warpage_prediction.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "warpage_prediction.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/lorransr/warpage_analysis/blob/master/warpage_prediction.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UO-GDNQVsn-2",
        "colab_type": "text"
      },
      "source": [
        "#Prevendo rechupe\n",
        "\n",
        "## Descrição dos Dados\n",
        "Os dados utilizado neste trabalho são baseados no artigo de [Taghizadeh](http://jps.shirazu.ac.ir/article_1744_b7eb8ed50d88a80fe3a7ade4b30059fb.pdf), que buscou prever o rechupe em uma peça circular de plástico, concebida pelo processo de injeção sob diversos parâmetros. Para obtenção dos dados Taghizadeh, utilizou o Moldflow Plastics Insight (MPI).\n",
        "A princípio foi concebido o desenho 3D de uma peça com geometria similar a calços, espaçadores automotivos, transmissão e geração de energia. A peça em questão foi escolhida, justamente, por conta da importância que a sua forma final em relação ao seu desempenho. O plástico utilizado na simulação foi o Polietileno de baixa densidade (LDPE), com as seguintes propriedades:\n",
        "\n",
        "* Temperatura de Injeção (°C): 180~280\n",
        "* Temperatura do Molde (°C): 20~70\n",
        "* Condutividade Térmica (W/m°C): 0.31\n",
        "* Calor Específico(J/kg°C): 3400\n",
        "\n",
        "O modelo foi importado para o MPI e em seguida foi aplicada uma malha, antes de se realizar a análise. Foi escolhida uma malha de elementos triangulares. O tipo de malha foi a malha para superfícies fundidas. De maneira que ao fim do processo foram obtidos 1174 nós, com 2204 elementos triangulares, e com Match ratio (**entender oq é**) de 97%. \n",
        "Vale ressaltar que foram também simulados canais de resfriamento situados a 15 mm da superfície da peça e com distâncias entre canais de 55mm.\n",
        "\n",
        "Ao todo foram selecionados para a simulação cinco níveis de temperatura de injeção (280,255,230,205 e 180°C), três níveis da pressão de recalque(90,75 e 60% da pressão máxima de injeção), três níveis de  temperatura do molde (70,45 e 20°C). A partir desses valores foi criado um conjunto de dados de treino, com 45 configurações, utilizando-se o método fatorial de design de experimento. A numeração de cada teste foi realizada de acordo com o nível de cada parâmetro. Foi convencionado o  número 1 para o nível mais baixo, 2 para o segundo nível e assim sucessivamente até o nível mais alto. De tal forma que para o teste T-523, o digito 4 indica o quarto nível da temperatura de injeção(280), 2 indica o segundo nível da pressão de recalque (75% da pressão de injeção), e 3 indica o terceiro nível da temperatura do molde (70).Em seguida foi simulado também um conjunto de teste contendo 10 amostras com valores aleatórios de parâmetros dentro das faixas recomendadas."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ty2qH-PZuexL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import altair as alt\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ms3tvuRHtHrd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Importando os dados\n",
        "url = \"https://raw.githubusercontent.com/lorransr/warpage_analysis/master/warpage.csv\"\n",
        "raw = pd.read_csv(url,sep = \";\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q0Turn2fu0lO",
        "colab_type": "code",
        "outputId": "e40823ae-fb2e-4658-b2be-e407fe5e67af",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "raw.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>teste</th>\n",
              "      <th>melt_temp</th>\n",
              "      <th>hold_press</th>\n",
              "      <th>mold_temp</th>\n",
              "      <th>warpage</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>T-111</td>\n",
              "      <td>180</td>\n",
              "      <td>60</td>\n",
              "      <td>20</td>\n",
              "      <td>1.328</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>T-112</td>\n",
              "      <td>180</td>\n",
              "      <td>60</td>\n",
              "      <td>45</td>\n",
              "      <td>1.328</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>T-113</td>\n",
              "      <td>180</td>\n",
              "      <td>60</td>\n",
              "      <td>70</td>\n",
              "      <td>1.310</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>T-121</td>\n",
              "      <td>180</td>\n",
              "      <td>75</td>\n",
              "      <td>20</td>\n",
              "      <td>0.981</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>T-122</td>\n",
              "      <td>180</td>\n",
              "      <td>75</td>\n",
              "      <td>45</td>\n",
              "      <td>0.989</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   teste  melt_temp  hold_press  mold_temp  warpage\n",
              "0  T-111        180          60         20    1.328\n",
              "1  T-112        180          60         45    1.328\n",
              "2  T-113        180          60         70    1.310\n",
              "3  T-121        180          75         20    0.981\n",
              "4  T-122        180          75         45    0.989"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kUZ3Ou6JwJeA",
        "colab_type": "code",
        "outputId": "cf4f477d-5981-4e6f-ecde-1d2279da176c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "raw.dtypes"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "teste          object\n",
              "melt_temp       int64\n",
              "hold_press      int64\n",
              "mold_temp       int64\n",
              "warpage       float64\n",
              "dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KdNJ9Dd86KuO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df = raw"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CqkxqXVpKVUW",
        "colab_type": "text"
      },
      "source": [
        "## Analise Exploratória"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oTFT6Upa6MtR",
        "colab_type": "code",
        "outputId": "1a80da82-894a-47f1-f2a8-ca69227e536c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 388
        }
      },
      "source": [
        "g = sns.pairplot(df,x_vars = df.columns[1:4], y_vars = \"warpage\", height = 5)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABCwAAAFgCAYAAABwsOggAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3X2cnGV97/Hvb3Z3ksmymHVJUsuC\nQQqxKV2RrLS61qK0Fi1K09AqilFrgxgfOedQtB7E1sPrGNMeejw1RmIRIhZLwSjHYxGP4kODqBuF\nCMEoCsoGmyx7Nrpslp3szu/8Mfess5t9mM3M/Tif9+uVV3Zn79m5rr3v+3dd85vrwdxdAAAAAAAA\nSZKLuwAAAAAAAAAzkbAAAAAAAACJQ8ICAAAAAAAkDgkLAAAAAACQOCQsAAAAAABA4pCwAAAAAAAA\niUPCAgAAAAAAJA4JCwAAAAAAkDgkLAAAAAAAQOK0xl2Axbrgggv8zjvvjLsYABAHi+qFiLUAmhix\nFgDCV1OsTd0IiyeeeCLuIgBA5hFrASB8xFoAmF/qEhYAAAAAACD7SFgAAAAAAIDEIWEBAAAAAAAS\nh4QFAAAAAABIHBIWAAAAAAAgcUhYAAAAAACAxCFhAQAAAAAAEoeEBQAAAAAASBwSFgAAAAAAIHFI\nWAAAAAAAgMQhYQEAAAAAABKHhAUAAAAAAEic1rgLAADNolRyDY0WVZyYVL61RV3teeVyFnexGiLL\ndQMAIEtos5EmJCwAIAKlkmv/wRFt2tmvgeExdXcWtGNjr9as6kh9JyHLdQMAIEtos5E2TAkBgAgM\njRanOgeSNDA8pk07+zU0Woy5ZPXLct0AAMgS2mykDSMsACACxYlJrThhia6+cK2WF9p0eOyotn/1\nxypOTMZdtLpluW4AAGRJcWJyKllRMTA8RpuNRYlyWhEJCwCIQCHfor+6YI2uvG3v1BDMrRf3qJBv\nibtodcty3QAAyJJ8a4u6OwvTkhbdnQXlW2mzUZuopxUxJQQAIjBR8qk39FL504wrb9uriZLHXLL6\nZbluAJAkpZJrcGRcB4aPaHBkXCXiLBapqz2vHRt71d1ZkKSpN5td7fmYS4a0iHpaESMsACACRydK\nsw7BPDpRiqlEjZPlugFAUrBYIhohlzOtWdWhXZv72CUExyXqaUWMsACACJjZ1KcZFd2dBZmlv4OQ\n5boBQFKwWCIaJZczrehYopM7l2lFxxKSFViUyrSiamFOKyJhAQARaDFpy4aeaUMwt2zoUUsG+ghZ\nrhsAJAWLJQJIgqinFTElBAAikMvldNM9j0zbSeOmex7Rtet74i5a3bJcNwBIChZLBJAEUU8rCi1h\nYWY3SLpQ0iF3P2uOY86T9A+S2iQ94e6/H1Z5ACBOXe15XfGHa46Ze5yFRa6yXDcASIrKp5rEWgBx\nq0wrioK5h7O6sJm9SNKTknbOlrAws+WS7pF0gbv/zMxWuvuhhX5vb2+v9/f3N77AABCyBuxZHdkk\ni8XG2ij34waAkBFrASB8NQWv0EZYuPvXzWz1PIe8RtJn3P1nwfELJisAIM2izEZHLct1A4CkINYC\naDZxLrp5pqROM/uqme0xs41zHWhml5lZv5n1Dw4ORlhEAGgexFoACB+xFgBqF2fColXSOkl/LOmP\nJF1tZmfOdqC7X+/uve7eu2LFiijLCABNg1gLAOEj1gJA7eLcJWRA0pC7j0oaNbOvS3qOpB/GWCYA\nAAAAAJAAcY6w+JykF5pZq5ktk/Q7kh6KsTwAAAAAACAhwtzW9BZJ50k6ycwGJF2j8valcvft7v6Q\nmd0paa+kkqSPu/sDYZUHAAAAAACkR5i7hFxSwzFbJW0NqwwAAAAAACCd4lzDAkCGsDd8c+P8AwCQ\nDrTZSBMSFgDqViq59h8c0aad/RoYHlN3Z0E7NvZqzaoOGsAmwPkHACAdaLORNiQsANRtaLSo6760\nX1dfuFbLC206PHZU131pv65d36MVHUsWfD6Z/nSr9/wnBdchgKQjTqFeQ6PFqWSFJA0Mj2nTzn7t\n2tyXqjY7qZrlHo2yniQsANStVCrp9S84TVfdvncqW79lQ49KpVINzyXTn3b1nP+k4DoEkHTEKTRC\ncWJyKllRMTA8puLEZEwlyo5muUejrmec25oCyIiJkk+9WZXKDd9Vt+/VRMkXfO5cmf6h0WKoZUbj\n1HP+k4LrEEDSEafQCPnWFnV3FqY91t1ZUL61JaYSZUez3KNR15OEBYC6TZZ81mz9ZA1vWMn0p189\n5z8puA4BJB1xCo3Q1Z7Xjo29U0mLyqfjXe35mEuWfs1yj0ZdT6aEAKhba0tO3Z2FacGru7Og1paF\nc6KVTP/M55LpT496zn9ScB0CSDriFBohlzOtWdWhXZv7Mr/OQtSa5R6Nup7p6U0CSKyVJyzR9kvX\nTcvWb790nVaesPDiTWT606+e858UXIcAko44hUbJ5UwrOpbo5M5lWtGxhGRFgzTLPRp1Pc09PUN2\nJam3t9f7+/vjLgaAGSYmSjr05LgmJktqbclp5QlL1NpaW060WVZUboDI/iiLjbX1nP+k4DoEEEhs\nrCVOAcnWLPdog+pZ0xOYEgKgIVpbc/r15YWFD5xFJdMfpWZpUKJSz/lPijiuw0bimgayL+1xCsci\ndmdLs9yjUdaThAWAptMs206heXBNA0D6ELuBhaVrvC4Qk4mJkh4/PKafDo3q8cNjmpgoxV0k1KFZ\ntp2KEvdIvLimASB94ordpZJrcGRcB4aPaHBkXKUU7eqFZIjyGmKEBbCAiYmSfnBwRJffvGcq+739\n0nV69qqO1M3RR1lc205lddgn90j8srKVWlbvEQCYTRyxu1RyPTo0qp8OHdGyfIuOFCf1zK5lWt3V\nTrxFTaIeGURPEljAoSfHp96ISeWG5PKb9+jQk+MxlwzHq7IdU7Wwt52qBPf123arb8vdWr9tt/Yf\nHMnEpxrcI/GL45putCzfIwAwmzhi9+Gxog7+8ild/bkH9Krr79XVn3tAB3/5lA6PMSIPtYl6ZBAJ\nC2ABRydLs2a/JyYZ8p5WcWw7leUh+9wj8cvCVmpZvkcAYDZxxO6x4qSuvG3vtFh75W17NVZM14g8\nxCfqkUFMCQEW0NaSU3dnYdqN2d1ZUGsL+b60yuVMa1Z1aNfmvsiGnhcnJrXihCW6+sK1Wl5o0+Gx\no9r+1R+nbsj+bLhH4hfHNd1oWb5HAGA2ccTuSfdZ32xOMpgNNcq3tuila1dqw7pTptrr2/c8FtrI\nIBIWwAJWnrBE2y9dd8z8/JUnZH/LoiyLetupQr5Ff3XBmqlPNbo7C9p6cY8K+fQM2Z8L90gypH0r\ntSzfI0CjsM5L9kQdu5e2tcz6IcPSNj5kQG06C216x/lnHtPv6yy0hfJ65p6udFpvb6/39/fHXQw0\nkVLJdeDwEY1PuHImlVxa0mo6efkyOgmo2aGRp/Sn2+45poPwmc0v0MqOpbX+msguuMXEWu4RNEKD\n7hGgERIba9kCE/XiOkK9BkfGtX7b7mPa612b+xabfKvpgmOEBbCAodGiLtnxrUbclGhiRydmX+fh\naAa2/+QeQSNk+R4BGmFotKjrvrR/2rSp6760X9eu7yHWomZZmEKIeLGGBZAwWdkuEPFqa519nYe2\nDGz7yT2CRsjyPQI0QqlU0utfcJquuv1X06a2bOhRqURSD4uT9imEiFdld5uZ7XVYa1jQCwAWkIXt\nAhG/1pxp68U901YC33pxj1oz8IkG9wgaIcv3CNAIk66pZIVUTgxfdfteFksEEKmod7dhhAWwgMpN\nOXOuX5q2C0T8xoqT+tCd04fyfujO/frH1zxXao+7dPXhHkEjZPkeARrB59jdIW3r0QFIt6inFZGw\nABbAXD80Qr61RYNPjuvNn9wz9VhWRiFwj6ARsnyPAI0Q9TBsAJhLlNOKmBIC1KByU57cuUwrOpbw\nRgyLFvXwuahxj6BeWb9HgHpxjwBoRoywANAQ7A0/P0YhAPPL5UxnrDhBt775+ZqYLKm1JaeVJ5D8\nAipoR9Ao9NmQJiQsANSNPb1rw6rcwNxKJdePBp8kjgDzoB1BveizIW1IWACoG3vDA6gXcQQAwjc0\nWpxKVkjlhVs37ezXrs19xFrULMpROiQsANSNveEB1Is4AiyMofyoV3FictbdZooTkzGVCGkT9Sgd\nFt0EUDf2hgdQL+IIML/Km4T123arb8vdWr9tt/YfHFGpxE2C2lV2m6nGbjNYjLlG6QyNFkN5PRIW\nAOo2WSrNmq2fpBMFoEbEEWB+Ub9JQDax2wzqFfUoHaaEQBJDDFGfnNmse8Mn+RLimm8s/p6oVxrj\nyEzcBwgTQ/nRCHHtNkN8zI7KKJ2Z7XVYo3QYYQGGGKJuhXyLtl7cMy1bv/XiHhXyyRxeyDXfWPw9\n0QhpiyMzcR8gbAzlR6NUdps5uXOZVnSEv3008TFbOgtt2n7pumnt9fZL16mz0BbK65l7ui6U3t5e\n7+/vj7sYmTI4Mq7123YfkyVjtWDUqlRyPTo0qp8OHdGyfIuOFCf1zK5lWt3VnsjseYqv+cj+mIuJ\ntSn+eyJB0hZHZuI+yJRExlq2o0SjRD3agfiYLYMj43rvrr3asO6UqV29bt/z2PHs6lXTRceUEDDE\nEHXL5Uyru9rVsbQtFUP9uOYbi78nGiFtcWQm7gOELa6h/MiWOBJfxMdsKU5M6q59h3TXvkPTHr/m\nFeGcT6aEQBbMG67W3VmQGQ0gahf18MJ6MKy2sYghaJQ0xZGZiCuIQprvESRDHIu3Eh+zJerzScIC\najFpy4bp84a3bOhRC20gIlIquQZHxnVg+IgGR8ZDn9PICtmNRQxBmjUq/hBXgPSLuj8ShzhGOxAf\nsyXq88mUECiXy+mmex7R1ReunZqHdNM9j+ja9T1xFw1NII6hiQyrbSxiCNKqkfGHuAKkW7OsERL1\nDg8S8TFroj6fLLqJpgnQSCYWYloUFoIDGoj4gzkkMtZKbA0ZpmaJB7TZSBAW3UTtlrTm9IGLzppa\nmX1JK7OFEA0WYsoGYgjSiPiDNOGNZriaJR4w2gFpE1rCwsxukHShpEPuftY8xz1P0jclvdrdbwur\nPJjb0GhRG2/4duYzykimOIYmorGIIUgr4g/SZK7FEom1jdFM8aCyeCuQBmF+BHajpAvmO8DMWiRt\nkXRXiOXAApolo4yFxbHYFAsxpR8xBElUSzwj/iBNiLXhIh4sTjMsUIpkCG2Ehbt/3cxWL3DY2yXd\nLul5YZUDC2umjDLmFtdQU4Ymph8xBElTazwj/iBNiLXhIh7UjulJiFJsk4zN7GRJ6yV9NK4yoIyM\nMqR49uWuYF/5dCOGIGkWE8+IP0gLYm34iAe1ibPPiOYT56Kb/yDpKncvmc0fDMzsMkmXSdKpp54a\nQdGaCxllSAw1xfHHWmIIkoZ4hiQj1iLtiLGIUpwJi15Jnw6SFSdJermZTbj7Z2ce6O7XS7peKm//\nFGkpmwSL74Chpqgn1hJDkCTEMyQZsRZpR4xFlGKbEuLup7n7andfLek2SZtnS1YAiAZDTQFkBfEM\nAMJDjEWUwtzW9BZJ50k6ycwGJF0jqU2S3H17WK8L4Pgw1BRAVhDPACA8xFhEKcxdQi5ZxLFvCKsc\nAGrHUNNwlUquodEijTsQAeJZcyLOAgtrxH1CjEVU4lzDAgCaBluAAUC4iLPAwrhPkDaxrWEBAM2E\nLcAAIFzEWWBh3CdIGxIWABABtgADgHARZ4GFcZ8gbZgSAgARyLe26KVrV2rDulO0vNCmw2NHdfue\nx9gCDAAahDgLLIwtSdEIUa4XRMICACLQWWjTO84/U5ffvGdqzuj2S9eps9AWd9EAIBOIs8DCKluS\nzlzDgi1JUauo10FhSggARGB47OhUJ1oqD7+8/OY9Gh47GnPJACAbmiHOlkquwZFxHRg+osGRcZVK\nHneRkDLVW5LuvurF2rW5jwU3sShRr4PCCAsAiABzRgEgXFmPs+zugEZhS1LUI+pYywgLAIiAmam7\nszDtse7OgszoZAJAI2Q9zrK7A4AkqKyDUi3MdVBIWABABFpM2rKhZyrAd3cWtGVDj1qy0Y8GgNhl\nPc5mfQQJgHSorINSHWvDXAelKaaETEyUdOjJcR2dLKmtJaeVJyxRayu5GiApolxpOC65XE433fOI\nrr5w7dTq9Tfd84iuXd8Td9EagjgLZEOa43HW4yy7O6BR0nyf41hRn8/qdVDYJaQBJiZK+sHBkWNW\njH72qg4600ACNMuc3K72vK74wzWZXJWbOAtkQ9rjcZbjrFTeBWX7pevYBQV1Sft9juniOp9RroNi\n7ulaXbi3t9f7+/trPv7xw2P6849985hs9K1vfr5+fXlhnmcCv0ImOjyDI+Nav233Mffors19mVsQ\nqgHXUWQX3WJiLXEWjUKsjVcW4nGDrqFExtrBkXG9d9debVh3ytQIktv3PKZr1/ek5vwgflxH2ZLy\nuF1TrM38CIujk6VZ5/tNTJZiKhHSplRyPTo0qp8OHdGyfIuOFCf1zK5lWt3VTke6AZppTm5WV+Um\nzqIRiLXxy0I8zmqclcrn5659h3TXvkPTHr/mFek5P4hfqVTS619wmq66fe/UJ/JbNvSoVKLNTqO4\n4naUHzBkPmHR1pLTS9euPCaL2NrCMGXU5vBYUQd/+ZSu/twDU4F968U9Wr6sTU9vD69T1CyfNDIn\nN/2Is2iEuGJtI6U9bhOPk43zg0aYdE0lK6Tym9urbt+rW9/8/JhLFp20x+pqccSFqKehZL43uaI9\nr7eff6Y+8Pl9etX19+oDn9+nt59/plZkZD4jwjdWnNSVt00P7FfetldjxfAyl5VAsH7bbvVtuVvr\nt+3W/oMjKpXSNYWrFlGvNIzGI86iEeKItY2UhbhNPE42zg8awd1n/UQ+bcsEHK8sxOpqccSFJ0bH\nZ91i+YnR8VBeL/MjLA4/NaG3BIsTSeU/6Ftu3lOe19NGRhoLm5wjsE+GGNfm2ms9JfPRFiXqlYbR\neMRZNEIcsbaRshC3icfJxvlBIzT7SJ0sxOpqccSFp47OPg3lqaPhTCvK/AiLLMzHRLyWtrVMZS0r\nujsLWtoW3u3TbNdtZc7xyZ3LtKJjCZ2vlGm26xXhiCPWNlJW7gPicbJxflCvZh+pk5VYXS3quNBi\nNmt73RLSy6ajF1CHShaxWjNlEVG/k9qXaMfrZgT21/XqpBDnVHPdIk24XtEIccTaRuI+QBRKJdfg\nyLgODB/R4Mh4aoexIz7Vn8jvvurF2rW5r6m2NCVW16+Qb9HWi3umtddbL+5RIR/O3zDz25qy1zDq\nFcfK9Vy3mEMit9rjekUjpH2XEO6DTCHWAhnFfVS/BrbXNR2c+YSFlK2VYBG9uPY35rrFLBLZiZa4\nXlG/lO8lL4n7IEMSGWuzcI8ASUCsrl+D/oY1PSHzi25K2d6TG+GLa65bPdctgRhRI86iXlmYV9zo\n+4BYjmpx3SNch8ga+iz1i/Jv2BQJC6AeaVtNmaFuANIobbE2bMRyzBTHPcJ1CCBumV90E6hX2lZT\nnmu7pqHRYswlA4C5pS3Who1YjpniuEe4DgHEjREWwALStu95FoZVA2g+aYu1YSOWY6Y47hGuQwBx\nI2EB1CBNc90YVg0grdIUa8NGLMdsor5HuA4BxI0pIUDGMKwaANKPWI4k4DoEEDdGWAAZw7BqAEg/\nYjmSgOsQQNxIWAAZxLBqAEg/YjmSgOsQQJyYEgIAAAAAABKHhAUAAAAAAEgcpoQAaIhSyTU0WmSO\nKwAgNrRFABC+KGMtCQsAdSuVXPsPjmjTzn4NDI9NrSK+ZlUHHUUAQCRoiwAgfFHHWqaEAKjb0Ghx\nKmhJ0sDwmDbt7NfQaDHmkgEAmgVtEVCbUsk1ODKuA8NHNDgyrlLJ4y4SUiTqWMsICwB1K05MTgWt\nioHhMRUnJmMqEQCg2TRDW8SUF9SLkUioV9SxlhEWAOpmZuruLEx7rLuzIDMaPgBANLLeFlXeaK7f\ntlt9W+7W+m27tf/gCJ+OY1EYiYR65VtbZo21+daWUF6PhAWAupm5tmzomQpe3Z0FbdnQo4z0EQEA\nKZD1tog3mmiEZhiJhHB1tee1Y2PvtFi7Y2OvutrzobweU0IA1K1Ukm665xFdfeFaLS+06fDYUd10\nzyO65hW/FXfRAABNIuttEW800QiVT8err6UwPx1H9uRypjWrOrRrcx+7hABIh0K+RW/sO01X3rZ3\naj7k1ot7VMjT+AEAopH1tog3mmiEyqfjM9ewCOvTcWRTLmda0bEkktciYQGgbssLea06cak+cNFZ\nWpZv0ZHipFaduFTLCzR+AIBoZL0t4o0mGiHqT8eBeoWWsDCzGyRdKOmQu581y89fK+kqSSZpRNJb\n3P3+sMoDIDy5nGl1V7s6lrbR+AEAYpH1tog3mmiUKD8dB+oV5giLGyX9o6Sdc/z8EUm/7+7DZvYy\nSddL+p0QywMgRDR+AIC4Zb0tynr9AGCm0BIW7v51M1s9z8/vqfr2XkndYZUFAAAAAACkS1K2NX2T\npH+LuxAAAAAAACAZYl9008xerHLC4oXzHHOZpMsk6dRTT42oZM2lVHINjRaZEwk0MWIt4kZbhGZA\nrAWQdlG217EmLMysR9LHJb3M3YfmOs7dr1d5jQv19vZ6RMVrGqWSa//BkWNWnV6zqoOOItBEiLWI\nE20RmgWxFnEjOYx6RN1exzYlxMxOlfQZSa9z9x/GVQ5IT4yOT11wkjQwPKZNO/v1xOh4zCUDADQL\n2iIACF/lzeb6bbvVt+Vurd+2W/sPjqhUIneG2gyNFmdtr4dGi6G8XmgJCzO7RdI3Ja0xswEze5OZ\nXW5mlweHvE9Sl6RtZnafmfWHVRbM76mjk1MXXMXA8JieOlqKqUQAgGZDWwQA4Yv6zSaypzgxe3td\nnJgM5fXC3CXkkgV+/peS/jKs10ftWszU3VmYduF1dxbUwsgwAEBEaIsAIHxRv9lE9uRbW2Ztr/Ot\nLaG8XlJ2CUGMCvkWbb24R92dBUnlC27rxT0q5MO56AAAmIm2CADCV3mzWS3MN5vInq72vHZs7J3W\nXu/Y2Kuu9nworxf7LiGI3/JCXqtOXKoPXHSWluVbdKQ4qVUnLtXyQjgXXb1YKAgAsidtbVEUaO8A\nNFrlzebMBRPDerOZVc0cn3M505pVHdq1uS/7u4QgGXI50+qudnUsbUv8Tccq8gCQTWlqi6JAewcg\nDFG/2cwi4nP5OlrRsSSa14rkVZB4lYvu5M5lWtGxJLE3GwsFAUB2paUtigLtHYCwEGvrQ3yOVk0J\nCzNbZmZXm9mO4PszzOzCcIsGHIuFggAAzYD2DgCSifgcrVpHWHxC0rik5wffH5D030IpEWJRKrkG\nR8Z1YPiIBkfGE7sXMwsFAUB2paUtigLtHWbDPQLEj/gcrVoTFqe7+4ckHZUkdz8iibFDGVGZh7V+\n2271bblb67ft1v6DI4lsBKNelRYAEI00tUVRoL3DTNwjQDIQn6Nl7gsHOTO7R9L5kna7+zlmdrqk\nW9z93LALOFNvb6/39/dH/bKZNjgyrvXbdh+zl+6uzX2RLaayGM28Ki+aXmQXOrEWUUtbWxQF2rvY\nJDLWco8AyUF8boia/mC17hJyjaQ7JZ1iZp+S1CfpDcdXLiRN2uZhRbkqLQAgGmlri6JAe4dq3CNA\nchCfo1PTlBB3/5KkP1U5SXGLpF53/2p4xUKUmIcFAIgbbREwP+4RAM2o1l1CzpH0TEk/l/S4pFPN\n7HQzq3WEBhKMeVgAgLjRFgHz4x4B0IxqTThsk3SOpL0qzzU5S9KDkp5mZm9x97tCKh8ikMuZ1qzq\n0K7NfczDAgDEgrYImB/3CIBmVGvC4nFJb3L3ByXJzNZK+ltJfyXpM5JIWKQc87AAAHGjLQLmxz0C\noNnUuq3pmZVkhSS5+z5Jz3b3n4RTLAAAAAAA0MxqHWHxoJl9VNKng+9fJWmfmS2RdDSUkgEAAAAA\ngKZV6wiLN0h6WNK7gn8/CR47KunFYRQMAAAAAAA0r5pGWLj7mKS/D/7N9GRDSwQkUKnkGhotssgV\nAAAJRnsNLIz7BGlSU8LCzM6Q9N8lrZW0tPK4uz8rpHIBiVEqufYfHNGmnf0aGB6b2kZszaoOgjsA\nAAlBew0sjPsEaVPrlJBPSPqopAmVp4DslHRzWIUCkmRotDgV1CVpYHhMm3b2a2i0GHPJAABABe01\nsDDuE6RNrQmLgrt/WZK5+0/d/f2S/ji8YgHJUZyYnArqFQPDYypOTMZUIgAAMBPtNbAw7hOkTa27\nhIybWU7Sj8zsbZIOSDohvGIByZFvbdFL167UhnWnaHmhTYfHjur2PY8p39oSd9EAAECA9hpYWL61\nRd2dhWlJi+7OAvcJFiXKdVBqTVi8U9IySe+Q9AFJL5H0+lBKBCRMZ6FN7zj/TF1+856puX7bL12n\nzkJb3EUDAAAB2mtgYV3tee3Y2HvMGhZd7fm4i4aUiHodFHP32g82O1GSu/tIw0tSo97eXu/v74/r\n5dGEBkfGtX7b7mMy0bs292lFx5IYS4YmFNlqWMRaAGnTwPaaWItMY5cQ1CPqWFvrLiG9Ki+82RF8\n/wtJf+HuexZTIiCNmOsHAEDy0V4DtcnljA/dcNyijrW1Lrp5g6TN7r7a3VdLeqvKCQwg8ypz/aox\n1w8AgGRphva6VHINjozrwPARDY6Mq1SqfaQ0ADRC1LG21oTFpLt/o/KNu/+7ylucApnXWWjT9kvX\nTd2YzIkFACB5st5eV+aNr9+2W31b7tb6bbu1/+AISQsAkaqsg1Ida8NcB6XWRTe/ZmYfk3SLJJf0\nKklfNbNzJMndvxtK6YAEGB47qg9/+Ye6+sK1U6uOf/jLP9S163sYTgcAQEJkvb0eGi1OLXInlYdg\nb9rZz5paACKVy5nWrOrQrs19idol5DnB/9fMePy5KicwXtKwEgEJU5yY1F37DumufYemPX7NK5gT\nCwBAUmS9vWaNDgBJEeU6KAsmLMwsJ+mj7n5rBOVBBqV9JWL2qwYAIPnyrS166dqV2rDulKkRFrfv\neSwz7XXa+yNp7w8CiMeCCQt3L5nZX0kiYYFFi3qf3jCwXzUAAMnXWWjTO84/U5ffvGeqvc7SGhZp\n7o9koT8IIB7mvvBCPWb2QUkGAAsRAAAf1klEQVRPSPoXSaOVx939/4VXtNmxX3W6NHCf3ljxqQAS\nIrKLjlgLIG0a2OdIbKxNa38kK/1BAA1VU/CqdQ2LVwX/v7XqMZf0rMWUCM0nK/Mt2a8aAIBky0qf\nYz5p7Y80w7kBEI6aEhbuflrYBUE2pX2+JQAASAf6HMnFuQFwvHK1HmhmZ5nZn5vZxsq/MAuGbIh6\nn14AANCc6HMkF+cGwPGqaYSFmV0j6TxJayV9QdLLJP27pJ2hlQyZEPU+vQAAoDnR50guzg2A41Xr\nGhYXS3qOpO+5+xvNbJWkm8MrFrIkrfMtAQBAutDnSC7ODYDjUeuUkKfcvSRpwsxOlHRI0inhFQsA\nAAAAADSzWkdYfMfMlkvaIWmPpCclfTO0UgEAAAAAgKZWa8LiREl/Jumrku6UdKK77w2rUAAAAAAA\noLnVOiXknyQ9Q9L/kvQVSdeY2Tvne4KZ3WBmh8zsgTl+bmb2YTN72Mz2mtk5iyo5AAAAAADIrJoS\nFu5+t6RrJV2t8rSQXklvWeBpN0q6YJ6fv0zSGcG/yyR9tJayAAAAAACA7Kt1W9MvS2pXed2Kb0h6\nnrsfmu857v51M1s9zyEXSdrp7i7pXjNbbmbPcPef11RyAAAAAACQWbVOCdkrqSjpLEk9ks4ys0Kd\nr32ypMeqvh8IHgMAAAAAAE2uphEW7n6FJJlZh6Q3SPqEpF+TFMlmymZ2mcrTRnTqqacu+vmlkmto\ntKjixKTyrS3qas8rl7NGFxMAUq3eWAvUi/YazYBYi7gRa5EmtU4JeZuk35O0TtKjkm5QeWpIPQ5I\nOqXq++7gsWO4+/WSrpek3t5eX8yLlEqu/QdHtGlnvwaGx9TdWdCOjb1as6qDGxMAqtQTa4F60V6j\nWRBrESdiLdKm1ikhSyX9D0nPdvc/cPe/cfev1Pnad0jaGOwW8ruSfhHG+hVDo8WpG1KSBobHtGln\nv4ZGi41+KQAAcJxor4GFlUquwZFxHRg+osGRcZVK5DuwOMRapE2tU0L+brG/2MxukXSepJPMbEDS\nNZLagt+3XdIXJL1c0sOSjkh642JfoxbFicmpG7JiYHhMxYnJMF4OAAAcB9prYH58Mo5GINYibWpK\nWBwPd79kgZ+7pLeG9foVZqbuzsK0G7O7syAzAjsAAElBew3Mb65Pxndt7tOKjkiWlUMG5FtbZo21\n+daWGEsFzK3WKSGp1WLSlg096u4sb2rS3VnQlg09aqH/AwBAYtBeA/Pjk3E0Qld7Xjs29k6LtTs2\n9qqrPR9zyYDZhTbCIilyuZy+vv+gPvGG56klZ5osuW7r/5nOXPUbcRcNAAAEaK+B+fHJOBohlzOt\nWdWhXZv72CUExy3KnWYyn7DoLLTpFWd36403fmdqvt/2S9eps9AWd9EAAECA9hqYX+WT8ZlrWPDJ\nOBYrlzOmEeG4Rb2eTuanhAyPHdXlN++ZNt/v8pv3aHjsaMwlAwAAFbTXwPyqPxnffdWLtWtzHwtu\nAohc1DvNZH6EBfP9AABIPtprYGF8Mg4gblG315kfYVGZ71eN+X4AACQL7TUAAMkXdXud+YQFK+EC\nAJB8tNcAACRf1O115qeEsBIuAADJR3sNAEDyRd1eZz5hITHfDwCANKC9BgAg+aJsrzM/JQQAAAAA\nAKQPCQsAAAAAAJA4JCwAAAAAAEDikLAAAAAAAACJQ8ICAAAAAAAkDgkLAAAAAACQOCQsAAAAAABA\n4pCwAAAAAAAAidMadwGiUCq5hkaLKk5MKt/aoq72vHI5i7tYAAAAaKCs9/myXj8AmCnzCYtSybX/\n4Ig27ezXwPCYujsL2rGxV2tWdRDgq9AAAgCANCuVXI8OjeqnQ0e0LN+iI8VJPbNrmVZ3tWeiT0Of\nFo1Cvx9pkvkpIUOjxanALkkDw2PatLNfQ6PFmEuWHJUG/oEDv9DA8JgeOPALPTo0qlLJ4y4aAABA\nTQ6PFTXy1NFpj408dVSHx7LR56NPi0aoJL7Wb9utvi13a/223dp/cIR+PxalVHINjozrwPARDY6M\nh3r9ZH6ERXFiciqwVwwMj6k4MRlTiZLn8FhRB3/5lK7+3ANTGfutF/do+bI2Pb19SdzFAwAAWNDR\niZKOFCeP6c8cnSjFXbSGoE+LRpgr8bVrc59WdNDvx8KiHu2V+REW+dYWdXcWpj3W3VlQvrUlphIl\nz1hxUlfetnda4Lrytr0aK9IAAgCAdDha8ln7M0cz8skxfVo0Aokv1Cvq0V6ZT1h0tee1Y2PvVICv\nZIC62vMxlyw5Jt1nDVyT2WjfAQBAEyjN0Z8peTY6NPRp0QgkvlCvqJNemZ8SksuZ1qzq0K7NfSws\nM4elbeXAVX3hdXcWtLQt8/ksAACQEXP3Z7LxRow+LRqhkviaOZyfxBdqVUl6zYy1YSW9muIdaS5n\nWtGxRCd3LtOKjiUE9hlOal8ya8b+JNavAAAAKdEM/Rn6tKhXdeJr91Uv1q7Nfew0g0WJerRX5kdY\nYGFk7AEAQNrRnwFqU0l8Accj6lhLwgKSCFwAACD96M8AQPiijLVNMSUEAAAAAACkCwkLAAAAAACQ\nOEwJgSSpVHINjRaZ8wkAAFIr6/2ZrNcPAGYiYQGVSq79B0eO2d6IFYMBAEBaZL0/k/X6AcBsmBIC\nDY0Wpxo/SRoYHtOmnf0aGi3GXDIAAIDaZL0/k/X6AcBsSFhAxYnJqcavYmB4TMWJyZhKBAAAsDhZ\n789kvX4AMBsSFlC+tUXdnYVpj3V3FpRvbYmpRAAAAIuT9f5M1usHALMhYQF1tee1Y2PvVCNYmRPZ\n1Z6PuWQAAAC1yXp/Juv1A4DZsOgmlMuZ1qzq0K7Nfaw6DQAAUinr/Zms1w8AZkPCApLKjeCKjiVx\nFwMAAOC4Zb0/k/X6AcBMTAkBAAAAAACJQ8ICAAAAAAAkDgkLAAAAAACQOKEmLMzsAjPbb2YPm9m7\nZ/n5qWZ2t5l9z8z2mtnLwywPAAAAAABIh9ASFmbWIukjkl4maa2kS8xs7YzD/qukW939uZJeLWlb\nWOUBAAAAAADpEeYIi3MlPezuP3H3oqRPS7poxjEu6cTg66dJejzE8gAAAAAAgJQIc1vTkyU9VvX9\ngKTfmXHM+yXdZWZvl9Qu6Q9m+0VmdpmkyyTp1FNPXXRBSiXX0GiRPasBYB71xloAiFsa+nzEWsQt\nDfcJUBFmwqIWl0i60d3/3syeL+mTZnaWu5eqD3L36yVdL0m9vb2+mBcolVz7D45o085+DQyPqbuz\noB0be7VmVQc3JgBUqSfWAkDc0tLnI9YiTmm5T4CKMKeEHJB0StX33cFj1d4k6VZJcvdvSloq6aRG\nFmJotDh1Q0rSwPCYNu3s19BosZEvAwAAgBjR5wMWxn2CtAkzYfEdSWeY2Wlmlld5Uc07ZhzzM0nn\nS5KZ/abKCYvBRhaiODE5dUNWDAyPqTgx2ciXAQAAQIzo8wEL4z5B2oSWsHD3CUlvk/RFSQ+pvBvI\ng2b2t2b2yuCw/yxpk5ndL+kWSW9w94YOjcu3tqi7szDtse7OgvKtLY18GQAAAMSIPh+wMO4TpE2Y\nIyzk7l9w9zPd/XR3vzZ47H3ufkfw9T5373P357j72e5+V6PL0NWe146NvVM3ZmWeVld7vtEvBQAA\ngJjQ5wMWxn2CtIl70c3Q5XKmNas6tGtzHyvhAgAAZBR9PmBh3CdIm8wnLKTyjbmiY0ncxQAAAECI\n6PMBC+M+QZqEOiUEAAAAAADgeJCwAAAAAAAAiUPCAgAAAAAAJA4JCwAAAAAAkDgkLAAAAAAAQOKQ\nsAAAAAAAAIlDwgIAAAAAACQOCQsAAAAAAJA4JCwAAAAAAEDikLAAAAAAAACJQ8ICAAAAAAAkDgkL\nAAAAAACQOCQsAAAAAABA4pCwAAAAAAAAiUPCAgAAAAAAJA4JCwAAAAAAkDgkLAAAAAAAQOKQsAAA\nAAAAAIlDwgIAAAAAACQOCQsAAAAAAJA4JCwAAAAAAEDikLAAAAAAAACJQ8ICAAAAAAAkDgkLAAAA\nAACQOCQsAAAAAABA4pCwAAAAAAAAiUPCAgAAAAAAJA4JCwAAAAAAkDgkLAAAAAAAQOKQsAAAAAAA\nAIlDwgIAAAAAACQOCQsAAAAAAJA4JCwAAAAAAEDikLAAAAAAAACJQ8ICAAAAAAAkDgkLAAAAAACQ\nOCQsAAAAAABA4oSasDCzC8xsv5k9bGbvnuOYPzezfWb2oJn9c5jlAQAAAAAA6dAa1i82sxZJH5H0\nh5IGJH3HzO5w931Vx5wh6T2S+tx92MxWhlUeAAAAAACQHmGOsDhX0sPu/hN3L0r6tKSLZhyzSdJH\n3H1Yktz9UIjlAQAAAAAAKRFmwuJkSY9VfT8QPFbtTElnmtluM7vXzC4IsTwAAAAAACAl4l50s1XS\nGZLOk3SJpB1mtnzmQWZ2mZn1m1n/4OBgxEUEgOZArAWA8BFrAaB2YSYsDkg6per77uCxagOS7nD3\no+7+iKQfqpzAmMbdr3f3XnfvXbFiRWgFBoBmRqwFgPARawGgdmEmLL4j6QwzO83M8pJeLemOGcd8\nVuXRFTKzk1SeIvKTEMsEAAAAAABSILSEhbtPSHqbpC9KekjSre7+oJn9rZm9Mjjsi5KGzGyfpLsl\nXenuQ2GVCQAAAAAApENo25pKkrt/QdIXZjz2vqqvXdJ/Cv4BAAAAAABIin/RTQAAAAAAgGOQsAAA\nAAAAAIlDwgIAAAAAACQOCQsAAAAAAJA4JCwAAAAAAEDikLAAAAAAAACJQ8ICAAAAAAAkDgkLAAAA\nAACQOCQsAAAAAABA4rTGXYAolEquodGiihOTyre2qKs9r1zO4i4WAAAAGog+HwCEL8pYm/mERank\n2n9wRJt29mtgeEzdnQXt2NirNas6aMAAAAAygj4fAIQv6lib+SkhQ6PFqT+mJA0Mj2nTzn4NjRZj\nLhkAAAAahT4fAIQv6lib+YRFcWJy6o9ZMTA8puLEZEwlAgAAQKPR5wOA8EUdazOfsMi3tqi7szDt\nse7OgvKtLTGVCAAAAI1Gnw8Awhd1rM18wqKrPa8dG3un/qiVOTZd7fmYSwYAAIBGoc8HAOGLOtZm\nftHNXM60ZlWHdm3uY8VoAACAjKLPBwDhizrWZj5hIZX/qCs6lsRdDAAAAISIPh8AhC/KWJv5KSEA\nAAAAACB9SFgAAAAAAIDEIWEBAAAAAAASh4QFAAAAAABIHBIWAAAAAAAgcUhYAAAAAACAxCFhAQAA\nAAAAEoeEBQAAAAAASBwSFgAAAAAAIHFIWAAAAAAAgMQhYQEAAAAAABKHhAUAAAAAAEgcc/e4y7Ao\nZjYo6afH+fSTJD3RwOIkFfXMlmapp9Q8dT3eej7h7hc0ujCzqSPWZuEcpr0OaS+/lP46pL38Uvrr\nUE/5ibXJ0Sz1lJqnrtQzW0KPtalLWNTDzPrdvTfucoSNemZLs9RTap66ZrmeWahb2uuQ9vJL6a9D\n2ssvpb8OaS//QrJev4pmqafUPHWlntkSRT2ZEgIAAAAAABKHhAUAAAAAAEicZktYXB93ASJCPbOl\nWeopNU9ds1zPLNQt7XVIe/ml9Nch7eWX0l+HtJd/IVmvX0Wz1FNqnrpSz2wJvZ5NtYYFAAAAAABI\nh2YbYQEAAAAAAFIgMwkLM7vBzA6Z2QNVj51tZvea2X1m1m9m5waPm5l92MweNrO9ZnZOfCVfHDM7\nxczuNrN9Zvagmb0zePzpZvYlM/tR8H9n8Hgq6zpPPbea2Q+Cuuwys+VVz3lPUM/9ZvZH8ZV+ceaq\na9XP/7OZuZmdFHyfqXMa/OztwXl90Mw+VPV46s7pPNdu5uKRJJnZcjO7LTh/D5nZ8+eKR0k0R/nf\nb2YHgnN1n5m9PO5yzsXM1lSV8z4z+6WZvSst52Ce8qfmHEiSmV0R3O8PmNktZrbUzE4zs28F9/a/\nmFk+7nLOZY7y32hmj1Sdg7PjLud8zOydQfkfNLN3BY+l4j5YyDztSibqVxFcd982s/uDev5N8Hhq\n7qXFMLMWM/uemX0++D5z9TSzR83s+5W+T/BYpq5bKf19oVrF1udw90z8k/QiSedIeqDqsbskvSz4\n+uWSvlr19b9JMkm/K+lbcZd/EfV8hqRzgq87JP1Q0lpJH5L07uDxd0vakua6zlPPl0pqDR7fUlXP\ntZLul7RE0mmSfiypJe561FPX4PtTJH1R5T3aT8roOX2xpP8raUnws5VpPqfz1DNz8Sgo/02S/jL4\nOi9p+VzxKIn/5ij/+yX9l7jLdhx1aZH0H5KemaZzMEf5U3MOJJ0s6RFJheD7WyW9Ifj/1cFj2yW9\nJe6yLrL8N0q6OO7y1ViHsyQ9IGmZpNagTfmNNN4Hc9RvUX2/tP4L2sETgq/bJH0raBdTcS8dR33/\nk6R/lvT54PvM1VPSowr6r1WPZeq6DeqR6r7QcdY5sj5HZkZYuPvXJf2/mQ9LOjH4+mmSHg++vkjS\nTi+7V9JyM3tGNCWtj7v/3N2/G3w9IukhlTsbF6l8syj4/0+Cr1NZ17nq6e53uftEcNi9krqDry+S\n9Gl3H3f3RyQ9LOncqMt9POY5p5J0naS/UvlarsjUOZX0FkkfdPfx4GeHgqek8pzOU8/MxSMze5rK\nyeJ/kiR3L7r7Yc0djxJlnvKn1fmSfuzuP1VKzsEM1eVPm1ZJBTNrVflN888lvUTSbcHPk34OZpb/\n8QWOT5rfVDnZeyToI3xN0p8qnffBMY6j75dKQTv4ZPBtW/DPla57qSZm1i3pjyV9PPjelMF6ziFT\n123a+0J1iKzPkZmExRzeJWmrmT0m6e8kvSd4/GRJj1UdN6BfvUFMDTNbLem5KmegV7n7z4Mf/Yek\nVcHXqa/rjHpW+wuVP5mWMlBPaXpdzewiSQfc/f4Zh6W+rjPO6ZmSfi8YBvk1M3tecFjW6pnFeHSa\npEFJnwiGtX7czNo1dzxKmrnKL0lvs/IUnRtSNIzz1ZJuCb5OyzmoVl1+KSXnwN0PqHxP/0zlRMUv\nJO2RdLgqwZ7Y+3q28rv7XcGPrw3OwXVmtiS2Qi7sAZXbkS4zW6byyLVTlM77YF419v1SK5gmcZ+k\nQ5K+pPLoylTcS4v0Dyp/IFUKvu9SNuvpku4ysz1mdlnwWNau27T3hY5XZH2OrCcs3iLpCnc/RdIV\nCjJfWWBmJ0i6XdK73P2X1T/z8nicTGz/Mlc9zey9kiYkfSqusjVadV1VrttfS3pfrIUKwSzntFXS\n01Ue9nmlpFuDTxpSbZZ6ZjEetao8Fe+j7v5cSaMqDwWckvB4NFf5PyrpdElnq/wG7u9jK2GNgrnO\nr5T0rzN/lvBzIGnW8qfmHATJlItU7rT+uqR2SRfEWqhFmK38ZnapyknVZ0t6nsox+qrYCrkAd39I\n5Wmid0m6U9J9kiZnHJP4+2AhzdD3c/dJdz9b5RG056p8DWaKmV0o6ZC774m7LBF4obufI+llkt5q\nZi+q/mFGrtu094UWLeo+R9YTFq+X9Jng63/Vr4aTH1A5817RHTyWCmbWpnKD9Sl3r9TvYGUYefB/\nZVh9aus6Rz1lZm+QdKGk1wY3hZTiekqz1vV0lTuP95vZoyrX57tm9mtKcV3nOKcDkj4TDAX9tsqf\nNpyk7NUzi/FoQNKAu1dGP92mcqM9VzxKmlnL7+4Hg05zSdIOpWAqksqdwe+6+8Hg+7Scg4pp5U/Z\nOfgDSY+4+6C7H1X5Pu9TeXpXa3BMku/r2cr/gmAaggfT9T6hZJ8Dufs/ufs6d3+RpGGV13lI230w\np0X2/VIvGFJ/t6TnKz33Uq36JL0y6N99WuWpIP9T2atnZQRXZbrvLpXjSNau27T3hY5HpH2OrCcs\nHpf0+8HXL5H0o+DrOyRttLLfVXn4489n+wVJE3zy/E+SHnL3/1H1oztUfkOk4P/PVT2eurrOVU8z\nu0DlIXSvdPcjVU+5Q9KrzWyJmZ0m6QxJ346yzMdrtrq6+/fdfaW7r3b31SoHw3Pc/T+UsXMq6bMq\nL7wpMztT5cWKnlBKz+k89cxcPAqux8fMbE3w0PmS9mnueJQoc5Xfpq8hsl7l4eZJd4mmT6dIxTmo\nMq38KTsHP5P0u2a2LLj/K/fB3ZIuDo5J8jmYrfwPVXU+TeX5yEk+BzKzlcH/p6q8fsU/K333wayO\no++XSma2woLd38ysIOkPVV6vIy33Uk3c/T3u3h30714t6Svu/lplrJ5m1m5mHZWvVV44/wFl7LpN\ne1/oOEXb5/AErDLaiH/BH+3nko6q/ObuTZJeqPI80vtVnuu3LjjWJH1E5Xlx35fUG3f5F1HPF6o8\nzGavykMe71N5rmaXpC+r/Cbo/0p6eprrOk89H1Z5vn/lse1Vz3lvUM/9CnZjSMO/ueo645hH9atd\nQrJ2TvOSbla5EfuupJek+ZzOU8/MxaOg/GdL6g/q+1lJnXPFoyT+m6P8nwzOxV6VG+FnxF3OBerQ\nLmlI0tOqHkvTOZit/Gk7B38j6QdBHPukyrsbPUvlJOvDKo+qWhJ3ORdZ/q8E5+CBIEafEHc5F6jD\nN1R+k3C/pPODx1JzHyxQt0X1/dL6T1KPpO8F9XxA0vuCx1NzLx1Hnc/Tr3YJyVQ9g/rcH/x7UNJ7\ng8czdd0GdUp1X2iRdY28z2HBiwAAAAAAACRG1qeEAAAAAACAFCJhAQAAAAAAEoeEBQAAAAAASBwS\nFgAAAAAAIHFIWAAAAAAAgMQhYQEAAAAAABKHhAWanpmdZ2afr/r6BQsc/ydmtjaa0gFAcpjZajN7\nYBHH32hmF8/y+FTcBQA0hpm9wcz+cY6fPTnP81ab2WvCKxlw/EhYANOdJ2nehIWkP5FEwgIAYmRm\nrXGXAQAyYrUkEhZIJBIWyIQgM/yD4NO8H5rZp8zsD8xst5n9yMzONbN2M7vBzL5tZt8zs4tm/g5J\nl0u6wszuM7Pfm+V1XiDplZK2BsecHvy708z2mNk3zOzZwbE3mtlHzexeM/tJ8IniDWb2kJndWPU7\nnzSz68zsQTP7spmtCPFPBQD1ajGzHUHMusvMCmZ2dhDr9prZLjPrnPkkM7sgiNPflfSn872Amb3f\nzD5pZt8MYvim4PHzgjh7h6R9wWOXBnH9PjP7mJm1BP9uNLMHzOz7ZnZFcOw7zGxfUM5PN/5PAwD1\nqbFP+3Qz+2wQy+41s55Zfs9pQQz9vpn9twVe9oOSfi+Io1cEMXSrmX0neI03B7/zPDP7mpl9Lujb\nftDMXhvE4O+b2enBcTea2XYz6w/qcGHj/1JoFiQskCW/IenvJT07+PcaSS+U9F8k/bWk90r6iruf\nK+nFKicd2itPdvdHJW2XdJ27n+3u35j5Au5+j6Q7JF0ZHPNjSddLeru7rwtea1vVUzolPV/SFcHz\nrpP0W5J+28zODo5pl9Tv7r8l6WuSrmnA3wIAwnKGpI8EMeuwpA2Sdkq6yt17JH1fM+KYmS2VtEPS\nKyStk/RrNbxOj6SXqBxD32dmvx48fo6kd7r7mWb2m5JeJanP3c+WNCnptZLOlnSyu5/l7r8t6RPB\nc98t6blBOS8/rtoDQPgW6tP+jaTvBbHsr1WOwTP9T0kfDWLgzxd4vXdL+kbQt71O0psk/cLdnyfp\neZI2mdlpwbHPUTl+/qak10k6M+hbf1zS26t+52pJ50r6Y0nbg3YAWDQSFsiSR9z9++5ekvSgpC+7\nu6vceV4t6aWS3m1m90n6qqSlkk6t5wXN7ASVp5D8a/B7PybpGVWH/O+qMhycUb7VwTElSf8SfH2z\nyg0SACTVI+5+X/D1HkmnS1ru7l8LHrtJ0otmPOfZwfN+FMTEm2t4nc+5+5i7PyHpbpU7vpL0bXd/\nJPj6fJUTIN8JYvD5kp4l6SeSnmVm/8vMLpD0y+D4vZI+ZWaXSppYRJ0BIEoL9WlfKOmTkuTuX5HU\nZWYnzvgdfZJuCb7+5CJf/6WSNgZx9VuSulROVkvSd9z95+4+LunHku4KHq+UreJWdy+5+49UjsnP\nXmQZAEkS8z+RJeNVX5eqvi+pfK1PStrg7vurn2Rmq+p4zZykw8Ene/OVqbo81WWajddRHgAIW3Us\nm5S0PKTXmRkLK9+PVj1mkm5y9/fMfLKZPUfSH6n8SeCfS/oLlT/pe5HKIz3ea2a/7e4kLgAkzUJ9\n2qM1/p7j7VOayqOHvzjtQbPzaijbXK9N/xbHhREWaCZflPR2MzNJMrPnznLMiKSOBX7P1DHu/ktJ\nj5jZnwW/04JO8mLkJFVW0X+NpH9f5PMBIE6/kDRcte7P61Se3lbtB5JWV+Y3S7qkht97kZktNbMu\nlRdE/s4sx3xZ0sVmtlKSgnndzzSzkyTl3P12Sf9V0jlmlpN0irvfLekqSU+TdELNtQSA5PiGytPf\nKkmEJ4I+abXdkl4dfP3aBX7fzP7vFyW9xczagtc4s3oadY3+zMxyQdx/lqT9Cz0BmA0JCzSTD0hq\nk7TXzB4Mvp/pf0tab3Msuhn4tKQrrbxw5+kqNwJvMrP7VR62d9Ecz5vLqKRzrbxV4Esk/e0inw8A\ncXu9yusC7VV5/Yhpcczdn5J0maT/Eyy6eaiG37lX5akg90r6gLs/PvMAd9+nckLiruC1v6TytLyT\nJX01GM58s6T3SGqRdLOZfV/S9yR92N0PH09lASBm75e0Loh7H1Q5Bs/0TklvDWLeyQv8vr2SJs3s\n/mCR4o+rvLDxd4P+6ce0+JH5P5P0bUn/JunyoB0AFs3K06EAxMXMnnR3PuUDgICZvV/Sk+7+d3GX\nBQCwOFbeDe/z7n5b3GVB+jHCAgAAAAAAJA4jLIA5mNl7Jf3ZjIf/1d2vjaM8AJA1ZvZGlYctV9vt\n7m+NozwAkGVm9ts6dseQcXf/nTjKA9SChAUAAAAAAEgcpoQAAAAAAIDEIWEBAAAAAAASh4QFAAAA\nAABIHBIWAAAAAAAgcUhYAAAAAACAxPn/OrL8RNkUOFoAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 1080x360 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6cdWL23ENCfq",
        "colab_type": "code",
        "outputId": "d49b13b8-0244-427e-e395-33cee91a0d51",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 890
        }
      },
      "source": [
        "\n",
        "\n",
        "grid = sns.JointGrid( \"melt_temp\",\"warpage\", data = df,height = 12)\n",
        "\n",
        "grid.plot_joint(plt.scatter)\n",
        "grid.plot_marginals(sns.distplot)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<seaborn.axisgrid.JointGrid at 0x7f7c013ba048>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 100
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA18AAANYCAYAAAA7Q1fRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nA4GnKwd0exooAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 864x864 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ri0FFdIuvJMx",
        "colab_type": "text"
      },
      "source": [
        "## Separando o dataframe de treino e o dataframe de teste\n",
        "\n",
        "Os primeiros 45 valores são referentes ao valores de treino e os ultimos 10 aos valores de teste. Sendo assim, basta separa-los."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zID9dedMu4fz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Ultimos 10 linhas\n",
        "X_test = raw.drop(\"warpage\",axis=1).tail(10)\n",
        "X_test.set_index(\"teste\",inplace = True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VAKxQziMu-VX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#primeiros 45 linhas\n",
        "X_train =  raw.drop(\"warpage\",axis=1).head(45)\n",
        "X_train.set_index(\"teste\",inplace= True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zGcGi7h09pra",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_test=raw[\"warpage\"].tail(10).to_numpy()\n",
        "y_train=raw[\"warpage\"].head(45).to_numpy()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LURmCWaJxNLr",
        "colab_type": "text"
      },
      "source": [
        "## Pré Processamento"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hGw02AhFxRbo",
        "colab_type": "text"
      },
      "source": [
        "A rede neural em Python pode ter algumas dificuldades em covergir antes do número máximo de interações permitidas se os dados não estiverem normalizados. O perceptron de multiplas-camadas é sensivel a escala das features, dessa forma é recomendavel que elas estejam normalizadas. Uma vez que os dados não possuem outliers, o StandardScaler do próprio scikit-learn é uma ferramenta adequada para a normalização dos dados."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hkxu2R4OwDUw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "scaler = StandardScaler()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LfehNt6pypa8",
        "colab_type": "code",
        "outputId": "f3bda76b-9b88-4d4d-e5ba-a4dc5ffbbe8d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "scaler.fit(X_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "StandardScaler(copy=True, with_mean=True, with_std=True)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8h8QDnmly0Ix",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train = scaler.transform(X_train)\n",
        "X_test = scaler.transform(X_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vTpRNHDD85OL",
        "colab_type": "text"
      },
      "source": [
        "## Criando o regressor"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zaARBBXZzHMK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.neural_network import MLPRegressor"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4BfDuoC-9HFm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "mlp = MLPRegressor(hidden_layer_sizes=(5,3),\n",
        "                   max_iter=700,\n",
        "                   solver = \"lbfgs\",\n",
        "                   random_state = 42)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zyJgT8V0Ayt6",
        "colab_type": "code",
        "outputId": "b677b25c-c49a-47c0-caa6-d426d412d329",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 141
        }
      },
      "source": [
        "mlp.fit(X_train,y_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "MLPRegressor(activation='relu', alpha=0.0001, batch_size='auto', beta_1=0.9,\n",
              "             beta_2=0.999, early_stopping=False, epsilon=1e-08,\n",
              "             hidden_layer_sizes=(5, 3), learning_rate='constant',\n",
              "             learning_rate_init=0.001, max_iter=700, momentum=0.9,\n",
              "             n_iter_no_change=10, nesterovs_momentum=True, power_t=0.5,\n",
              "             random_state=42, shuffle=True, solver='lbfgs', tol=0.0001,\n",
              "             validation_fraction=0.1, verbose=False, warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VlURJMy_A20N",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predictions_test = mlp.predict(X_test)\n",
        "predictions_train = mlp.predict(X_train)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DTdyxZ6aDIxX",
        "colab_type": "code",
        "outputId": "2e916c9e-a437-45bb-b6df-43e00363f903",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "predictions_test"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1.43578665, 1.51975274, 1.54105277, 1.42460129, 1.34500927,\n",
              "       1.41011028, 1.29834162, 1.3331364 , 0.95444932, 1.1481733 ])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 50
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "02NrmRoDDQM2",
        "colab_type": "code",
        "outputId": "14bda317-4bd0-4bfa-97b1-b717da8e1343",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "y_test"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1.44 , 1.527, 1.557, 1.417, 1.328, 1.396, 1.28 , 1.312, 1.05 ,\n",
              "       1.162])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wLch0seLA8Es",
        "colab_type": "code",
        "outputId": "7de4e9c6-421b-46cb-dae1-8edd5881f828",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn import metrics\n",
        "\n",
        "r2_test = metrics.r2_score(y_test,predictions_test)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9503298709006565"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5RP1eM3xA9QA",
        "colab_type": "code",
        "outputId": "49a21420-c796-45d9-cd69-ea1bdf0cf681",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "r2_train = metrics.r2_score(y_train,predictions_train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9941764091705162"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jG5f60QeoTWk",
        "colab_type": "text"
      },
      "source": [
        "O coeficiente de determinação($R^2$) é uma forma de se avaliar o ajuste do modelo. Em linhas gerais é uma forma de quantificar o quanto o modelo foi capaz de explicar os dados coletados. Vale ressaltar que:\n",
        "\n",
        "$0 \\leq R^2 \\leq 1$ \n",
        "\n",
        "Dessa forma, pode-se observar que ambos para o dataset de treino, quanto para o dataset de test, a rede neural foi capaz de predizer o rechupe com elevado coeficiente de determinação."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_rOJqpNIGFzr",
        "colab_type": "code",
        "outputId": "35c412c0-8f7c-4b6a-847d-b5401266b407",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        }
      },
      "source": [
        "\n",
        "#The weight matrix at index i holds the weights between the layer i and layer i + 1. https://www.python-course.eu/images/mlp_example_layer.png\n",
        "print(\"weights between input and first hidden layer:\")\n",
        "print(mlp.coefs_[0])\n",
        "print(\"\\nweights between first hidden and second hidden layer:\")\n",
        "print(mlp.coefs_[1])\n",
        "print(\"\\nweights between second hidden and result:\")\n",
        "print(mlp.coefs_[2])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "weights between input and first hidden layer:\n",
            "[[ 0.07001345  0.89787471 -0.32653017 -0.37640497 -0.83223389]\n",
            " [-0.63099473 -0.97344795  0.42708507  0.34677342  0.69222905]\n",
            " [-0.9073803   0.29512139  0.86688819 -0.17088574 -0.26087691]]\n",
            "\n",
            "weights between first hidden and second hidden layer:\n",
            "[[ 0.11234135 -0.28616841 -0.37218335]\n",
            " [ 0.36105603 -0.20679787  0.17747356]\n",
            " [-0.12967757  0.61806258  0.27368371]\n",
            " [-0.35675513 -0.15442684 -0.69582342]\n",
            " [-0.60760199  0.95991727  0.62966412]]\n",
            "\n",
            "weights between second hidden and result:\n",
            "[[ 0.2565208 ]\n",
            " [-0.0799621 ]\n",
            " [-0.95474526]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2-olkHEBzQtn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}