{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "94dcd061",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-04-01 14:12:48.477745: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory\n",
      "2022-04-01 14:12:48.477765: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.\n"
     ]
    }
   ],
   "source": [
    "from library.generators import Autoencoder\n",
    "from library.dataset import *\n",
    "import random\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "30b8f7d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def mkVector(n):\n",
    "    return np.array([random.uniform(-1.0, 1.0) for _k in range(n)])\n",
    "\n",
    "u = 0.4 * np.array([1.0,  1.0, 1.0, 1.0])\n",
    "v = 0.4 * np.array([1.0, -1.0, 0.0, 0.0])\n",
    "\n",
    "def mkPVector(a):\n",
    "    return (a[0] * u) + (a[1] * v)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "644b3d2e",
   "metadata": {},
   "outputs": [],
   "source": [
    "d2 = np.array([mkVector(2) for n in range(1000)])\n",
    "d1 = [2.0, 2.0, 1.0, 1.0] + np.array([mkPVector(x) for x in d2])\n",
    "d0 = np.array([u, v])\n",
    "data = DataSet(d0, d1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3d97e25d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "encoder\n",
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_1 (InputLayer)        [(None, 4)]               0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 4)                 20        \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 2)                 10        \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 2)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 36\n",
      "Trainable params: 36\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "decoder\n",
      "Model: \"model_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_2 (InputLayer)        [(None, 2)]               0         \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 4)                 12        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 12\n",
      "Trainable params: 12\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "autoencoder\n",
      "Model: \"model_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_3 (InputLayer)        [(None, 4)]               0         \n",
      "                                                                 \n",
      " model (Functional)          (None, 2)                 36        \n",
      "                                                                 \n",
      " model_1 (Functional)        (None, 4)                 12        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 48\n",
      "Trainable params: 48\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-04-01 14:12:50.467745: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory\n",
      "2022-04-01 14:12:50.467774: W tensorflow/stream_executor/cuda/cuda_driver.cc:269] failed call to cuInit: UNKNOWN ERROR (303)\n",
      "2022-04-01 14:12:50.467797: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (sbi-klabautermann): /proc/driver/nvidia/version does not exist\n",
      "2022-04-01 14:12:50.467955: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
     ]
    }
   ],
   "source": [
    "ae = Autoencoder(4, 2, eps=0.1**6)\n",
    "ae.reset()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "02b99128",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "scaler: 4.128243404530178\n",
      "32/32 [==============================] - 0s 745us/step - loss: 0.0284\n",
      "32/32 [==============================] - 0s 731us/step - loss: 0.0150\n",
      "Loss: 0.028437774628400803 → 0.014993255026638508\n",
      "32/32 [==============================] - 0s 735us/step - loss: 0.0142\n",
      "Loss: 0.014993255026638508 → 0.014239023439586163\n",
      "32/32 [==============================] - 0s 785us/step - loss: 0.0139\n",
      "Loss: 0.014239023439586163 → 0.013911601155996323\n",
      "32/32 [==============================] - 0s 780us/step - loss: 0.0137\n",
      "Loss: 0.013911601155996323 → 0.0136729646474123\n",
      "32/32 [==============================] - 0s 861us/step - loss: 0.0135\n",
      "Loss: 0.0136729646474123 → 0.013468525372445583\n",
      "32/32 [==============================] - 0s 849us/step - loss: 0.0133\n",
      "Loss: 0.013468525372445583 → 0.013311195187270641\n",
      "32/32 [==============================] - 0s 817us/step - loss: 0.0132\n",
      "Loss: 0.013311195187270641 → 0.013190635479986668\n",
      "32/32 [==============================] - 0s 768us/step - loss: 0.0131\n",
      "Loss: 0.013190635479986668 → 0.013096977956593037\n",
      "32/32 [==============================] - 0s 745us/step - loss: 0.0130\n",
      "Loss: 0.013096977956593037 → 0.013028563000261784\n",
      "32/32 [==============================] - 0s 820us/step - loss: 0.0130\n",
      "Loss: 0.013028563000261784 → 0.012951376847922802\n",
      "32/32 [==============================] - 0s 836us/step - loss: 0.0129\n",
      "Loss: 0.012951376847922802 → 0.012917964719235897\n",
      "32/32 [==============================] - 0s 816us/step - loss: 0.0129\n",
      "Loss: 0.012917964719235897 → 0.012865377590060234\n",
      "32/32 [==============================] - 0s 838us/step - loss: 0.0128\n",
      "Loss: 0.012865377590060234 → 0.012804212048649788\n",
      "32/32 [==============================] - 0s 778us/step - loss: 0.0128\n",
      "Loss: 0.012804212048649788 → 0.012793153524398804\n",
      "32/32 [==============================] - 0s 866us/step - loss: 0.0128\n",
      "Loss: 0.012793153524398804 → 0.012774690985679626\n",
      "32/32 [==============================] - 0s 780us/step - loss: 0.0128\n",
      "Loss: 0.012774690985679626 → 0.01275850273668766\n",
      "32/32 [==============================] - 0s 767us/step - loss: 0.0127\n",
      "Loss: 0.01275850273668766 → 0.012729167938232422\n",
      "32/32 [==============================] - 0s 825us/step - loss: 0.0127\n",
      "Loss: 0.012729167938232422 → 0.012719701044261456\n",
      "32/32 [==============================] - 0s 765us/step - loss: 0.0127\n",
      "Loss: 0.012719701044261456 → 0.012731422670185566\n",
      "32/32 [==============================] - 0s 830us/step - loss: 0.0127\n",
      "Loss: 0.012731422670185566 → 0.012705178000032902\n",
      "32/32 [==============================] - 0s 798us/step - loss: 0.0127\n",
      "Loss: 0.012705178000032902 → 0.012730122543871403\n",
      "32/32 [==============================] - 0s 782us/step - loss: 0.0127\n",
      "Loss: 0.012730122543871403 → 0.01272022444754839\n",
      "32/32 [==============================] - 0s 873us/step - loss: 0.0127\n",
      "Loss: 0.01272022444754839 → 0.01273187156766653\n",
      "32/32 [==============================] - 0s 812us/step - loss: 0.0127\n",
      "Loss: 0.01273187156766653 → 0.01270010694861412\n",
      "32/32 [==============================] - 0s 804us/step - loss: 0.0127\n",
      "Loss: 0.01270010694861412 → 0.012693691067397594\n",
      "32/32 [==============================] - 0s 822us/step - loss: 0.0127\n",
      "Loss: 0.012693691067397594 → 0.012692615389823914\n",
      "32/32 [==============================] - 0s 852us/step - loss: 0.0127\n",
      "Loss: 0.012692615389823914 → 0.012691832147538662\n",
      "32/32 [==============================] - 0s 861us/step - loss: 0.0127\n",
      "Loss: 0.012691832147538662 → 0.012697849422693253\n",
      "32/32 [==============================] - 0s 881us/step - loss: 0.0127\n",
      "Loss: 0.012697849422693253 → 0.01271753292530775\n",
      "32/32 [==============================] - 0s 863us/step - loss: 0.0127\n",
      "Loss: 0.01271753292530775 → 0.012720879167318344\n",
      "32/32 [==============================] - 0s 879us/step - loss: 0.0127\n",
      "Loss: 0.012720879167318344 → 0.012691931799054146\n",
      "32/32 [==============================] - 0s 851us/step - loss: 0.0127\n",
      "Loss: 0.012691931799054146 → 0.012697676196694374\n",
      "32/32 [==============================] - 0s 848us/step - loss: 0.0127\n",
      "Loss: 0.012697676196694374 → 0.012715937569737434\n",
      "32/32 [==============================] - 0s 871us/step - loss: 0.0127\n",
      "Loss: 0.012715937569737434 → 0.012685921043157578\n",
      "32/32 [==============================] - 0s 873us/step - loss: 0.0127\n",
      "Loss: 0.012685921043157578 → 0.01268226932734251\n",
      "32/32 [==============================] - 0s 861us/step - loss: 0.0127\n",
      "Loss: 0.01268226932734251 → 0.012741411104798317\n",
      "32/32 [==============================] - 0s 833us/step - loss: 0.0127\n",
      "Loss: 0.012741411104798317 → 0.01268570777028799\n",
      "32/32 [==============================] - 0s 877us/step - loss: 0.0127\n",
      "Loss: 0.01268570777028799 → 0.012686950154602528\n",
      "32/32 [==============================] - 0s 841us/step - loss: 0.0127\n",
      "Loss: 0.012686950154602528 → 0.012699298560619354\n",
      "32/32 [==============================] - 0s 826us/step - loss: 0.0127\n",
      "Loss: 0.012699298560619354 → 0.012684759683907032\n",
      "32/32 [==============================] - 0s 801us/step - loss: 0.0127\n",
      "Loss: 0.012684759683907032 → 0.012687243521213531\n",
      "32/32 [==============================] - 0s 871us/step - loss: 0.0127\n",
      "Loss: 0.012687243521213531 → 0.012707813642919064\n",
      "32/32 [==============================] - 0s 849us/step - loss: 0.0127\n",
      "Loss: 0.012707813642919064 → 0.012719701044261456\n",
      "32/32 [==============================] - 0s 852us/step - loss: 0.0127\n",
      "Loss: 0.012719701044261456 → 0.012706177309155464\n",
      "32/32 [==============================] - 0s 819us/step - loss: 0.0127\n",
      "Loss: 0.012706177309155464 → 0.012707564048469067\n",
      "32/32 [==============================] - 0s 721us/step - loss: 0.0127\n",
      "Loss: 0.012707564048469067 → 0.012691184878349304\n",
      "32/32 [==============================] - 0s 751us/step - loss: 0.0127\n",
      "Loss: 0.012691184878349304 → 0.012684602290391922\n",
      "32/32 [==============================] - 0s 796us/step - loss: 0.0127\n",
      "Loss: 0.012684602290391922 → 0.012679639272391796\n",
      "32/32 [==============================] - 0s 796us/step - loss: 0.0127\n",
      "Loss: 0.012679639272391796 → 0.012698126956820488\n",
      "32/32 [==============================] - 0s 786us/step - loss: 0.0127\n",
      "Loss: 0.012698126956820488 → 0.012684212066233158\n",
      "32/32 [==============================] - 0s 783us/step - loss: 0.0127\n",
      "Loss: 0.012684212066233158 → 0.0126904072239995\n",
      "32/32 [==============================] - 0s 828us/step - loss: 0.0127\n",
      "Loss: 0.0126904072239995 → 0.012686283327639103\n",
      "32/32 [==============================] - 0s 775us/step - loss: 0.0127\n",
      "Loss: 0.012686283327639103 → 0.01268336083739996\n",
      "32/32 [==============================] - 0s 812us/step - loss: 0.0127\n",
      "Loss: 0.01268336083739996 → 0.012685326859354973\n",
      "32/32 [==============================] - 0s 760us/step - loss: 0.0127\n",
      "Loss: 0.012685326859354973 → 0.01270284503698349\n",
      "32/32 [==============================] - 0s 856us/step - loss: 0.0127\n",
      "Loss: 0.01270284503698349 → 0.012690501287579536\n",
      "32/32 [==============================] - 0s 875us/step - loss: 0.0127\n",
      "Loss: 0.012690501287579536 → 0.012686098925769329\n",
      "32/32 [==============================] - 0s 895us/step - loss: 0.0127\n",
      "Loss: 0.012686098925769329 → 0.01268165186047554\n",
      "32/32 [==============================] - 0s 874us/step - loss: 0.0127\n",
      "Loss: 0.01268165186047554 → 0.012708653695881367\n",
      "32/32 [==============================] - 0s 832us/step - loss: 0.0127\n",
      "Loss: 0.012708653695881367 → 0.01268619392067194\n",
      "32/32 [==============================] - 0s 739us/step - loss: 0.0127\n",
      "Loss: 0.01268619392067194 → 0.012690472416579723\n",
      "32/32 [==============================] - 0s 776us/step - loss: 0.0127\n",
      "Loss: 0.012690472416579723 → 0.012682026252150536\n",
      "32/32 [==============================] - 0s 878us/step - loss: 0.0127\n",
      "Loss: 0.012682026252150536 → 0.012682165950536728\n",
      "32/32 [==============================] - 0s 820us/step - loss: 0.0127\n",
      "Loss: 0.012682165950536728 → 0.012707279063761234\n",
      "32/32 [==============================] - 0s 834us/step - loss: 0.0127\n",
      "Loss: 0.012707279063761234 → 0.012696998193860054\n",
      "32/32 [==============================] - 0s 853us/step - loss: 0.0127\n",
      "Loss: 0.012696998193860054 → 0.012690282426774502\n",
      "32/32 [==============================] - 0s 884us/step - loss: 0.0127\n",
      "Loss: 0.012690282426774502 → 0.012691528536379337\n",
      "32/32 [==============================] - 0s 917us/step - loss: 0.0127\n",
      "Loss: 0.012691528536379337 → 0.012698102742433548\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "32/32 [==============================] - 0s 867us/step - loss: 0.0127\n",
      "Loss: 0.012698102742433548 → 0.012690417468547821\n",
      "32/32 [==============================] - 0s 852us/step - loss: 0.0127\n",
      "Loss: 0.012690417468547821 → 0.012707076966762543\n",
      "32/32 [==============================] - 0s 917us/step - loss: 0.0127\n",
      "Loss: 0.012707076966762543 → 0.012682667933404446\n",
      "32/32 [==============================] - 0s 844us/step - loss: 0.0127\n",
      "Loss: 0.012682667933404446 → 0.01268991082906723\n",
      "32/32 [==============================] - 0s 840us/step - loss: 0.0127\n",
      "Loss: 0.01268991082906723 → 0.012703971937298775\n",
      "32/32 [==============================] - 0s 878us/step - loss: 0.0127\n",
      "Loss: 0.012703971937298775 → 0.012693480588495731\n",
      "32/32 [==============================] - 0s 839us/step - loss: 0.0127\n",
      "Loss: 0.012693480588495731 → 0.012680095620453358\n",
      "32/32 [==============================] - 0s 903us/step - loss: 0.0127\n",
      "Loss: 0.012680095620453358 → 0.012688150629401207\n",
      "32/32 [==============================] - 0s 791us/step - loss: 0.0127\n",
      "Loss: 0.012688150629401207 → 0.012692077085375786\n",
      "32/32 [==============================] - 0s 803us/step - loss: 0.0127\n",
      "Loss: 0.012692077085375786 → 0.012682580389082432\n",
      "32/32 [==============================] - 0s 819us/step - loss: 0.0127\n",
      "Loss: 0.012682580389082432 → 0.012701192870736122\n",
      "32/32 [==============================] - 0s 798us/step - loss: 0.0127\n",
      "Loss: 0.012701192870736122 → 0.01269218698143959\n",
      "32/32 [==============================] - 0s 784us/step - loss: 0.0127\n",
      "Loss: 0.01269218698143959 → 0.012690702453255653\n",
      "32/32 [==============================] - 0s 792us/step - loss: 0.0127\n",
      "Loss: 0.012690702453255653 → 0.012688483111560345\n",
      "32/32 [==============================] - 0s 820us/step - loss: 0.0127\n",
      "Loss: 0.012688483111560345 → 0.012696083635091782\n",
      "32/32 [==============================] - 0s 818us/step - loss: 0.0127\n",
      "Loss: 0.012696083635091782 → 0.012682049535214901\n",
      "32/32 [==============================] - 0s 863us/step - loss: 0.0127\n",
      "Loss: 0.012682049535214901 → 0.012689575552940369\n",
      "32/32 [==============================] - 0s 902us/step - loss: 0.0127\n",
      "Loss: 0.012689575552940369 → 0.012679762206971645\n",
      "32/32 [==============================] - 0s 839us/step - loss: 0.0127\n",
      "Loss: 0.012679762206971645 → 0.01268184743821621\n",
      "32/32 [==============================] - 0s 852us/step - loss: 0.0127\n",
      "Loss: 0.01268184743821621 → 0.012686855159699917\n",
      "32/32 [==============================] - 0s 898us/step - loss: 0.0127\n",
      "Loss: 0.012686855159699917 → 0.01270339172333479\n",
      "32/32 [==============================] - 0s 851us/step - loss: 0.0127\n",
      "Loss: 0.01270339172333479 → 0.012698620557785034\n",
      "32/32 [==============================] - 0s 791us/step - loss: 0.0127\n",
      "Loss: 0.012698620557785034 → 0.012685850262641907\n",
      "32/32 [==============================] - 0s 725us/step - loss: 0.0127\n",
      "Loss: 0.012685850262641907 → 0.012688320130109787\n",
      "32/32 [==============================] - 0s 764us/step - loss: 0.0127\n",
      "Loss: 0.012688320130109787 → 0.012695953249931335\n",
      "32/32 [==============================] - 0s 811us/step - loss: 0.0127\n",
      "Loss: 0.012695953249931335 → 0.012689359486103058\n",
      "32/32 [==============================] - 0s 806us/step - loss: 0.0127\n",
      "Loss: 0.012689359486103058 → 0.012678883038461208\n",
      "32/32 [==============================] - 0s 808us/step - loss: 0.0127\n",
      "Loss: 0.012678883038461208 → 0.012700718827545643\n",
      "32/32 [==============================] - 0s 845us/step - loss: 0.0127\n",
      "Loss: 0.012700718827545643 → 0.01269646268337965\n",
      "32/32 [==============================] - 0s 834us/step - loss: 0.0127\n",
      "Loss: 0.01269646268337965 → 0.012700562365353107\n",
      "32/32 [==============================] - 0s 797us/step - loss: 0.0127\n",
      "Loss: 0.012700562365353107 → 0.012682804837822914\n",
      "32/32 [==============================] - 0s 723us/step - loss: 0.0127\n",
      "Loss: 0.012682804837822914 → 0.012677977792918682\n",
      "32/32 [==============================] - 0s 772us/step - loss: 0.0127\n",
      "Loss: 0.012677977792918682 → 0.012681444175541401\n",
      "32/32 [==============================] - 0s 796us/step - loss: 0.0127\n",
      "Loss: 0.012681444175541401 → 0.012686187401413918\n",
      "32/32 [==============================] - 0s 801us/step - loss: 0.0127\n",
      "Loss: 0.012686187401413918 → 0.012681050226092339\n",
      "32/32 [==============================] - 0s 843us/step - loss: 0.0127\n",
      "Loss: 0.012681050226092339 → 0.012674792669713497\n",
      "32/32 [==============================] - 0s 843us/step - loss: 0.0127\n",
      "Loss: 0.012674792669713497 → 0.012680117040872574\n",
      "32/32 [==============================] - 0s 782us/step - loss: 0.0127\n",
      "Loss: 0.012680117040872574 → 0.012677757069468498\n",
      "32/32 [==============================] - 0s 794us/step - loss: 0.0127\n",
      "Loss: 0.012677757069468498 → 0.012689709663391113\n",
      "32/32 [==============================] - 0s 738us/step - loss: 0.0127\n",
      "Loss: 0.012689709663391113 → 0.012686925008893013\n",
      "32/32 [==============================] - 0s 868us/step - loss: 0.0127\n",
      "Loss: 0.012686925008893013 → 0.012681815773248672\n",
      "32/32 [==============================] - 0s 819us/step - loss: 0.0127\n",
      "Loss: 0.012681815773248672 → 0.012678584083914757\n",
      "32/32 [==============================] - 0s 813us/step - loss: 0.0127\n",
      "Loss: 0.012678584083914757 → 0.012691756710410118\n",
      "32/32 [==============================] - 0s 831us/step - loss: 0.0127\n",
      "Loss: 0.012691756710410118 → 0.012686770409345627\n",
      "32/32 [==============================] - 0s 852us/step - loss: 0.0127\n",
      "Loss: 0.012686770409345627 → 0.012677742168307304\n",
      "32/32 [==============================] - 0s 984us/step - loss: 0.0127\n",
      "Loss: 0.012677742168307304 → 0.012687593698501587\n",
      "32/32 [==============================] - 0s 797us/step - loss: 0.0127\n",
      "Loss: 0.012687593698501587 → 0.012696138583123684\n",
      "32/32 [==============================] - 0s 730us/step - loss: 0.0127\n",
      "Loss: 0.012696138583123684 → 0.012677235528826714\n",
      "32/32 [==============================] - 0s 788us/step - loss: 0.0127\n",
      "Loss: 0.012677235528826714 → 0.012680829502642155\n",
      "32/32 [==============================] - 0s 801us/step - loss: 0.0127\n",
      "Loss: 0.012680829502642155 → 0.01268478762358427\n",
      "32/32 [==============================] - 0s 802us/step - loss: 0.0127\n",
      "Loss: 0.01268478762358427 → 0.012682672590017319\n",
      "32/32 [==============================] - 0s 782us/step - loss: 0.0127\n",
      "Loss: 0.012682672590017319 → 0.012682433240115643\n",
      "32/32 [==============================] - 0s 786us/step - loss: 0.0127\n",
      "Loss: 0.012682433240115643 → 0.012681093066930771\n",
      "32/32 [==============================] - 0s 801us/step - loss: 0.0127\n",
      "Loss: 0.012681093066930771 → 0.01269302424043417\n",
      "32/32 [==============================] - 0s 742us/step - loss: 0.0127\n",
      "Loss: 0.01269302424043417 → 0.012691360898315907\n",
      "32/32 [==============================] - 0s 788us/step - loss: 0.0127\n",
      "Loss: 0.012691360898315907 → 0.012699367478489876\n",
      "32/32 [==============================] - 0s 866us/step - loss: 0.0127\n",
      "Loss: 0.012699367478489876 → 0.012684808112680912\n",
      "32/32 [==============================] - 0s 820us/step - loss: 0.0127\n",
      "Loss: 0.012684808112680912 → 0.01268677320331335\n",
      "32/32 [==============================] - 0s 769us/step - loss: 0.0127\n",
      "Loss: 0.01268677320331335 → 0.012688460759818554\n",
      "32/32 [==============================] - 0s 832us/step - loss: 0.0127\n",
      "Loss: 0.012688460759818554 → 0.012684906832873821\n",
      "32/32 [==============================] - 0s 853us/step - loss: 0.0127\n",
      "Loss: 0.012684906832873821 → 0.012681410647928715\n",
      "32/32 [==============================] - 0s 892us/step - loss: 0.0127\n",
      "Loss: 0.012681410647928715 → 0.012687462382018566\n",
      "32/32 [==============================] - 0s 879us/step - loss: 0.0127\n",
      "Loss: 0.012687462382018566 → 0.012688900344073772\n",
      "32/32 [==============================] - 0s 819us/step - loss: 0.0127\n",
      "Loss: 0.012688900344073772 → 0.012693721801042557\n",
      "32/32 [==============================] - 0s 846us/step - loss: 0.0127\n",
      "Loss: 0.012693721801042557 → 0.012676053680479527\n",
      "32/32 [==============================] - 0s 774us/step - loss: 0.0127\n",
      "Loss: 0.012676053680479527 → 0.012686432339251041\n",
      "32/32 [==============================] - 0s 797us/step - loss: 0.0127\n",
      "Loss: 0.012686432339251041 → 0.012684906832873821\n",
      "32/32 [==============================] - 0s 834us/step - loss: 0.0127\n",
      "Loss: 0.012684906832873821 → 0.012699801474809647\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "32/32 [==============================] - 0s 787us/step - loss: 0.0127\n",
      "Loss: 0.012699801474809647 → 0.012678155675530434\n",
      "32/32 [==============================] - 0s 870us/step - loss: 0.0127\n",
      "Loss: 0.012678155675530434 → 0.012682000175118446\n",
      "32/32 [==============================] - 0s 888us/step - loss: 0.0127\n",
      "Loss: 0.012682000175118446 → 0.01269276812672615\n",
      "32/32 [==============================] - 0s 910us/step - loss: 0.0127\n",
      "Loss: 0.01269276812672615 → 0.012681564316153526\n",
      "32/32 [==============================] - 0s 870us/step - loss: 0.0127\n",
      "Loss: 0.012681564316153526 → 0.012699519284069538\n",
      "32/32 [==============================] - 0s 864us/step - loss: 0.0127\n",
      "Loss: 0.012699519284069538 → 0.012680163607001305\n",
      "32/32 [==============================] - 0s 747us/step - loss: 0.0127\n",
      "Loss: 0.012680163607001305 → 0.012685640715062618\n",
      "32/32 [==============================] - 0s 732us/step - loss: 0.0127\n",
      "Loss: 0.012685640715062618 → 0.012675425037741661\n",
      "32/32 [==============================] - 0s 786us/step - loss: 0.0127\n",
      "Loss: 0.012675425037741661 → 0.012680189684033394\n",
      "32/32 [==============================] - 0s 797us/step - loss: 0.0127\n",
      "Loss: 0.012680189684033394 → 0.012681961990892887\n",
      "32/32 [==============================] - 0s 860us/step - loss: 0.0127\n",
      "Loss: 0.012681961990892887 → 0.01268579438328743\n",
      "32/32 [==============================] - 0s 856us/step - loss: 0.0127\n",
      "Loss: 0.01268579438328743 → 0.012698291800916195\n",
      "32/32 [==============================] - 0s 808us/step - loss: 0.0127\n",
      "Loss: 0.012698291800916195 → 0.012691364623606205\n",
      "32/32 [==============================] - 0s 752us/step - loss: 0.0127\n",
      "Loss: 0.012691364623606205 → 0.012697343714535236\n",
      "32/32 [==============================] - 0s 801us/step - loss: 0.0127\n",
      "Loss: 0.012697343714535236 → 0.01269889809191227\n",
      "32/32 [==============================] - 0s 996us/step - loss: 0.0127\n",
      "Loss: 0.01269889809191227 → 0.012682449072599411\n",
      "32/32 [==============================] - 0s 848us/step - loss: 0.0127\n",
      "Loss: 0.012682449072599411 → 0.012706371955573559\n",
      "32/32 [==============================] - 0s 748us/step - loss: 0.0127\n",
      "Loss: 0.012706371955573559 → 0.012693656608462334\n",
      "32/32 [==============================] - 0s 757us/step - loss: 0.0127\n",
      "Loss: 0.012693656608462334 → 0.012678011320531368\n",
      "32/32 [==============================] - 0s 783us/step - loss: 0.0127\n",
      "Loss: 0.012678011320531368 → 0.012677457183599472\n",
      "32/32 [==============================] - 0s 772us/step - loss: 0.0127\n",
      "Loss: 0.012677457183599472 → 0.012680075131356716\n",
      "32/32 [==============================] - 0s 815us/step - loss: 0.0127\n",
      "Loss: 0.012680075131356716 → 0.012684504501521587\n",
      "32/32 [==============================] - 0s 786us/step - loss: 0.0127\n",
      "Loss: 0.012684504501521587 → 0.012679524719715118\n",
      "32/32 [==============================] - 0s 779us/step - loss: 0.0127\n",
      "Loss: 0.012679524719715118 → 0.012676456943154335\n",
      "32/32 [==============================] - 0s 832us/step - loss: 0.0127\n",
      "Loss: 0.012676456943154335 → 0.012684903107583523\n",
      "32/32 [==============================] - 0s 822us/step - loss: 0.0127\n",
      "Loss: 0.012684903107583523 → 0.01268160529434681\n",
      "32/32 [==============================] - 0s 765us/step - loss: 0.0127\n",
      "Loss: 0.01268160529434681 → 0.012683557346463203\n",
      "32/32 [==============================] - 0s 838us/step - loss: 0.0127\n",
      "Loss: 0.012683557346463203 → 0.012680706568062305\n",
      "32/32 [==============================] - 0s 889us/step - loss: 0.0127\n",
      "Loss: 0.012680706568062305 → 0.012679923325777054\n",
      "32/32 [==============================] - 0s 900us/step - loss: 0.0127\n",
      "Loss: 0.012679923325777054 → 0.012672575190663338\n",
      "32/32 [==============================] - 0s 887us/step - loss: 0.0127\n",
      "Loss: 0.012672575190663338 → 0.012684354558587074\n",
      "32/32 [==============================] - 0s 866us/step - loss: 0.0127\n",
      "Loss: 0.012684354558587074 → 0.012690423987805843\n",
      "32/32 [==============================] - 0s 854us/step - loss: 0.0127\n",
      "Loss: 0.012690423987805843 → 0.01267862506210804\n",
      "32/32 [==============================] - 0s 841us/step - loss: 0.0127\n",
      "Loss: 0.01267862506210804 → 0.012680350802838802\n",
      "32/32 [==============================] - 0s 879us/step - loss: 0.0127\n",
      "Loss: 0.012680350802838802 → 0.012677581049501896\n",
      "32/32 [==============================] - 0s 835us/step - loss: 0.0127\n",
      "Loss: 0.012677581049501896 → 0.012681134976446629\n",
      "32/32 [==============================] - 0s 750us/step - loss: 0.0127\n",
      "Loss: 0.012681134976446629 → 0.012686280533671379\n",
      "32/32 [==============================] - 0s 778us/step - loss: 0.0127\n",
      "Loss: 0.012686280533671379 → 0.012688659131526947\n",
      "32/32 [==============================] - 0s 800us/step - loss: 0.0127\n",
      "Loss: 0.012688659131526947 → 0.01267960574477911\n",
      "32/32 [==============================] - 0s 790us/step - loss: 0.0127\n",
      "Loss: 0.01267960574477911 → 0.012677302584052086\n",
      "32/32 [==============================] - 0s 846us/step - loss: 0.0127\n",
      "Loss: 0.012677302584052086 → 0.012692135758697987\n",
      "32/32 [==============================] - 0s 745us/step - loss: 0.0127\n",
      "Loss: 0.012692135758697987 → 0.012680561281740665\n",
      "32/32 [==============================] - 0s 767us/step - loss: 0.0127\n",
      "Loss: 0.012680561281740665 → 0.012683815322816372\n",
      "32/32 [==============================] - 0s 769us/step - loss: 0.0127\n",
      "Loss: 0.012683815322816372 → 0.012683389708399773\n",
      "32/32 [==============================] - 0s 877us/step - loss: 0.0127\n",
      "Loss: 0.012683389708399773 → 0.012680170126259327\n",
      "32/32 [==============================] - 0s 835us/step - loss: 0.0127\n",
      "Loss: 0.012680170126259327 → 0.012686174362897873\n",
      "32/32 [==============================] - 0s 756us/step - loss: 0.0127\n",
      "Loss: 0.012686174362897873 → 0.012676029466092587\n",
      "32/32 [==============================] - 0s 774us/step - loss: 0.0127\n",
      "Loss: 0.012676029466092587 → 0.012686045840382576\n",
      "32/32 [==============================] - 0s 763us/step - loss: 0.0127\n",
      "Loss: 0.012686045840382576 → 0.012685501947999\n",
      "32/32 [==============================] - 0s 808us/step - loss: 0.0127\n",
      "Loss: 0.012685501947999 → 0.01267742458730936\n",
      "32/32 [==============================] - 0s 845us/step - loss: 0.0127\n",
      "Loss: 0.01267742458730936 → 0.012682407163083553\n",
      "32/32 [==============================] - 0s 791us/step - loss: 0.0127\n",
      "Loss: 0.012682407163083553 → 0.012688462622463703\n",
      "32/32 [==============================] - 0s 820us/step - loss: 0.0127\n",
      "Loss: 0.012688462622463703 → 0.012681788764894009\n",
      "32/32 [==============================] - 0s 926us/step - loss: 0.0127\n",
      "Loss: 0.012681788764894009 → 0.012680851854383945\n",
      "32/32 [==============================] - 0s 792us/step - loss: 0.0127\n",
      "Loss: 0.012680851854383945 → 0.012701311148703098\n",
      "32/32 [==============================] - 0s 817us/step - loss: 0.0127\n",
      "Loss: 0.012701311148703098 → 0.012691560201346874\n",
      "32/32 [==============================] - 0s 830us/step - loss: 0.0127\n",
      "Loss: 0.012691560201346874 → 0.012689854018390179\n",
      "32/32 [==============================] - 0s 777us/step - loss: 0.0127\n",
      "Loss: 0.012689854018390179 → 0.01268122997134924\n",
      "32/32 [==============================] - 0s 794us/step - loss: 0.0127\n",
      "Loss: 0.01268122997134924 → 0.01268131472170353\n",
      "converged in 195 rounds\n"
     ]
    }
   ],
   "source": [
    "ae.train(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6fe3f3ef",
   "metadata": {},
   "outputs": [],
   "source": [
    "p = ae.encoder.predict(d1)\n",
    "e = ae.autoencoder.predict(d1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "b3adea8c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter([x[0] for x in d2], [x[1] for x in d2])\n",
    "plt.show()\n",
    "\n",
    "plt.scatter([x[0] for x in p], [x[1] for x in p])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "45e4b1f5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter([x[0] for x in d1], [x[1] for x in d1])\n",
    "plt.show()\n",
    "\n",
    "plt.scatter([x[0] for x in e], [x[1] for x in e])\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}