vqc_pca_mnist_sebnotes.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "57bfd5ed-186f-461b-b796-3f8a6e2be619",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc3ebf33-fb21-4755-91f2-46a23f947d4f",
   "metadata": {},
   "source": [
    "**Importing MNIST dataset using Tensorflow**\n",
    "\n",
    "In this step, we assume you have intmernet connectivity. Connectivity is necessary for this tutorial because the tensorflow library imports the dataset from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz by default. If you do not have internet connectivity, for some strange reason, you can follow our \"How to import the MNIST dataset from a local directory\" tutorial.\n",
    "\n",
    "To load the MNIST dataset, use the following code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "150f3f87-b03c-4b4c-88d2-d5fb4d5e884d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\scipy\\__init__.py:155: UserWarning: A NumPy version >=1.18.5 and <1.25.0 is required for this version of SciPy (detected version 1.26.3\n",
      "  warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\ProgramData\\Anaconda3\\lib\\site-packages\\keras\\src\\losses.py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use tf.compat.v1.losses.sparse_softmax_cross_entropy instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.datasets import mnist"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "003e8e30-d6f5-4653-8161-90276461b29f",
   "metadata": {},
   "source": [
    "The MNIST dataset will be loaded as a set of training and test inputs (X) and outputs (Y). The imputs are samples of digit images while the outputs contain the numerical value each input represents. The exact shape of each dataset is provided below for your convenience."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "d8aaae1b-fc94-4e60-9aa1-295a6f6b9d09",
   "metadata": {},
   "outputs": [],
   "source": [
    "(x_train, y_train), (x_test, y_test) = mnist.load_data() #loads everything, we only want some specific stuff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3e79a11b-c759-4902-8d03-d476a1744f1c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MNIST Dataset Shape:\n",
      "x_train: (60000, 28, 28)\n",
      "y_train: (60000,)\n",
      "x_test:  (10000, 28, 28)\n",
      "y_test:  (10000,)\n"
     ]
    }
   ],
   "source": [
    "print('MNIST Dataset Shape:')\n",
    "print('x_train: ' + str(x_train.shape))\n",
    "print('y_train: ' + str(y_train.shape))\n",
    "print('x_test:  '  + str(x_test.shape))\n",
    "print('y_test:  '  + str(y_test.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f43133fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(28, 28)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train[0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "460ddc1e-d37b-4626-8e63-dfa9eb7d064f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\sebts\\AppData\\Local\\Temp\\ipykernel_10016\\4053146574.py:8: DeprecationWarning: ``qiskit.algorithms`` has been migrated to an independent package: https://github.com/qiskit-community/qiskit-algorithms. The ``qiskit.algorithms`` import path is deprecated as of qiskit-terra 0.25.0 and will be removed no earlier than 3 months after the release date. Please run ``pip install qiskit_algorithms`` and use ``import qiskit_algorithms`` instead.\n",
      "  from qiskit.algorithms.optimizers import ADAM, SPSA\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn import datasets\n",
    "import tensorflow as tf\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from qiskit.algorithms.optimizers import ADAM, SPSA\n",
    "from qiskit.circuit.library import TwoLocal, ZZFeatureMap\n",
    "from qiskit.primitives import BackendEstimator, BackendSampler\n",
    "from qiskit.quantum_info import Pauli\n",
    "from qiskit_aer import AerSimulator\n",
    "from qiskit_algorithms.gradients import ParamShiftEstimatorGradient, ParamShiftSamplerGradient\n",
    "\n",
    "from qiskit_machine_learning.algorithms.classifiers import NeuralNetworkClassifier\n",
    "from qiskit_machine_learning.neural_networks import EstimatorQNN, SamplerQNN\n",
    "from qiskit_machine_learning.utils.loss_functions import CrossEntropyLoss"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bc32a3c-6817-4f01-bc06-bc32b36467db",
   "metadata": {},
   "source": [
    "**Load the MNIST Dataset**\n",
    "\n",
    "MNIST dataset consists of thousand images of hand-written digits from 0 to 9. Every image has 28x28 pixels and every pixel's value is between 0 to 255. This dataset is too large for us. So we need to reduce it.\n",
    "\n",
    "1). Normalizing the pixel's value in the range of [0,1].\n",
    "\n",
    "2). Filtering images from the dataset, leaving only the images with number 3 and 6 as their labels.\n",
    "\n",
    "3). Reducing the total number of images in the dataset and divide the dataset into training dataset and testing dataset.\n",
    "\n",
    "4). Resize the image from 28x28 pixels to 16x16 pixels because QPCA needs the size of the image in the form of 2^n x 2^n pixels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "8f86806a-b8ad-4f98-b6b3-b1440ffd4817",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of original training examples: 60000\n",
      "Number of original test examples: 10000\n"
     ]
    }
   ],
   "source": [
    "(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()\n",
    "\n",
    "# Rescale the images from [0,255] to the [0.0,1.0] range.\n",
    "x_train, x_test = x_train[..., np.newaxis]/255.0, x_test[..., np.newaxis]/255.0 #this rescales the int intensity to 0-1 float\n",
    "#also for some reason adds another axis? (see shape below)\n",
    "\n",
    "print(\"Number of original training examples:\", len(x_train))\n",
    "print(\"Number of original test examples:\", len(x_test))\n",
    "     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ee04d089",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28, 1)\n",
      "(28, 28, 1)\n",
      "[[[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.01176471]\n",
      "  [0.07058824]\n",
      "  [0.07058824]\n",
      "  [0.07058824]\n",
      "  [0.49411765]\n",
      "  [0.53333333]\n",
      "  [0.68627451]\n",
      "  [0.10196078]\n",
      "  [0.65098039]\n",
      "  [1.        ]\n",
      "  [0.96862745]\n",
      "  [0.49803922]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.11764706]\n",
      "  [0.14117647]\n",
      "  [0.36862745]\n",
      "  [0.60392157]\n",
      "  [0.66666667]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.88235294]\n",
      "  [0.6745098 ]\n",
      "  [0.99215686]\n",
      "  [0.94901961]\n",
      "  [0.76470588]\n",
      "  [0.25098039]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.19215686]\n",
      "  [0.93333333]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.98431373]\n",
      "  [0.36470588]\n",
      "  [0.32156863]\n",
      "  [0.32156863]\n",
      "  [0.21960784]\n",
      "  [0.15294118]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.07058824]\n",
      "  [0.85882353]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.77647059]\n",
      "  [0.71372549]\n",
      "  [0.96862745]\n",
      "  [0.94509804]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.31372549]\n",
      "  [0.61176471]\n",
      "  [0.41960784]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.80392157]\n",
      "  [0.04313725]\n",
      "  [0.        ]\n",
      "  [0.16862745]\n",
      "  [0.60392157]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.05490196]\n",
      "  [0.00392157]\n",
      "  [0.60392157]\n",
      "  [0.99215686]\n",
      "  [0.35294118]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.54509804]\n",
      "  [0.99215686]\n",
      "  [0.74509804]\n",
      "  [0.00784314]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.04313725]\n",
      "  [0.74509804]\n",
      "  [0.99215686]\n",
      "  [0.2745098 ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.1372549 ]\n",
      "  [0.94509804]\n",
      "  [0.88235294]\n",
      "  [0.62745098]\n",
      "  [0.42352941]\n",
      "  [0.00392157]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.31764706]\n",
      "  [0.94117647]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.46666667]\n",
      "  [0.09803922]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.17647059]\n",
      "  [0.72941176]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.58823529]\n",
      "  [0.10588235]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.0627451 ]\n",
      "  [0.36470588]\n",
      "  [0.98823529]\n",
      "  [0.99215686]\n",
      "  [0.73333333]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.97647059]\n",
      "  [0.99215686]\n",
      "  [0.97647059]\n",
      "  [0.25098039]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.18039216]\n",
      "  [0.50980392]\n",
      "  [0.71764706]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.81176471]\n",
      "  [0.00784314]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.15294118]\n",
      "  [0.58039216]\n",
      "  [0.89803922]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.98039216]\n",
      "  [0.71372549]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.09411765]\n",
      "  [0.44705882]\n",
      "  [0.86666667]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.78823529]\n",
      "  [0.30588235]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.09019608]\n",
      "  [0.25882353]\n",
      "  [0.83529412]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.77647059]\n",
      "  [0.31764706]\n",
      "  [0.00784314]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.07058824]\n",
      "  [0.67058824]\n",
      "  [0.85882353]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.76470588]\n",
      "  [0.31372549]\n",
      "  [0.03529412]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.21568627]\n",
      "  [0.6745098 ]\n",
      "  [0.88627451]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.95686275]\n",
      "  [0.52156863]\n",
      "  [0.04313725]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.53333333]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.99215686]\n",
      "  [0.83137255]\n",
      "  [0.52941176]\n",
      "  [0.51764706]\n",
      "  [0.0627451 ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]\n",
      "\n",
      " [[0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]\n",
      "  [0.        ]]]\n"
     ]
    }
   ],
   "source": [
    "print(x_train.shape) # from (60000, 28, 28)?\n",
    "print(x_train[0].shape) #-- a single image but with an extra set of []\n",
    "print(x_train[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a7ddf841-356c-4e3f-9ac6-03a8ff46f0f4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5, 0, 4, ..., 5, 6, 8], dtype=uint8)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e21251ee-0cbd-402b-8aac-4bff71c15974",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of filtered training examples: 12665\n",
      "Number of filtered test examples: 2115\n"
     ]
    }
   ],
   "source": [
    "# choose the number to be classified here, in this case I selected number 0 and 1\n",
    "number_set = [0, 1]\n",
    "\n",
    "mask_train = np.isin(y_train, number_set) #is in is useful\n",
    "mask_test = np.isin(y_test, number_set)\n",
    "\n",
    "x_train = x_train[mask_train]\n",
    "y_train = y_train[mask_train]\n",
    "x_test = x_test[mask_test]\n",
    "y_test = y_test[mask_test]\n",
    "\n",
    "print(\"Number of filtered training examples:\", len(x_train))\n",
    "print(\"Number of filtered test examples:\", len(x_test))\n",
    "     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "26ebc1a3-85e2-4233-bc69-04efbcc58a32",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 1, ..., 1, 0, 1], dtype=uint8)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train # check we only have 1 and 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "439acb74",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(80,)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d7168a2a-7d00-4dba-bd3a-57be26db9566",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(80, 28, 28, 1) (80,)\n",
      "(20, 28, 28, 1) (20,)\n"
     ]
    }
   ],
   "source": [
    "# next, reduce the number of images inside the dataset\n",
    "\n",
    "index_train = np.array([], dtype=int)\n",
    "index_test = np.array([], dtype=int)\n",
    "\n",
    "# decide how many images we want inside the dataset\n",
    "num_data = 100\n",
    "# decide what is the percentage of the dataset we want for the training dataset\n",
    "train_percentage = 0.8\n",
    "\n",
    "# train data\n",
    "counter = np.array([0,0], dtype=int)\n",
    "for i in range (len(x_train)):\n",
    "    if y_train[i] == True and counter[0] < int(train_percentage*num_data/2): # num_data/2 because we want a 50-50 split of 1's and 0's\n",
    "        index_train = np.append(index_train, np.array([int(i)]), axis=0)\n",
    "        counter[0] += 1\n",
    "\n",
    "    if y_train[i] == False and counter[1] < int(train_percentage*num_data/2): # the == True and False works to select 0 and 1\n",
    "        index_train = np.append(index_train, np.array([int(i)]), axis=0)\n",
    "        counter[1] += 1\n",
    "\n",
    "# test data\n",
    "counter = np.array([0,0], dtype=int)\n",
    "for i in range (len(x_test)):\n",
    "    if y_test[i] == True and counter[0] < int(np.round(1-train_percentage, 2)*num_data/2):\n",
    "        index_test = np.append(index_test, np.array([int(i)]), axis=0)\n",
    "        counter[0] += 1\n",
    "\n",
    "    if y_test[i] == False and counter[1] < int(np.round(1-train_percentage, 2)*num_data/2):\n",
    "        index_test = np.append(index_test, np.array([int(i)]), axis=0)\n",
    "        counter[1] += 1\n",
    "\n",
    "x_train = x_train[index_train]\n",
    "y_train = y_train[index_train]\n",
    "x_test = x_test[index_test]\n",
    "y_test = y_test[index_test]\n",
    "\n",
    "print(x_train.shape, y_train.shape)\n",
    "print(x_test.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f9af6816-856c-4fc1-a9ab-03d3faab30c1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x500 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Print some images from the final dataset\n",
    "fig, axes = plt.subplots(2, 5, figsize=(10, 5))\n",
    "\n",
    "for i in range(2):\n",
    "    for j in range(5):\n",
    "        axes[i, j].imshow(x_train[i + j].squeeze(), cmap='gray') #imshow() is for images, and cmap grey i think interprets 0-1\n",
    "        # the .squeeze() removes and dimensions of 1, so the extra dimension we added?\n",
    "        axes[i, j].set_title(\"Label: {}\".format(y_train[i + j]))\n",
    "        axes[i, j].axis('off')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8dad08b0-a3a7-44f1-b2f5-996292252afa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x20006c9faf0>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot an example of 1 image\n",
    "\n",
    "print(y_train[0])\n",
    "\n",
    "plt.imshow(x_train[0, :, :, 0]) #this is 1st image [0, data :,:, and then removing last dim 0]\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b9170ef0-c8f2-476d-9094-e1346d4c0762",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(80, 16, 16, 1)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "80"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# resize the images\n",
    "\n",
    "x_train_small = tf.image.resize(x_train, (16,16)).numpy() #tf is tensor flow, which I imagine is the resizing method?\n",
    "x_test_small = tf.image.resize(x_test, (16,16)).numpy() #.numpy() puts into np array?\n",
    "print(x_train_small.shape)\n",
    "len(x_train_small)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "cae94f70-fbc0-497b-b296-5c9485918633",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x20006d325e0>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot an example of 1 resized image\n",
    "\n",
    "print(y_train[0])\n",
    "\n",
    "plt.imshow(x_train_small[0,:,:,0], vmin=0, vmax=1)\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "57dd93e9-bc71-47ec-8329-966df90f5c66",
   "metadata": {},
   "source": [
    "**Classical Principal Component Analysis**\n",
    "\n",
    "We will use the well-known scikit-learn package for the classical PCA (https://scikit-learn.org/stable/modules/generated/sklearn.decomposition.PCA.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "1d6e2484-3444-4eb7-886b-0ae8af713189",
   "metadata": {},
   "outputs": [],
   "source": [
    "# decide the number of principal components to be used\n",
    "#from sklearn.decomposition import PCA -- can do 2^n pixels, (16x16=256) above it says 2^n x 2^n but this doesn't seem right\n",
    "n_components_classic = 8\n",
    "\n",
    "pca = PCA(n_components=n_components_classic, whiten = False, random_state = 2020) #reduces dimensionality"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "9a5cc4e4-afdb-4b87-875e-5d583da65b9b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((80, 8), (20, 8))"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_small_pca = pca.fit_transform(x_train_small.reshape(len(x_train_small),16**2)) #reshapes to 80,256 then reduces to 80,8 using pca\n",
    "x_test_small_pca = pca.fit_transform(x_test_small.reshape(len(x_test_small),16**2))\n",
    "x_train_small_pca.shape, x_test_small_pca.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "7d0a487c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 2.149694   -1.0433712  -0.18994118 -1.1095574  -0.06876877  0.44443968\n",
      "  0.78277516  0.3613933 ]\n"
     ]
    }
   ],
   "source": [
    "print(x_train_small_pca[0]) #not normalised, hence next step, could probably wait until this point to normalise?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55ad742a-b5ec-43a7-82e3-7db3a9eb78f2",
   "metadata": {},
   "source": [
    "**Dataset Reformatting**\n",
    "\n",
    "VQC class from Qiskit need a specific format for the dataset. This section is dedicated to do that. \n",
    "You can read the documentation here (https://qiskit.org/documentation/stubs/qiskit.aqua.algorithms.VQC.html). THIS LINK DOESN'T WORK BUT IT'S A VQC SO SIMILAR TO THE VQC EXAMPLE WE HAVE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "cfc3803c-b0c2-4b2b-9b53-d1b197704334",
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalize(x):\n",
    "  return (x - x.min())/(x.max() - x.min())\n",
    "\n",
    "x_train_small_pca = normalize(x_train_small_pca)\n",
    "x_test_small_pca = normalize(x_test_small_pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "97c1ca14-3b3c-4c0b-a600-f4704bac3cf3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100, 8)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_total = np.concatenate((x_train_small_pca, x_test_small_pca))\n",
    "x_total.shape #so we could have just taken all data from test or train, just keeping convention of data set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d55f3282-5a09-4701-bd30-e590dc349c4a",
   "metadata": {},
   "source": [
    "**Variational Quantum Classifier**\n",
    "\n",
    "This is the final section and finally we can play with the VQC.\n",
    "\n",
    "We need to tune the VQC by trying different parameters until we get good results. The best parameters for VQC that is trained with QPCA dataset maybe (and by 'maybe' I mean high chance) different with the one that is trained with classical PCA. There are many good tutorials on VQC here https://github.com/qiskit-community/qiskit-community-tutorials/tree/master/machine_learning ."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2d3fe035-ae89-4925-907d-187feaefc413",
   "metadata": {},
   "source": [
    "**Data embedding circuits**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "91bf8ba3-7aa9-4814-b226-2f33a5c6c450",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\sebts\\AppData\\Roaming\\Python\\Python39\\site-packages\\qiskit\\visualization\\circuit\\matplotlib.py:266: FutureWarning: The default matplotlib drawer scheme will be changed to \"iqp\" in a following release. To silence this warning, specify the current default explicitly as style=\"clifford\", or the new default as style=\"iqp\".\n",
      "  self._style, def_font_ratio = load_style(self._style)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1547.39x1545.13 with 1 Axes>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_zz_feature_map(feature_dim, feature_map_rep, ent): #this is our generic feature map for 8 components (qubits)\n",
    "    return ZZFeatureMap(feature_dimension=feature_dim, reps=feature_map_rep, entanglement=ent)\n",
    "\n",
    "fmap_zz = get_zz_feature_map(n_components_classic, 2, 'linear') #can't remember what rep is? possibly just visual rep depth\n",
    "input_params = fmap_zz.parameters\n",
    "fmap_zz.decompose().draw('mpl', scale=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "b0083d62-786a-4c81-b56e-bcecdbe09771",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ParameterView([ParameterVectorElement(x[0]), ParameterVectorElement(x[1]), ParameterVectorElement(x[2]), ParameterVectorElement(x[3]), ParameterVectorElement(x[4]), ParameterVectorElement(x[5]), ParameterVectorElement(x[6]), ParameterVectorElement(x[7])])"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fmap_zz.parameters #8 parameters to encode the 8 features of our reduced dimensionality data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "39d55540-5b19-4532-8a82-9800d66c28ba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 962.109x491.633 with 1 Axes>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_two_locals(feature_dim, rotations, var_form_rep, ent): #sets up our general ansatz\n",
    "    return TwoLocal(num_qubits=feature_dim, rotation_blocks=rotations, entanglement_blocks='cx', entanglement=ent, reps=var_form_rep)\n",
    "\n",
    "ansatz_tl = get_two_locals(n_components_classic, ['ry', 'rz'], 2, 'linear') #do we just do ry and rz bc we have x in cx?\n",
    "weights_params = ansatz_tl.parameters\n",
    "ansatz_tl.decompose().draw('mpl', scale=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "e427580b-7d54-4de3-8b54-9b48509e5e08",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 962.109x491.633 with 1 Axes>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights = np.random.random(len(ansatz_tl.parameters)) #get random weights from 0 to 1 for each rotation\n",
    "ansatz_tl.assign_parameters({k:v for (k,v) in zip(ansatz_tl.parameters, weights)}).decompose().draw('mpl', scale=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "01145898-0d7f-460e-a0d6-9bf184ed32e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 845.053x491.633 with 1 Axes>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "var_circuit = fmap_zz.compose(ansatz_tl) #variational circuit, I am not sure if there is more to the circuit past two local?\n",
    "var_circuit.draw('mpl', scale=0.7)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ebe84b00-e757-4be5-9cd2-61cdc80818b5",
   "metadata": {},
   "source": [
    "**Primitives**\n",
    "\n",
    "For info, see https://docs.quantum.ibm.com/run/primitives"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "8c27a1e0-398c-46f5-a8dd-43e5b323eaf8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#it seems like there might be better options for this? am I accessing the online simulator?\n",
    "\n",
    "# for qiskit_ibm_runtime.Options, see https://docs.quantum.ibm.com/api/qiskit-ibm-runtime/qiskit_ibm_runtime.options.Options\n",
    "qasm_sim = AerSimulator()\n",
    "options = {}\n",
    "sampler = BackendSampler(backend=qasm_sim, options=options)\n",
    "estimator = BackendEstimator(backend=qasm_sim, options=options)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83d1b4ab-36d0-498b-b5b7-1a2effadeaee",
   "metadata": {},
   "source": [
    "**Quantum Neural Network**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c58bfb2c-3324-4950-8e46-182fef617abb",
   "metadata": {},
   "source": [
    "**SamplerQNN**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "658f518b-4eba-4c1a-bd86-78802048548f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#these are the loss (or cost?) fns from the example, I think parity might work for more classes? but we only have two so can use one qubit binary?\n",
    "\n",
    "def parity(x, num_classes):\n",
    "    return f\"{x:b}\".count(\"1\") % num_classes\n",
    "\n",
    "def one_qubit_binary(x):\n",
    "    return x % 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "87faaaee-ed57-4940-8bba-9dbce09c853e",
   "metadata": {},
   "outputs": [],
   "source": [
    "var_circuit_with_meas = var_circuit.copy()\n",
    "var_circuit_with_meas.measure_all() #adds classical measurement (in Z?) to the circuit, I think this would be done automatically?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "f24afc9a-3488-4561-aec5-d61c34bd8fd1",
   "metadata": {},
   "outputs": [],
   "source": [
    "#is this just to define the parameterised ciruit sampler as the AerSimulator?\n",
    "paramShiftSampGrad = ParamShiftSamplerGradient(sampler=sampler)\n",
    "\n",
    "sampler_qnn = SamplerQNN(sampler=sampler, #calling the sampler again\n",
    "                         circuit=var_circuit_with_meas, #this is our ciruit we built above\n",
    "                         input_params=input_params, #from fmap_zz.parameters, I think this is where the 8 points per data set are put in\n",
    "                                                           \n",
    "                         weight_params=weights_params, #this is from our two local (ansatz_tl.parameters)\n",
    "                                                           \n",
    "                         interpret=one_qubit_binary, #the one qubit binary loss fn (same = 0, dif = 1)\n",
    "                         output_shape=2, #gives us [a,b] output, not sure what this refers to? chance of 0 and 1?\n",
    "                         gradient=paramShiftSampGrad) #why are we putting the sampler in again?\n",
    "\n",
    "\n",
    "# if your embedding strategy is not a parametrized circuit\n",
    "# (e.g. amplitude encoding) you will have to do some extra work!\n",
    "# if you train your embedding as well, the trainable\n",
    "# parameters of the embedding must be added"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b70344b6-c769-497c-bec7-1e1dfce2d396",
   "metadata": {},
   "source": [
    "**Forward pass**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "5eebb0fd-61d0-4707-935c-67744eb02401",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output shape for 100 samples: (100, 2)\n",
      "Output of the forward pass for first sample: [0.66503906 0.33496094]\n"
     ]
    }
   ],
   "source": [
    "# The result of the neural network on a forward pass.\n",
    "output = sampler_qnn.forward(x_total, weights) #(100, 8) the full (reduced) x data set, \n",
    "#weights = np.random.random(len(ansatz_tl.parameters)) -- our random set of starting parameters in the two local\n",
    "print(f\"Output shape for {len(x_total)} samples: {output.shape}\")\n",
    "print(f\"Output of the forward pass for first sample: {output[0]}\") #i think this is chance of outputting 0 or 1, \n",
    "                                                                   #but I am not sure if this means correct/wrong or 0,1 the pictures\n",
    "                                                                   # I don't think this knows about y? so I think it's picture 0 or 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c911a33-66d7-4966-86a0-6151899f1c61",
   "metadata": {},
   "source": [
    "**Backward pass**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "0addf495-edd8-4d6e-aa9a-2d10f7c6f641",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output shape for 100 samples: (100, 2, 48)\n",
      "Output of the backward pass for first sample: [[ 0.15136719  0.12011719  0.00732422  0.00585938 -0.00048828  0.00195312\n",
      "   0.02246094  0.00585938  0.32519531  0.01171875 -0.00244141 -0.00732422\n",
      "  -0.0078125  -0.00976562 -0.0234375  -0.00585938 -0.07080078 -0.02148438\n",
      "   0.00097656  0.01757812 -0.01464844 -0.00976562  0.00048828  0.00830078\n",
      "   0.0078125   0.00439453 -0.01074219 -0.00439453 -0.01269531 -0.00390625\n",
      "  -0.01025391 -0.00244141 -0.01855469  0.00195312  0.00683594 -0.00195312\n",
      "  -0.00976562  0.00488281  0.01611328  0.00830078 -0.00390625  0.00048828\n",
      "   0.01074219 -0.0078125  -0.02929688 -0.00341797  0.00439453 -0.00244141]\n",
      " [-0.15136719 -0.12011719 -0.00732422 -0.00585938  0.00048828 -0.00195312\n",
      "  -0.02246094 -0.00585938 -0.32519531 -0.01171875  0.00244141  0.00732422\n",
      "   0.0078125   0.00976562  0.0234375   0.00585938  0.07080078  0.02148438\n",
      "  -0.00097656 -0.01757812  0.01464844  0.00976562 -0.00048828 -0.00830078\n",
      "  -0.0078125  -0.00439453  0.01074219  0.00439453  0.01269531  0.00390625\n",
      "   0.01025391  0.00244141  0.01855469 -0.00195312 -0.00683594  0.00195312\n",
      "   0.00976562 -0.00488281 -0.01611328 -0.00830078  0.00390625 -0.00048828\n",
      "  -0.01074219  0.0078125   0.02929688  0.00341797 -0.00439453  0.00244141]]\n"
     ]
    }
   ],
   "source": [
    "# The result of the neural network on a backward pass\n",
    "_, weights_grad = sampler_qnn.backward(x_total, weights) #same thing but backwards\n",
    "print(f\"Output shape for {len(x_total)} samples: {weights_grad.shape}\")\n",
    "print(f\"Output of the backward pass for first sample: {weights_grad[0]}\") \n",
    "#these weights are the parameter values for the two local?\n",
    "#but there's two sets of 48? (0 and 1 values?)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58524679-a163-4e8b-b500-738cb387258b",
   "metadata": {},
   "source": [
    "**EstimatorQNN**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "982a7d11-4371-41d1-8bd1-235777e65926",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parity observable: ZZZZZZZZ\n",
      "First qubit observable: IIIIIIIZ\n"
     ]
    }
   ],
   "source": [
    "obs_parity = Pauli('Z' * n_components_classic)\n",
    "obs_one_qubit = Pauli('I' * (n_components_classic - 1) + 'Z')\n",
    "print(f\"Parity observable: {obs_parity}\")\n",
    "print(f\"First qubit observable: {obs_one_qubit}\") #ahhh so these are our choices of observables to then turn into loss fn?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "2e90289c-89f7-4e3e-b5f6-3b1867a08f05",
   "metadata": {},
   "outputs": [],
   "source": [
    "paramShiftEstGrad = ParamShiftEstimatorGradient(estimator=estimator) #this time we're using the estimator rather than sampler, this just does the matrices?\n",
    "\n",
    "estimator_qnn = EstimatorQNN(estimator=estimator,\n",
    "                             circuit=var_circuit,\n",
    "                             observables=obs_one_qubit, #this time we're applying the observable (IIIIIIIZ) rather than 1 qubit binary\n",
    "                                                        # i'm not sure how the 1 qubit binary in the sampler is calculated then\n",
    "                                                        # possibly something to do with output shape?\n",
    "                             \n",
    "                             input_params=input_params,     # if your embedding strategy is not a parametrized circuit\n",
    "                                                            # (e.g. amplitude encoding) you will have to do some extra work!\n",
    "                             weight_params=weights_params,  # if you train your embedding as well, the trainable\n",
    "                                                            # parameters of the embedding must be added\n",
    "                             gradient=paramShiftEstGrad)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d3b6aef3-3b32-4546-ac25-c9afec639222",
   "metadata": {},
   "source": [
    "**Forward pass**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "268a6bfd-5949-4f77-becc-cf529b10feaf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output shape for 100 samples: (100, 1)\n",
      "Output of the forward pass for first sample: [0.35546875]\n"
     ]
    }
   ],
   "source": [
    "# The result of the neural network on a forward pass.\n",
    "output = estimator_qnn.forward(x_total, weights)\n",
    "print(f\"Output shape for {len(x_total)} samples: {output.shape}\")\n",
    "print(f\"Output of the forward pass for first sample: {output[0]}\") \n",
    "#this time we're simply outputting the expectation value of the observable IIIIIIIZ"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a662dad-1c7f-44e1-b02c-2f717e1402eb",
   "metadata": {},
   "source": [
    "**Backward pass**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "d74cec27-826c-4b99-afc2-e1288447c62f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output shape for 100 samples: (100, 1, 48)\n",
      "Output of the backward pass for first sample: [[ 0.33691406  0.23144531 -0.00195312  0.00195312 -0.0078125  -0.02832031\n",
      "  -0.00097656  0.00097656  0.64355469  0.01953125 -0.00976562  0.02246094\n",
      "  -0.01171875 -0.01464844  0.04101562  0.00195312 -0.13964844  0.00390625\n",
      "  -0.00097656 -0.00292969  0.00097656 -0.00878906  0.01953125 -0.0078125\n",
      "  -0.00195312  0.02636719  0.00878906 -0.03613281 -0.02539062 -0.00390625\n",
      "  -0.04492188 -0.00976562 -0.0234375   0.01367188 -0.00585938  0.01367188\n",
      "  -0.02734375  0.00683594  0.02636719  0.01367188 -0.00195312  0.01269531\n",
      "   0.00390625  0.00488281 -0.01660156  0.00683594 -0.00097656 -0.0546875 ]]\n"
     ]
    }
   ],
   "source": [
    "# The result of the neural network on a backward pass\n",
    "_, weights_grad = estimator_qnn.backward(x_total, weights)\n",
    "print(f\"Output shape for {len(x_total)} samples: {weights_grad.shape}\")\n",
    "print(f\"Output of the backward pass for first sample: {weights_grad[0]}\")\n",
    "# i think these are directly the parameter values for two local?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "85311c2a-080f-4f54-9ec1-f5be64551640",
   "metadata": {},
   "source": [
    "**Train a Neural Network Classifier**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "41f552d6-96d7-405b-82df-9db00bc44a3f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 1, 1, 1, 0, 1, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1,\n",
       "       1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1,\n",
       "       1, 1, 0, 1, 1, 1, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 1, 0,\n",
       "       1, 1, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0], dtype=uint8)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train #just our y data of if it's a 0 or 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "16f26f9c-0870-4e40-bb3d-d34663da6b1b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Label 1 converted to [0 1]\n",
      "Label 0 converted to [1 0]\n"
     ]
    }
   ],
   "source": [
    "def get_one_hot_encoding(y): #this converts the labels 0,1 to vectors |0>,|1> (maybe transpose)\n",
    "    unique_labels = np.unique(y, axis=0)\n",
    "    y_one_hot = [(np.eye(len(unique_labels), dtype=int)[np.where(unique_labels == y_i)]).reshape(len(unique_labels)) for y_i in y]\n",
    "\n",
    "    return np.array(y_one_hot)\n",
    "\n",
    "y_train_1h = get_one_hot_encoding(y_train)\n",
    "y_test_1h = get_one_hot_encoding(y_test)\n",
    "\n",
    "print(f\"Label {y_train[2]} converted to {y_train_1h[2]}\")\n",
    "print(f\"Label {y_train[0]} converted to {y_train_1h[0]}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "bcb66377-7752-4379-bea0-87ddc670fc12",
   "metadata": {},
   "outputs": [],
   "source": [
    "def spsa_callback(nfev, params, fval, stepsize, accepted=None): #this is just to pull out loss every 3 iterations\n",
    "    #doesn't use params or stepsize (i think it's hard set to 3 below)\n",
    "    \"\"\"\n",
    "    Can be used for SPSA and GradientDescent optimizers\n",
    "    nfev: the number of function evals\n",
    "    params: the current parameters\n",
    "    fval: the current function value\n",
    "    stepsize: size of the update step\n",
    "    accepted: whether the step was accepted (not used for )\n",
    "    \"\"\"\n",
    "    global spsa_loss_recorder\n",
    "\n",
    "    if (nfev % 3) == 0:\n",
    "        spsa_loss_recorder.append(fval)\n",
    "        # print(f\"SPSA Epoch {len(spsa_loss_recorder)}: {fval:.5f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "7c72ae40-d51d-4808-936f-d0c513ac4682",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(80, 8)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_small_pca.shape #previously had wrong x_train here, we want the reduced dimensionality set of 80"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "e57c4aab-5c4e-416d-9904-d61f4482eeb6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(80, 2)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_1h.shape #2 bc we've converted into one hot encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "8613b552-2fcb-46b4-9556-247937bee46e",
   "metadata": {},
   "outputs": [],
   "source": [
    "max_itr = 10\n",
    "spsa_opt = SPSA(maxiter=max_itr, callback=spsa_callback) # from qiskit.algorithms.optimizers\n",
    "spsa_loss_recorder = [] #empty list to append training loss into \n",
    "initial_point = np.random.random((len(ansatz_tl.parameters),)) #random starting parameter set for our twolocal ansatz\n",
    "\n",
    "vqc = NeuralNetworkClassifier(neural_network=sampler_qnn, #SamplerQNN(sampler=sampler, circuit=var_circuit_with_meas,input_params=input_params,\n",
    "                              # weight_params=weights_params, interpret=one_qubit_binary, output_shape=2,gradient=paramShiftSampGrad) \n",
    "                              # so outputs [a,b] vector, for a|0>+b|1> from 1st qubit?\n",
    "                              one_hot=True, #does loss as vectors\n",
    "                              optimizer=spsa_opt, #above\n",
    "                              initial_point=initial_point) #random twolocal starting point\n",
    "\n",
    "vqc = vqc.fit(x_train_small_pca, y_train_1h) #run, I think the .fit runs the training?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "fad41f82-3ddc-4fc5-ad5d-be1c1bfa6369",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Training loss')"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(spsa_loss_recorder) #plots fval, I'm confused about the % 3 bc this would imply only every third point?\n",
    "#maybe 3 per itteration? so 30 total?\n",
    "plt.xlabel(\"Number of epochs\")\n",
    "plt.title(\"Training loss\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "73bc199b-df0d-4fed-af6c-624d80cde86d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score on the train set 0.8375\n",
      "Score on the test set 0.9\n"
     ]
    }
   ],
   "source": [
    "score_train = vqc.score(x_train_small_pca, y_train_1h) #how well does it run on the training set\n",
    "score_test = vqc.score(x_test_small_pca, y_test_1h) #how well does it run on the test set (assesses how good in more general case)\n",
    "print(f'Score on the train set {score_train}')\n",
    "print(f'Score on the test set {score_test}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "d45bcd1f-cfdb-43a0-8571-59783ca1ae80",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_pred = vqc.predict(x_train_small_pca) #what's the predicted y outcome [0,1] / [1,0] for pca'ed training data\n",
    "y_test_pred = vqc.predict(x_test_small_pca) #same for test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "9f99d61e-c86b-4e74-8572-dae671990ddd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#corret answer mask\n",
    "train_pred_acc = np.all(y_train_pred == y_train_1h, axis=1)\n",
    "test_pred_acc = np.all(y_test_pred == y_test_1h, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "32082ac0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False,  True,  True,  True,  True, False,  True,  True,  True,\n",
       "        True,  True, False, False,  True,  True,  True,  True, False,\n",
       "        True,  True,  True,  True,  True, False,  True, False,  True,\n",
       "        True,  True,  True,  True,  True,  True, False,  True,  True,\n",
       "        True,  True,  True,  True,  True,  True, False,  True,  True,\n",
       "        True, False,  True,  True,  True,  True,  True,  True,  True,\n",
       "        True,  True,  True,  True,  True, False, False,  True,  True,\n",
       "        True,  True,  True,  True,  True,  True,  True,  True, False,\n",
       "        True,  True,  True, False, False,  True,  True,  True])"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_pred_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "a839a3e0-c5da-47b0-9cd5-e2018d328116",
   "metadata": {},
   "outputs": [],
   "source": [
    "#split x data by whether the trained algo will give correct answer\n",
    "x_all = np.concatenate((x_train_small_pca, x_test_small_pca))\n",
    "y_all= np.concatenate((y_train_1h, y_test_1h))\n",
    "y_pred_acc_all = np.concatenate((train_pred_acc, test_pred_acc))\n",
    "\n",
    "x_b = x_all[np.all(y_all == [1, 0], axis=1)]\n",
    "x_b_good = x_b[(y_pred_acc_all[np.all(y_all == [1, 0], axis=1)])]\n",
    "x_b_bad = x_b[np.logical_not(y_pred_acc_all[np.all(y_all == [1, 0], axis=1)])]\n",
    "\n",
    "x_r = x_all[np.all(y_all == [0, 1], axis=1)]\n",
    "x_r_good = x_r[(y_pred_acc_all[np.all(y_all == [0, 1], axis=1)])]\n",
    "x_r_bad = x_r[np.logical_not(y_pred_acc_all[np.all(y_all == [0, 1], axis=1)])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "1cb8ead1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(32, 8)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_b_good.shape #good 0's, all x data?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "3a7a37af",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.5717578 , 0.0197521 , 0.4566309 , 0.54267925, 0.38951346,\n",
       "       0.33407667, 0.17543308, 0.5658302 ], dtype=float32)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_b_good[0] #the 8 parameters for this data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "3153ca1e-cad7-4b5f-887e-13235186a3e2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x201097a7310>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot whether the algo gives the correct outcome for each x, by the first two variables in x\n",
    "plt.scatter(x_b_good[:,0], x_b_good[:,1], c='b', marker=\".\", label=\"Good class 0\")\n",
    "plt.scatter(x_b_bad[:,0], x_b_bad[:,1], c='b', marker=\"x\", label=\"Bad class 0\")\n",
    "plt.scatter(x_r_good[:,0], x_r_good[:,1], c='r', marker=\".\", label=\"Good class 1\")\n",
    "plt.scatter(x_r_bad[:,0], x_r_bad[:,1], c='r', marker=\"x\", label=\"Bad class 1\")\n",
    "\n",
    "plt.title(\"Mnist dataset classification\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "1669baea-d5a6-4c00-b141-0b7e385b6cac",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2004e417c70>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot whether the algo gives the correct outcome for each x, by 3rd and 4th variable in x\n",
    "plt.scatter(x_b_good[:,2], x_b_good[:,3], c='b', marker=\".\", label=\"Good class 0\")\n",
    "plt.scatter(x_b_bad[:,2], x_b_bad[:,3], c='b', marker=\"x\", label=\"Bad class 0\")\n",
    "plt.scatter(x_r_good[:,2], x_r_good[:,3], c='r', marker=\".\", label=\"Good class 1\")\n",
    "plt.scatter(x_r_bad[:,2], x_r_bad[:,3], c='r', marker=\"x\", label=\"Bad class 1\")\n",
    "\n",
    "plt.title(\"Mnist dataset classification\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4d07fc22-df36-44d2-81d4-660a6b045dfd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2469c6c8-36ce-4be3-b3ce-003e10992350",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}