, lower-case letters denote scalars, while upper-case letters denote vectors or matrices. No dot or $\\cdot$ denotes a dot product of vectors or matrices.\n",
    "\n",
    "### Single neuron\n",
    "Given a subjected neuron $l$ receives inputs from an upstream layer $k$ and the upstream layer consists of three neurons with the outputs $x_1, x_2, x_3$ via connections with weights $\\theta_{1,l}, \\theta_{2,l}, \\theta_{3,l}$ (Figure 3), we can compute the weighted sum as state $h_l(X)$ of the subjected neuron by linear algebra as the product of the row vector $X_k$ and the column vector $\\Theta_{k,l}$. Additionally, we add a bias $b$ to the neuron.\n",
    "\n",
    "\\begin{align}\n",
    "h_l(X) &= X_k \\cdot \\Theta_{k,l} + b_l\\\\\n",
    "&=\n",
    "\\begin{bmatrix}\n",
    "    x_1 & x_2 & x_3\n",
    "\\end{bmatrix} \n",
    "\\cdot\n",
    "\\begin{bmatrix}\n",
    "    \\theta_{1,l} \\\\ \\theta_{2,l} \\\\ \\theta_{3,l}\n",
    "\\end{bmatrix}\n",
    "+ b_l\\\\\n",
    "&=x_1 \\theta_{1,l} +  x_2 \\theta_{2,l} + x_3 \\theta_{3,l} + b_l\n",
    "\\end{align}\n",
    "\n",
    "And the output or activation $a_l(x)$ of the subjected neuron in terms of spike-frequency is computed using a non-linear activation function $\\sigma()$:\n",
    "\n",
    "\\begin{align}\n",
    "a_l(X) &= \\sigma(h_l(X))\n",
    "\\end{align}\n",
    "\n",
    "<div>\n",
    "<img src=\"attachment:ff_b1.png\" width=\"30%\"/>\n",
    "</div>\n",
    "<p style=\"text-align:center;font-size:80%;font-style:italic\">\n",
    "Figure 3: A neuron layer k with three neurons has feed-forward connections to a single neuron l.\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "db7043aef1b70f848d5cf0f1458c1cba",
     "grade": false,
     "grade_id": "cell-3bb0abd58852685e",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "### Neuron Layer\n",
    "Now given that the subjected $l$ is not only a single neuron but e.g. a layer of two neurons (Figure 4), the notation is the same with the only difference that we add a column to the weight column vector $\\Theta_{k,l}$ changing it to a $3 \\times 2$ dimensional matrix to calculate the $1 \\times 2$ dimensional row state vector $H_l(X)$:\n",
    "\n",
    "\\begin{align}\n",
    "    H_l(X) &= X_k \\cdot \\Theta_{k,l} + B_l\\\\\n",
    "    &=\n",
    "    \\begin{bmatrix}\n",
    "        x_1 & x_2 & x_3\n",
    "    \\end{bmatrix} \n",
    "    \\cdot\n",
    "    \\begin{bmatrix}\n",
    "        \\theta_{1,1} & \\theta_{1,2} \\\\\n",
    "        \\theta_{2,1} & \\theta_{2,2} \\\\\n",
    "        \\theta_{3,1} & \\theta_{3,2} \n",
    "    \\end{bmatrix}\n",
    "    +\n",
    "    \\begin{bmatrix}\n",
    "        b_1 & b_2 \n",
    "    \\end{bmatrix} \\\\\n",
    "    &=\n",
    "    \\begin{bmatrix}\n",
    "        \\theta_{1,1} x_1 + \\theta_{1,2} x_2 + \\theta_{1,3} x_3 + b_1 &\n",
    "        \\theta_{2,1} x_1 + \\theta_{2,2} x_2 + \\theta_{2,3} x_3 + b_2\n",
    "    \\end{bmatrix}\n",
    "\\end{align}\n",
    "\n",
    "And the activation of the layer:\n",
    "\n",
    "\\begin{align}\n",
    "A_i(X) &= \\sigma(H_i(X))\n",
    "\\end{align}\n",
    "\n",
    "<div>\n",
    "<img src=\"attachment:ff_b2.png\" width=\"30%\"/>\n",
    "</div>\n",
    "<p style=\"text-align:center;font-size:80%;font-style:italic\">\n",
    "Figure 4: A neuron layer j with three neurons has feed-forward connections to a neuron layer i with two neurons.\n",
    "</p>\n",
    "\n",
    "For the bias there exist two approaches:\n",
    "1. One can see the bias as an additional input with a constant $1$ from the previous layer and a learnable weight vector.\n",
    "2. Adding the bias simply as row vector that added to the neuron state $H(x)$.\n",
    "\n",
    "Here we use approach 2."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "49521b88440b3051c8cd09e7638442d6",
     "grade": false,
     "grade_id": "cell-c59d409c87e07103",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "### Batches\n",
    "Usually data is processed in batches, so not only one single sample is processed at a time but multiple samples as this is much faster when calculated using vectorization. In our case, samples are rows with the length of the features. So when we pass 10 samples of our features (the `mol_descriptors`), we pass a matrix of shape $10 \\times 63$. This already fits well with our current notation and it will return a $10 \\times 32$ state matrix.\n",
    "\n",
    "\n",
    "### Weight initialization\n",
    "To compute anything, the weights must be different from $0$. There are several approaches to initialize weights. We will use the approach of [Glorot et Al., 2010](http://proceedings.mlr.press/v9/glorot10a.html).\n",
    "The initial weights are drawn form a uniform distribution $U(-z, z)$ with the limit $z$ calculated as:\n",
    "\\begin{align}\n",
    "    z &= \\sqrt{\\frac{6}{k+l}}\n",
    "\\end{align}\n",
    "...with input length $k$ and output length $l$.\n",
    "\n",
    "\n",
    "Just to get an idea how this all works in python we will build a layer with 32 neurons and feed all `mol_descriptors`.  \n",
    "Initialize the weight matrix `theta` using numpy functions:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "68161c0f161bbdbbbe4af6fa44453eec",
     "grade": false,
     "grade_id": "cell-d04191b4871926d9",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "sample = X_train[0:10]\n",
    "n_inputs = sample.shape[1]\n",
    "n_outputs = 32\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "e01d3deacd16eecf176ecc8087b9ea5e",
     "grade": false,
     "grade_id": "Weights_Init_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# Initialize the weights theta after Glorot et Al. 2010 (glorot uniform):\n",
    "theta = None\n",
    "\n",
    "z = np.sqrt(6/(n_inputs+n_outputs))\n",
    "theta = np.random.uniform(low=-z, high=z, size=(n_inputs, n_outputs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "3549875c2c12f3efc024d5b3eeb798b4",
     "grade": true,
     "grade_id": "Weights_Init_Test",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: Weights_Init_Test - possible points: 1\n",
    "\n",
    "# Weights - 1 point\n",
    "\n",
    "assert theta.shape[0] == n_inputs, \"Your weight shape doesn't match!\"\n",
    "assert theta.shape[1] == n_outputs, \"Your weight shape doesn't match!\"\n",
    "\n",
    "# Hidden asserts for the limits of the uniform distribution\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "ae1d129cf07164e34e2e64d2050a8ca7",
     "grade": false,
     "grade_id": "cell-c06d0e6e31460559",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Now calculate h(x) for the sample using your weights:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "27f82b54e470bee4ab57a418ff6ac3ba",
     "grade": false,
     "grade_id": "Bias_Init_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# Initialize the bias row vector with zeros:\n",
    "b = None\n",
    "\n",
    "b = np.zeros(n_outputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "9d66a647c59de39dd2324c4ef33e1d82",
     "grade": false,
     "grade_id": "hx_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# Calculate h(x) using sample, theta and b\n",
    "# The bias won't change anything but it will check for correct shapes.\n",
    "h = None\n",
    "\n",
    "h = sample.dot(theta) + b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "281e4a11f9b6a9966d056b105f4dd3c2",
     "grade": true,
     "grade_id": "hx_Test",
     "locked": true,
     "points": 2,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: hx_Test - possible points: 2\n",
    "\n",
    "# States - 2 points\n",
    "\n",
    "assert h.shape[0] == sample.shape[0]\n",
    "assert h.shape[1] == n_outputs\n",
    "# Possible hidden asserts\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "824f2e53f5b960d56ada84d76073907e",
     "grade": false,
     "grade_id": "cell-61b2ec56bb5109fc",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Activation functions\n",
    "\n",
    "Our model will be a regression model. For the output of the model we need to fit the unscaled `y`. Hence we need an unbound linear activation function for the output:\n",
    "\n",
    "\\begin{align}\n",
    "    linear(H(x)) &= H(x)\n",
    "\\end{align}\n",
    "\n",
    "As non-linear activation function for the hidden layers we will use the ReLu function:\n",
    "\n",
    "\\begin{align}\n",
    "    relu(H(x)) &= \n",
    "\\begin{cases} \n",
    "    0~\\text{ if }~h_l(x) \\leq 0\\\\\n",
    "    h_l(x)~\\text{ if }~h_l(x)>0\n",
    "\\end{cases}\n",
    "\\end{align}\n",
    "\n",
    "The ReLu has the biological motivation of having a minimal threshold until it starts outputting values different from 0. Additionally, it yields sparse activations in the network, which is usually favorable, and its gradient is trivial to calculate.\n",
    "\n",
    "Please calculate the ReLu of your previously calculated `h`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "070c2ce28349809a7f1cbb6c87bf9947",
     "grade": false,
     "grade_id": "relu_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "# Calculate the activation a by computing the ReLu of h\n",
    "a = None\n",
    "\n",
    "relu = lambda x: x*(x > 0)\n",
    "a = relu(h)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "5e7c4473ba10c3120b4c99bda1151726",
     "grade": true,
     "grade_id": "relu_Test",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: relu_Test - possible points: 1\n",
    "\n",
    "# ReLu - 1 point\n",
    "\n",
    "assert np.min(a) >= 0\n",
    "# Possible hidden asserts\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "06a7215672d538fc71d0e3175c0acc3e",
     "grade": false,
     "grade_id": "cell-3ef043f5a305266c",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## Backpropagation\n",
    "\n",
    "### Weights\n",
    "From the lecture recall the following chain to get the gradient of the error $J$ with respect to the weights $\\Theta$ of the output layer:\n",
    "\n",
    "\\begin{align}\n",
    "\\frac{\\partial J}{\\partial \\Theta} &= \\underset{(1)}{\\frac{\\partial J}{\\partial a}} \n",
    "                                      \\underset{(2)}{\\frac{\\partial a}{\\partial h}}\n",
    "                                      \\underset{(3)}{\\frac{\\partial h}{\\partial \\Theta}} \\\\\n",
    "\\end{align}\n",
    "\n",
    "For our case the error is the MSE defined as:\n",
    "\\begin{align}\n",
    "J &= \\frac{1}{2m} \\sum_{i=1}^m \\left( a^{(i)} - y^{(i)} \\right)^2\n",
    "\\end{align}\n",
    "...for $m$ samples $i$.\n",
    "\n",
    "Everything else is given. Now calculate the partial derivatives for a single sample $i$ of:  \n",
    "**(1) the MSE  \n",
    "(2) the linear and ReLu functions  \n",
    "(3) $h(x)$**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "512fd4ca60faeccbfec481da732716da",
     "grade": false,
     "grade_id": "cell-f60273a53983a843",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "\\begin{align}\\frac{\\partial \\text{MSE}}{\\partial a^{(i)}} &= \\frac{1}{m} \\left( a^{(i)} - y^{(i)} \\right) \\tag{1}\\\\\\frac{\\partial \\text{ReLu}^{(i)}}{\\partial h^{(i)}} &= \\begin{cases} 0~\\text{ if }~h^{(i)}(x) \\leq 0\\\\1~\\text{ if }~h^{(i)}(x)>0\\end{cases} \\tag{2}\\\\\\frac{\\partial \\text{linear}^{(i)}}{\\partial h^{(i)}} &= 1 \\tag{3}\\\\\\frac{\\partial h^{(i)}}{\\partial \\theta^{(i)}} &= x^{(i)} \\tag{4}\\end{align}"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "from IPython.display import display, Markdown\n",
    "\n",
    "display(Markdown(r\"\\begin{align}\"\n",
    "                 r\"\\frac{\\partial \\text{MSE}}{\\partial a^{(i)}} &= \\frac{1}{m} \\left( a^{(i)} - y^{(i)} \\right) \\tag{1}\\\\\"\n",
    "                 r\"\\frac{\\partial \\text{ReLu}^{(i)}}{\\partial h^{(i)}} &= \"\n",
    "                 r\"\\begin{cases} \"\n",
    "                 r\"0~\\text{ if }~h^{(i)}(x) \\leq 0\\\\\"\n",
    "                 r\"1~\\text{ if }~h^{(i)}(x)>0\"\n",
    "                 r\"\\end{cases} \\tag{2}\\\\\"\n",
    "                 r\"\\frac{\\partial \\text{linear}^{(i)}}{\\partial h^{(i)}} &= 1 \\tag{3}\\\\\"\n",
    "                 r\"\\frac{\\partial h^{(i)}}{\\partial \\theta^{(i)}} &= x^{(i)} \\tag{4}\"\n",
    "                 r\"\\end{align}\"))\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "2c162fb8d9235955d3e1ab6a7206f90f",
     "grade": false,
     "grade_id": "cell-a9b4a77b59c24fe3",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "### in case the above cell does not render, this is what the equations should look like ###\n",
    "import requests\n",
    "import matplotlib.image as mpimg\n",
    "url = 'https://bwsyncandshare.kit.edu/s/Tc8T2DjNaqqFENw/download'\n",
    "response: str = requests.get(url).content\n",
    "with open('equations.png','wb') as img:\n",
    "    img.write(response)\n",
    "img = mpimg.imread('equations.png')\n",
    "\n",
    "# Display the image\n",
    "plt.imshow(img)\n",
    "plt.axis('off')  # Hide the axes\n",
    "plt.show()\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "9127f8fdc3a9eed48231c48ae3550442",
     "grade": false,
     "grade_id": "cell-1941fc0d8efbd043",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Now let's take the above calculated `a` as output and calculate some random true labels `y_` for testing:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "075d73dba42a607e292e8f82d936beb9",
     "grade": false,
     "grade_id": "cell-fe1cdbcf98ba7063",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "y_ = np.random.uniform(low=0, high=3, size=[10, 32])\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "3bb439257470f04a1c070b89035b308c",
     "grade": false,
     "grade_id": "cell-3cc2c401b157a957",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Next we can use the derivatives from above to calculate the gradient for one layer:\n",
    "\n",
    "\\begin{align}\n",
    "\\frac{\\partial J}{\\partial \\Theta} &= \\underset{(1)}{\\frac{\\partial J}{\\partial a}} \n",
    "                                      \\underset{(2)}{\\frac{\\partial a}{\\partial h}}\n",
    "                                      \\underset{(3)}{\\frac{\\partial h}{\\partial \\Theta}} \\\\\n",
    "\\end{align}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "136c230a2e950a8353b2617d2aec6051",
     "grade": false,
     "grade_id": "gradient_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(63, 32) (63, 32)\n"
     ]
    }
   ],
   "source": [
    "# We start from the back:\n",
    "# (3) is trivial\n",
    "dh_dtheta = None\n",
    "\n",
    "dh_dtheta = sample\n",
    "\n",
    "# (2) for ReLu: Here you need a reference to h and maybe two steps\n",
    "da_dh = None\n",
    "\n",
    "da_dh = h>0\n",
    "\n",
    "# (1) m is the batch size!\n",
    "dJ_da = None\n",
    "\n",
    "dJ_da = 1/sample.shape[0]*(a-y_)\n",
    "\n",
    "dJ_dTheta = np.dot(dh_dtheta.T, (dJ_da * da_dh))\n",
    "# The error gradient with respect to the weights and the shape of the weights should agree:\n",
    "print(dJ_dTheta.shape, theta.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "65bf6306991bd6c8e3e9304105321e79",
     "grade": true,
     "grade_id": "gradient_Tests1",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: gradient_Tests1 - possible points: 1\n",
    "\n",
    "# dh dtheta - 1 point\n",
    "\n",
    "assert dh_dtheta.shape[0] == 10\n",
    "assert dh_dtheta.shape[1] == 63\n",
    "# Hidden test for the content of dh_dtheta\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "2fde87ba639b990f6b7f8de496a56802",
     "grade": true,
     "grade_id": "gradient_Tests2",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: gradient_Tests2 - possible points: 1\n",
    "\n",
    "# da dh - 1 point\n",
    "\n",
    "assert da_dh.shape[0] == 10\n",
    "assert da_dh.shape[1] == 32\n",
    "# Hidden test for the content of drelu_dh\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "c9d0c3ce0a0489fcd049e7f43a9c2998",
     "grade": true,
     "grade_id": "gradient_Tests3",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: gradient_Tests3 - possible points: 1\n",
    "\n",
    "# dJ dh - 1 point\n",
    "\n",
    "assert dJ_da.shape[0] == 10\n",
    "assert dJ_da.shape[1] == 32\n",
    "# Hidden test for the content of dJ_da\n",
    "\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "ef3779a3348fe0eebfea67a16cb684e9",
     "grade": false,
     "grade_id": "cell-d11a585e611a83a3",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "We can then update the weights for the next step using:\n",
    "\\begin{align}\n",
    "    \\Theta_{t+1} &= \\Theta_t - \\alpha \\cdot \\frac{\\partial J}{\\partial \\Theta} \\\\\n",
    "\\end{align}\n",
    "...with learning rate $\\alpha$."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "1823b6c4cf60af6f982eab9b18931d6a",
     "grade": false,
     "grade_id": "cell-e8f62b0ae9c7cc0d",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "### Bias \n",
    "\n",
    "Besides the weights we also need to fit the bias. The bias can be derived in the same manner, only that instead of $x$ the input is $1$. We simply multiply $1$ with the bias weights. Therefore (3) only for the bias collapses to:\n",
    "\\begin{align}\n",
    "\\frac{\\partial h^{(i)}_b}{\\partial \\theta^{(i)}_b} &= x^{(i)}_b = 1\\tag{3}\n",
    "\\end{align}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "8a65171be581659af0ed9758be9117f9",
     "grade": false,
     "grade_id": "cell-3032aeb4b06ebcbf",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Next we will build a simple feed forward network.\n",
    "\n",
    "Recall from the lecture (Figure 5):\n",
    "\n",
    "<div>\n",
    "<img src=\"attachment:backprop_1.png\" width=\"80%\"/>\n",
    "</div>\n",
    "<p style=\"text-align:center;font-size:80%;font-style:italic\">\n",
    "Figure 5: Backpropagation.\n",
    "</p>\n",
    "\n",
    "For backpropagation we are still missing the one element connecting the layers, namely $\\frac{\\partial a^2}{\\partial a^1}$:\n",
    "\\begin{align}\n",
    "a^2 &= \\Theta^2 a^1 \\\\\n",
    "\\frac{\\partial a^2}{\\partial a^1} &= \\Theta^2\n",
    "\\end{align}\n",
    "\n",
    "With this and the above derived equations we can calculated the gradient of the first (hidden) layer weights regarding the output error as:\n",
    "\\begin{align}\n",
    "\\frac{\\partial J}{\\partial \\Theta^1} &= \\underbrace{\\frac{\\partial J}{\\partial a^2}\n",
    "                                                    \\frac{\\partial a^2}{\\partial a^1}}_\\text{(a.)}\n",
    "                                        \\underbrace{\\frac{\\partial a^1}{\\partial h^1}\n",
    "                                                    \\frac{\\partial h^1}{\\partial \\Theta^1}}_\\text{(b.)} \\\\\n",
    "\\end{align}\n",
    "**The part (a.) can be calculated in the second (output) layer and is returned as upstream gradient.**\n",
    "\n",
    "**In the first (hidden) layer we can then use this gradient and combine it with part (b.) to obtain the gradient for the weights of layer 1 with respect to the output error.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "5c0b5cdb8859706de835f08112313de5",
     "grade": false,
     "grade_id": "cell-8988c9ddb69636de",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Let's put this into work by building a feed forward network for our regression task with:\n",
    "\n",
    "1. The 63 `mol_descriptors` as input $x^{(i)}$\n",
    "2. One hidden layer $1$ with 32 neurons and a ReLu activation: `class HiddenLayer`\n",
    "3. One output layer $2$ with two outputs (LUMO and HOMO) with a linear (no) activation function: `class OutputLayer`\n",
    "\n",
    "Both have a forward pass, which calculates the output of the layer, a backward pass which calculates the gradients and an update function that updates the weights using the gradients and the learning rate.\n",
    "\n",
    "Only the `OutputLayer` has to return the part (a.) from above to then pass it on to the backward pass of the `HiddenLayer`.\n",
    "\n",
    "We start implementing from the bottom up with the `OutputLayer` (layer 2):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "bf5d87aa465bff6e8c7385ba9094da4e",
     "grade": false,
     "grade_id": "outputlayer_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "class OutputLayer:\n",
    "\n",
    "    def __init__(self, n_inputs: int, n_outputs: int):\n",
    "        # Initialize (n_inputs X n_outputs)-dimensional weight matrix self.theta with a\n",
    "        # Glorot et Al. 2010 uniform initialization:\n",
    "        z = np.sqrt(6/(n_inputs+n_outputs))\n",
    "        self.theta = np.random.uniform(low=-z, high=z, size=(n_inputs, n_outputs))\n",
    "\n",
    "        # Initialize the bias vector self.b to zeros:\n",
    "        self.b = np.zeros(n_outputs)\n",
    "\n",
    "    def forward(self, input_vector):\n",
    "        self.input = input_vector\n",
    "        # Compute the states h(x) as self.h\n",
    "        self.h = np.dot(input_vector, self.theta) + self.b\n",
    "\n",
    "        # As this is a linear layer a(x) = h(x)\n",
    "        self.a = self.h\n",
    "        return self.a\n",
    "\n",
    "    def backward(self, y_predicted, y_true):\n",
    "        # HINT: as we do things backwards you might have to transpose some matrices\n",
    "\n",
    "        # partial derivative of the states with respect to the weights\n",
    "        dh2_dtheta2 = self.input\n",
    "\n",
    "        # partial derivative of activations with respect to the states\n",
    "        # One as we have a linear/no activation for the regression output\n",
    "        da2_dh2 = 1\n",
    "\n",
    "        # partial derivative of the error (MSE) with respect to the acivation\n",
    "        # infer the batch size from the input shape for normalization\n",
    "        dJ_da2 = 1/y_predicted.shape[0]*y_predicted.shape[1]*(y_predicted-y_true)\n",
    "\n",
    "        # Gradient of the weights with respect to the error\n",
    "        # for the weight updates for this layer:\n",
    "        self.dJ_dTheta2 = np.dot(dh2_dtheta2.transpose(), dJ_da2)\n",
    "        #print(dJ_da2.shape, y_predicted.shape)\n",
    "        #print(self.dJ_dTheta2.shape, self.theta.shape)\n",
    "        \n",
    "        # Gradient of the bias with respect to the error\n",
    "        # Recall using dh2_db2 = 1 and handle the batch size by summation\n",
    "        self.dJ_db2 = np.sum(np.dot(da2_dh2,dJ_da2), axis = 0)\n",
    "        #print(self.dJ_db2.shape, dJ_da2.shape, self.b.shape, self.theta.shape)\n",
    "        \n",
    "        # The downstream gradient for layer 1.\n",
    "        # employing da2_da1 = theta\n",
    "        downstream_gradient = np.dot(dJ_da2, self.theta.transpose())\n",
    "        #print(downstream_gradient.shape)\n",
    "        \n",
    "        return downstream_gradient\n",
    "\n",
    "    def update(self, learning_rate):\n",
    "        # You don't need to change this\n",
    "        # HINT:\n",
    "        # If your model gets worse instead of better make sure you calculate the correct MSE\n",
    "        self.theta = self.theta - learning_rate * self.dJ_dTheta2\n",
    "        self.b = self.b - learning_rate * self.dJ_db2\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "7977bbf24c4e5e11ff636999bc6085af",
     "grade": false,
     "grade_id": "cell-f6e7004a8a264380",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "X_sample = X_train[0:100, :]\n",
    "y_sample = y_train[0:100, :]\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "2ab61635926de664fe68f4a4c3df8925",
     "grade": true,
     "grade_id": "outputlayer_Test1",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: outputlayer_Test1 - possible points: 1\n",
    "\n",
    "# Output Forward Pass - 1 point\n",
    "\n",
    "l2 = OutputLayer(X_sample.shape[1], y_sample.shape[1])\n",
    "y_pred_sample = l2.forward(X_sample)\n",
    "\n",
    "assert y_sample.shape[0] == y_pred_sample.shape[0]\n",
    "assert y_sample.shape[1] == y_pred_sample.shape[1]\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "d02c25e1461758be1b980853bad10673",
     "grade": true,
     "grade_id": "outputlayer_Test2",
     "locked": true,
     "points": 2,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: outputlayer_Test2 - possible points: 2\n",
    "\n",
    "# Output Backward Pass - 2 points\n",
    "\n",
    "downstream_gradient = l2.backward(y_pred_sample, y_sample)\n",
    "\n",
    "assert downstream_gradient.shape[0] == X_sample.shape[0]\n",
    "assert downstream_gradient.shape[1] == X_sample.shape[1]\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "b8a2f3c8eb9303be0ee1c44dbb2a4975",
     "grade": true,
     "grade_id": "outputlayer_Test3",
     "locked": true,
     "points": 2,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: outputlayer_Test3 - possible points: 2\n",
    "\n",
    "# Output Weight Update - 2 points\n",
    "\n",
    "l2 = OutputLayer(X_sample.shape[1], y_sample.shape[1])\n",
    "y_pred_before = l2.forward(X_sample)\n",
    "\n",
    "for i in range(100):\n",
    "    y_pred_sample = l2.forward(X_sample)\n",
    "    l2.backward(y_pred_sample, y_sample)\n",
    "    l2.update(0.05)\n",
    "\n",
    "y_pred_after = l2.forward(X_sample)\n",
    "r2_before = r2_score(y_sample, y_pred_before)\n",
    "r2_after = r2_score(y_sample, y_pred_after)\n",
    "\n",
    "assert r2_before < r2_after\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "059fcb0020113e908311655a8d303b8c",
     "grade": false,
     "grade_id": "cell-7730164366507bcf",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Next is the `HiddenLayer`. There are mainly two main differences you have to keep in mind here:\n",
    "1. You have to work with the ReLu activation, so $\\frac{\\partial a}{\\partial h}$ isn't $1$ anymore.\n",
    "2. We use the `upstream_gradient` for the backward pass, provided by the `backward` method of the `OutputLayer`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "fd6d2580eff9e669d69b4f727d760a99",
     "grade": false,
     "grade_id": "hiddenlayer_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "class HiddenLayer:\n",
    "\n",
    "    def __init__(self, n_inputs: int, n_outputs: int):\n",
    "        # Initialize (n_inputs X n_outputs)-dimensional weight matrix self.theta with a\n",
    "        # Glorot et Al. 2010 uniform initialization:\n",
    "        z = np.sqrt(6/(n_inputs+n_outputs))\n",
    "        self.theta = np.random.uniform(low=-z, high=z, size=(n_inputs, n_outputs))\n",
    "\n",
    "        # Initialize the bias vector self.b to zeros:\n",
    "        self.b = np.zeros(n_outputs)\n",
    "\n",
    "    def forward(self, input_vector):      \n",
    "        self.input = input_vector\n",
    "        # Compute the states h(x) as self.h\n",
    "        self.h = input_vector.dot(self.theta) + self.b\n",
    "\n",
    "        # As this is a linear layer a(x) = h(x)\n",
    "        self.a = relu(self.h)\n",
    "        return self.a\n",
    "\n",
    "\n",
    "    def backward(self, upstream_gradient):\n",
    "        # HINT: as we do things backwards you might have to transpose some matrices\n",
    "\n",
    "        # Gradient of the states with respect to the inputs (trivial):\n",
    "        dh1_dtheta1 = self.input\n",
    "\n",
    "        # Gradient of the activations with respect to the states:\n",
    "        # Remember to apply ReLu here\n",
    "        da1_dh1 = self.h>0\n",
    "        \n",
    "        # Gradient of the error with respect to the weights:\n",
    "        # Now we can finally use the upstream gradient...\n",
    "        self.dJ_dTheta1 = np.dot(dh1_dtheta1.transpose(), (upstream_gradient*da1_dh1))\n",
    "        #print(dh1_dtheta1.shape, da1_dh1.shape, upstream_gradient.shape)\n",
    "        #print(f\"Ziel: {self.theta.shape}, Stand: {self.dJ_dTheta1.shape}\")\n",
    "\n",
    "        # And for the bias, similar to the output layer but this time\n",
    "        # using the upstream gradient:\n",
    "        self.dJ_db1 = np.sum(da1_dh1*upstream_gradient, axis=0)\n",
    "        #print(da1_dh1.shape, upstream_gradient.shape)\n",
    "        #print(\"dJ_db1\", self.dJ_db1.shape) \n",
    "\n",
    "    def update(self, learning_rate):\n",
    "        # You don't need to change this\n",
    "        self.theta = self.theta - learning_rate * self.dJ_dTheta1\n",
    "        self.b = self.b - learning_rate * self.dJ_db1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "a69b7a4a492e35f8b945e814402e5919",
     "grade": false,
     "grade_id": "cell-e60eb9f1a0ccc85a",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# We create a sample from the existing data that matches the output dimension \n",
    "# and also shift it above zero, so it is learnable with ReLu:\n",
    "\n",
    "n_hidden = 32\n",
    "\n",
    "X_sample = X_train[0:100, :]\n",
    "y_sample = np.array([y_train[i * 100:i * 100 + 100, 0] for i in range(n_hidden)]).T\n",
    "y_sample = y_sample - y_sample.min()\n",
    "\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 280,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "32e30608b6692222fc1253e1aa05d394",
     "grade": true,
     "grade_id": "hiddenlayer_Test1",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: hiddenlayer_Test1 - possible points: 1\n",
    "\n",
    "# Hidden Forward Pass - 1 point\n",
    "\n",
    "l1 = HiddenLayer(X_sample.shape[1], n_hidden)\n",
    "\n",
    "y_pred_sample = l1.forward(X_sample)\n",
    "assert y_pred_sample.shape[0] == X_sample.shape[0]\n",
    "assert y_pred_sample.shape[1] == n_hidden\n",
    "assert y_pred_sample.min() >= 0\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 281,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "822fb3143774661694bd3170c74580fc",
     "grade": true,
     "grade_id": "hiddenlayer_Test2",
     "locked": true,
     "points": 2,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: hiddenlayer_Test2 - possible points: 2\n",
    "\n",
    "# Hidden Weight Update - 2 points\n",
    "\n",
    "l1 = HiddenLayer(X_sample.shape[1], n_hidden)\n",
    "y_pred_before = l1.forward(X_sample)\n",
    "\n",
    "for i in range(100):\n",
    "    y_pred_sample = l1.forward(X_sample)\n",
    "    downstream_gradient = 1 / 100 * (y_pred_sample - y_sample)\n",
    "    l1.backward(downstream_gradient)\n",
    "    l1.update(0.05)\n",
    "\n",
    "y_pred_after = l1.forward(X_sample)\n",
    "r2_before = r2_score(y_sample, y_pred_before)\n",
    "r2_after = r2_score(y_sample, y_pred_after)\n",
    "\n",
    "assert r2_before < r2_after\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "b9bbe9d9c2449b2a9a74a5618c1bef63",
     "grade": false,
     "grade_id": "cell-692f27c0b8177b60",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "## NN Training\n",
    "\n",
    "Okay now let's see whether we can be better with our network than the benchmark ridge regression.\n",
    "\n",
    "As said before we will train our network in batches (`batch_size`) and for `n_epochs`.  \n",
    "Per epoch we therefore have to pass `X_train.shape[0] // batch_size` batches.  \n",
    "For each epoch we randomly shuffle the dataset. Using `np.random.permutation(X_train.shape[0])` we generate a randomly shuffled index. Indexing X and y with slices of this shuffled index, creates differently shuffled batches for each epoch."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 284,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "d0fa25ef5694e060cf8b5ed441678fb4",
     "grade": false,
     "grade_id": "cell-3aa9a880d4250f22",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "n_hidden = 64\n",
    "lr = 0.05\n",
    "n_epochs = 60\n",
    "batch_size = 100\n",
    "n_batches = X_train.shape[0] // batch_size\n",
    "\n",
    "l1 = HiddenLayer(X_train.shape[1], n_hidden)\n",
    "l2 = OutputLayer(n_hidden, y_train.shape[1])\n",
    "\n",
    "y_pred_before = l2.forward(l1.forward(X_test))\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {
    "deletable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "2438bc2e2affff486537185957c8d9dd",
     "grade": false,
     "grade_id": "training_Question",
     "locked": false,
     "schema_version": 3,
     "solution": true,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0: Test R2 = 0.24044537959708923\n",
      "Epoch 4: Test R2 = 0.716972551980976\n",
      "Epoch 8: Test R2 = 0.7791430858835686\n",
      "Epoch 12: Test R2 = 0.7899162249818088\n",
      "Epoch 16: Test R2 = 0.7934712322585863\n",
      "Epoch 20: Test R2 = 0.807017405223877\n",
      "Epoch 24: Test R2 = 0.7506001469059513\n",
      "Epoch 28: Test R2 = 0.792146002145901\n",
      "Epoch 32: Test R2 = 0.8249750324139051\n",
      "Epoch 36: Test R2 = 0.8279065315528997\n",
      "Epoch 40: Test R2 = 0.8358052319946528\n",
      "Epoch 44: Test R2 = 0.8294147510754667\n",
      "Epoch 48: Test R2 = 0.8299158542989007\n",
      "Epoch 52: Test R2 = 0.8360773761077643\n",
      "Epoch 56: Test R2 = 0.8348684326710152\n"
     ]
    }
   ],
   "source": [
    "for epoch in range(n_epochs):\n",
    "    permutation = np.random.permutation(X_train.shape[0])\n",
    "\n",
    "    for batch in range(n_batches):\n",
    "        start = batch * batch_size\n",
    "        end = start + batch_size\n",
    "        X_batch = X_train[permutation[start:end]]\n",
    "        y_batch = y_train[permutation[start:end]]\n",
    "\n",
    "        # Get the predictions\n",
    "        y_pred = None\n",
    "        \n",
    "        y_pred = l2.forward(l1.forward(X_batch))\n",
    "\n",
    "        # Do the backward pass of both layers and update the weights\n",
    "        downstream_gradient = l2.backward(y_pred, y_batch)\n",
    "        l1.backward(downstream_gradient)\n",
    "        l1.update(0.05)\n",
    "        l2.update(0.05)\n",
    "    \n",
    "    y_pred = l2.forward(l1.forward(X_test))\n",
    "    if epoch % 4 == 0:\n",
    "        print(f\"Epoch {epoch}: Test R2 = {r2_score(y_test, y_pred)}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "ad1b321f2e64fe325d04725084c7d53f",
     "grade": true,
     "grade_id": "training_Test",
     "locked": true,
     "points": 2,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: training_Test - possible points: 2\n",
    "\n",
    "# Full Training - 2 points\n",
    "\n",
    "y_pred_after = l2.forward(l1.forward(X_test))\n",
    "\n",
    "r2_before = r2_score(y_test, y_pred_before)\n",
    "r2_after = r2_score(y_test, y_pred_after)\n",
    "\n",
    "assert r2_before < r2_after\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "2d4a45d489566ca200e57234ff252400",
     "grade": false,
     "grade_id": "cell-d07c5db63d278253",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R2 LUMO: 0.7988797639044443\n",
      "R2 HOMO: 0.8841329392763239\n"
     ]
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "y_pred = l2.forward(l1.forward(X_test))\n",
    "\n",
    "r2_lumo_nn = r2_score(y_test[:, 0], y_pred[:, 0])\n",
    "r2_homo_nn = r2_score(y_test[:, 1], y_pred[:, 1])\n",
    "print(f'R2 LUMO: {r2_lumo_nn}\\nR2 HOMO: {r2_homo_nn}')\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "e926183a09b909610ed5a2bda665f725",
     "grade": true,
     "grade_id": "beating_ridge_test",
     "locked": true,
     "points": 1,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tRidge\t\t\tNN\n",
      "LUMO\t0.7165783593656454\t0.7988797639044443\n",
      "HOMO\t0.7971706345668933\t0.8841329392763239\n"
     ]
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "# ID: beating_ridge_test - possible points: 1\n",
    "\n",
    "# Beating Ridge - 1 point\n",
    "\n",
    "print(f'\\tRidge\\t\\t\\tNN\\nLUMO\\t{r2_lumo_ridge}\\t{r2_lumo_nn}\\nHOMO\\t{r2_homo_ridge}\\t{r2_homo_nn}')\n",
    "\n",
    "assert r2_lumo_nn > r2_lumo_ridge and r2_homo_nn > r2_homo_ridge\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "markdown",
     "checksum": "a21f12fc09b8b559bbe587e742877afb",
     "grade": false,
     "grade_id": "cell-20f695070da729a2",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "source": [
    "Quite possibly you were able to beat the ridge regression at that point, only with one hidden ReLu layer. Of course the hyperparameters of both models haven't been optimized yet, so you can get even better. But please do this in a separate notebook ;)\n",
    "\n",
    "Let's have a look at the final plots:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {
    "deletable": false,
    "editable": false,
    "nbgrader": {
     "cell_type": "code",
     "checksum": "239fcfd4410a71c9304ff3aca8eee8f5",
     "grade": false,
     "grade_id": "cell-28f5b2fb240f0b25",
     "locked": true,
     "schema_version": 3,
     "solution": false,
     "task": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "##### DO NOT CHANGE #####\n",
    "fig, axs = plt.subplots(1, 2, figsize=(12, 6))\n",
    "axs[0].scatter(y_test[:, 0], y_pred[:, 0], alpha=0.2, s=5)\n",
    "axs[0].plot([-6, -1], [-6, -1], 'k')\n",
    "axs[0].set_title(f'R2 LUMO: {r2_lumo_nn}')\n",
    "axs[0].set_xlabel('true LUMO')\n",
    "axs[0].set_ylabel('predicted LUMO')\n",
    "axs[0].set_xlim([-6, -1])\n",
    "axs[0].set_ylim([-6, -1])\n",
    "axs[1].scatter(y_test[:, 1], y_pred[:, 1], alpha=0.2, s=5)\n",
    "axs[1].plot([-9, -3], [-9, -3], 'k')\n",
    "axs[1].set_title(f'R2 HOMO: {r2_homo_nn}')\n",
    "axs[1].set_xlabel('true HOMO')\n",
    "axs[1].set_ylabel('predicted HOMO')\n",
    "axs[1].set_xlim([-9, -3])\n",
    "axs[1].set_ylim([-9, -3])\n",
    "\n",
    "plt.show()\n",
    "\n",
    "##### DO NOT CHANGE #####"
   ]
  }
 ],
 "metadata": {
  "jupytext": {
   "formats": "ipynb,py:light"
  },
  "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": "3.11.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}