Visualizing_tweets_and_the_Logistic_Regression_model_ipynb.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyOVamLOXXcSndYFnDg5i6vy",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/hussain0048/Natural-language-processing/blob/main/Visualizing_tweets_and_the_Logistic_Regression_model_ipynb.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Objectives:**"
      ],
      "metadata": {
        "id": "fKHv_IqjaMJF"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Visualize and interpret the logistic regression model"
      ],
      "metadata": {
        "id": "nLHbvflCaVu9"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Import the required libraries**\n"
      ],
      "metadata": {
        "id": "fraRKpp2d4QC"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "trgYpW7FYidf"
      },
      "outputs": [],
      "source": [
        "import nltk                         # NLP toolbox\n",
        "from os import getcwd\n",
        "import pandas as pd                 # Library for Dataframes \n",
        "from nltk.corpus import twitter_samples \n",
        "import matplotlib.pyplot as plt     # Library for visualization\n",
        "import numpy as np                  # Library for math functions\n",
        "#from utils import process_tweet, build_freqs # Our functions for NLP"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Load the NLTK sample dataset**\n"
      ],
      "metadata": {
        "id": "aux68uzdemGA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# select the set of positive and negative tweets\n",
        "import nltk\n",
        "nltk.download('twitter_samples')\n",
        "all_positive_tweets = twitter_samples.strings('positive_tweets.json')\n",
        "all_negative_tweets = twitter_samples.strings('negative_tweets.json')\n",
        "\n",
        "tweets = all_positive_tweets + all_negative_tweets ## Concatenate the lists. \n",
        "labels = np.append(np.ones((len(all_positive_tweets),1)), np.zeros((len(all_negative_tweets),1)), axis = 0)\n",
        "\n",
        "# split the data into two pieces, one for training and one for testing (validation set) \n",
        "train_pos  = all_positive_tweets[:4000]\n",
        "train_neg  = all_negative_tweets[:4000]\n",
        "\n",
        "train_x = train_pos + train_neg \n",
        "\n",
        "print(\"Number of tweets: \", len(train_x))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oUDEGFGWekIL",
        "outputId": "cfe616a0-f9f5-448f-d6f6-eee75bd965e8"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[nltk_data] Downloading package twitter_samples to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/twitter_samples.zip.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of tweets:  8000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Load the extracted features**"
      ],
      "metadata": {
        "id": "9rdqRfaEfrJY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gj35hz1zsW5a",
        "outputId": "a9a2e66b-71f6-4738-e4e6-2a9c7ee86fee"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd \n",
        "data = pd.read_csv('/content/drive/MyDrive/Courses /Data Science /NLP/Natural Language Processing Specialization/Natural Language Processing with Classification and Vector Spaces/Dataset/logistic_features.csv',encoding = 'latin')"
      ],
      "metadata": {
        "id": "IMS1G5szshsZ"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "data.head(10) # Print the first three data entries"
      ],
      "metadata": {
        "id": "MDUIyyhus5L8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Each feature is labeled as bias, positive and negative\n",
        "X = data[['bias', 'positive', 'negative']].values # Get only the numerical values of the dataframe\n",
        "Y = data['sentiment'].values; # Put in Y the corresponding labels or sentiments\n",
        "\n",
        "print(X.shape) # Print the shape of the X part\n",
        "print(X) # Print some rows of X"
      ],
      "metadata": {
        "id": "2lmbKS3ut2tS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Load a pretrained Logistic Regression model**\n"
      ],
      "metadata": {
        "id": "aR4JQ3eluWHP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "theta = [7e-08, 0.0005239, -0.00055517]"
      ],
      "metadata": {
        "id": "ZOG_vAx2ufDg"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Plot the samples in a scatter plot**\n"
      ],
      "metadata": {
        "id": "rxO2fV-6uoFJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the samples using columns 1 and 2 of the matrix\n",
        "fig, ax = plt.subplots(figsize = (8, 8))\n",
        "\n",
        "colors = ['red', 'green']\n",
        "\n",
        "# Color based on the sentiment Y\n",
        "ax.scatter(X[:,1], X[:,2], c=[colors[int(k)] for k in Y], s = 0.1)  # Plot a dot for each pair of words\n",
        "plt.xlabel(\"Positive\")\n",
        "plt.ylabel(\"Negative\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 514
        },
        "id": "LhRU-ZIIuw__",
        "outputId": "b43c9b0f-7264-41f3-acda-f33e417a24f6"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Negative')"
            ]
          },
          "metadata": {},
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgEAAAHgCAYAAAA8Fr7bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5gV5d3/8feXtiC9SZUigsaOYu8msT55ME+KLepjYkmipptomqZcifGJJcYWa8wvscSuiYrGGDUWFJRQpIgoCiJFOuzCwt6/P2bQpS9sOSzzfl3Xuc4595TznWHZ89mZe+6JlBKSJKl4mpS6AEmSVBqGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgmpW6gIbWpUuX1K9fv1KXIUlSgxg5cuSclFLXdU0rXAjo168fI0aMKHUZkiQ1iIiYur5png6QJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAF17frr4c9/LnUVkiRtVLNSF7DVOfJIaNWq1FVIkrRRhoC6ttNOpa5AkqQa8XSAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFVW8hICK2i4hnIuKNiBgXEd/M2ztFxFMR8Wb+3DFvj4i4JiImR8ToiNir2rrOyOd/MyLOqNa+d0SMyZe5JiKivrZHkqStTX0eCVgBfDeltDOwP3BeROwMXAQ8nVIaCDydvwc4FhiYP84BboAsNACXAPsB+wKXrAoO+TxnV1vumHrcHkmStir1FgJSSjNSSq/lrxcB44FewFDgjny2O4AT8tdDgT+lzMtAh4joARwNPJVSmptSmgc8BRyTT2uXUno5pZSAP1VblyRJ2ogG6RMQEf2AwcBwoFtKaUY+6QOgW/66F/BetcWm5W0bap+2jnZJklQD9R4CIqINcD/wrZTSwurT8r/gUwPUcE5EjIiIEbNnz67vj5MkqVGo1xAQEc3JAsBfUkoP5M0z80P55M+z8vbpwHbVFu+dt22ovfc62teSUroppTQkpTSka9eutdsoSZK2EvV5dUAAtwLjU0pXVpv0CLCqh/8ZwMPV2k/PrxLYH1iQnzYYBhwVER3zDoFHAcPyaQsjYv/8s06vti5JkrQRzepx3QcBpwFjImJU3vZD4DLgrxHxFWAq8MV82mPAccBkYClwJkBKaW5E/AJ4NZ/v5ymlufnrrwN/BFoBj+cPSZJUA5Gdli+OIUOGpBEjRpS6DEmSGkREjEwpDVnXNEcMlCSpoAwBkiQVlCFAkqSCMgRIklRQhgBJkgrKECBJUkEZAiRJKihDgCRJBWUIkCSpoAwBkiQVlCFAkqSCMgRIklRQhgBJkgrKECBJUkEZAiRJKihDgCRJBWUIqCsrVkBFRamrkCSpxgwBdeWOO+Cqq0pdhSRJNdas1AVsNb7wBSgvL3UVkiTVmEcC6kq7dlBWBpdfDosXl7oaSZI2yhBQl5o3z8JAMw+wSJK2fIaA2po4ES64AJYsgdat4atfhZYtS12VJEkbZQiorf79oUMHuOyyUlciSdIm8bh1bbVoAd//PsyYUepKJEnaJIaAutC2bfaQJKkR8XSAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQdVbCIiI2yJiVkSMrdZ2aURMj4hR+eO4atMujojJETExIo6u1n5M3jY5Ii6q1t4/Iobn7fdERIv62hZJkrZG9Xkk4I/AMetovyqltGf+eAwgInYGTgJ2yZe5PiKaRkRT4DrgWGBn4OR8XoDf5OvaAZgHfKUet2XTjBsHL79c6iokSdqgegsBKaXngLk1nH0ocHdKaVlK6W1gMrBv/picUpqSUloO3A0MjYgAjgTuy5e/AzihTjegNt55B8aPL3UVkiRtULMSfOb5EXE6MAL4bkppHtALqP6n87S8DeC9Ndr3AzoD81NKK9Yxf+kdf3ypK5AkaaMaumPgDcAAYE9gBnBFQ3xoRJwTESMiYsTs2bPr50MWLYJrr4WKivpZvyRJdaxBQ0BKaWZKaWVKqQq4mexwP8B0YLtqs/bO29bX/iHQISKardG+vs+9KaU0JKU0pGvXrnWzMWtauTILAFVV9bN+SZLqWIOGgIjoUe3tZ4FVVw48ApwUEWUR0R8YCLwCvAoMzK8EaEHWefCRlFICngE+ny9/BvBwQ2zDenXoAN/7HmyzTUnLkCSppuqtT0BE3AUcDnSJiGnAJcDhEbEnkIB3gHMBUkrjIuKvwBvACuC8lNLKfD3nA8OApsBtKaVx+Uf8ALg7In4JvA7cWl/bIknS1iiyP6qLY8iQIWnEiBF1s7KqKpg6Ffr3r5v1SZJUxyJiZEppyLqmOWJgbbz+OvzkJ1l/AEmSGhlDQG3svTfccAM0bVrqSiRJ2mSGgNpq27bUFUiStFkMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhCwJXvySXj00VJXIUnaShkCtmT33w8PPFDqKiRJW6lmpS5AG/CHP5S6AknSVswjAZIkFZQhQJKkgqpxCIiIvhHxqfx1q4hoW39lSZKk+lajEBARZwP3AatOUvcGHqqvoiRJUv2r6ZGA84CDgIUAKaU3gW3rqyhJklT/ahoClqWUlq96ExHNgFQ/JUmSpIZQ0xDwbET8EGgVEZ8G7gUcxUaSpEaspiHgImA2MAY4F3gM+HF9FSVJkupfTQcLOgH4U0rp5vosRpIkNZyaHgn4DDApIv5fRPxX3idAkiQ1YjUKASmlM4EdyPoCnAy8FRG31Gdhjdry5TBmTKmrkCRpg2o8WFBKqRJ4HLgbGEl2ikDrMnw4XH99qauQJGmDanRYPyKOBU4EDgf+BdwCfLHeqmrsDjkEDjig1FVIkrRBNT23fzpwD3BuSmlZPdaz9WhmtwlJ0patRt9UKaWT67uQrdrixdCmTamrkCRpNRvsExAR/86fF0XEwmqPRRGxsGFKbOSmTYMzz4S5c0tdiSRJq9ngkYCU0sH5s3cM3Fy9e8Nll0GnTqWuRJKk1dT0LoL/ryZtWo8BA0pdgSRJa6npJYK7VH+TDxa0d92XsxVLCR57LBtDQJKkLcDG+gRcHBGLgN2r9wcAZgIPN0iFjcWVV8KkSeufXlEBTz4J773XcDVJkrQBGwwBKaVf5/0B/i+l1C5/tE0pdU4pXdxANW75li6Fjh3h6adh5cp1z9OqFVx9tacGJElbjJoOG3xxRHSMiH0j4tBVj/ourlG44go44QTYfffsSoAlS0pdkSRJNVLTEQPPAr4J9AZGAfsDLwFH1l9pjcTdd2cDA915ZxYIJElqJGraMfCbwD7A1JTSEcBgYH69VdVYPPwwTJ8OXbvCNtuUuhpJkjZJTce2rUgpVUQEEVGWUpoQETvWa2WNweOPQ7t2cNNNjgMgSWp0ahoCpkVEB+Ah4KmImAdMrb+yGonrroMVK6CsrNSVSJK0yWp674DP5i8vjYhngPbAE/VWVWPRtGn2kCSpEappx8Dqx7rH5M+p7suRJEkNpaYdA18DZgOTgDfz1+9ExGsR4ciBkiQ1QjUNAU8Bx6WUuqSUOgPHAn8Dvg5cX1/FSZKk+lPTELB/SmnYqjcppSeBA1JKLwP2ipMkqRGq6dUBMyLiB8Dd+fsTgZkR0RSoqpfKJElSvarpkYBTyEYLfAh4ENgub2sKfLF+SpMkSfWpppcIzgEuiIjWKaU1B8efXPdlSZKk+lajIwERcWBEvAGMz9/vERF2CJQkqRGr6emAq4CjgQ8BUkr/AbyLoCRJjVhNQwAppffWaFpZx7VIkqQGVNOrA96LiAOBFBHNye4qOL7+ypIkSfWtpkcCvgqcB/QCpgN75u/XKyJui4hZETG2WluniHgqIt7Mnzvm7RER10TE5IgYHRF7VVvmjHz+NyPijGrte0fEmHyZayIiar7ZkiSpRiEgpTQnpXRqSqlbSmnblNKXUkofbmSxPwLHrNF2EfB0Smkg8HT+HrIRCAfmj3OAG+CjexZcAuwH7Atcsio45POcXW25NT9LkiRtwAZPB0TETzcwOaWUfrGBic9FRL81mocCh+ev7wD+Bfwgb/9TSikBL0dEh4jokc/7VEppbl7PU8AxEfEvoF0+YiER8SfgBODxDW2PJEn62Mb6BKw5JgBAa+ArQGdgvSFgPbqllGbkrz8AuuWvewHVOx5Oy9s21D5tHe2SJKmGNhgCUkpXrHodEW3JOgSeSTZ88BXrW64mUkopIhrkdsQRcQ7ZaQb69OnTEB8pSdIWb6N9AvLOfL8ERpOFhr1SSj9IKc3ajM+bmR/mJ39etY7pZEMRr9I7b9tQe+91tK9TSummlNKQlNKQrl27bkbZkiRtfTYYAiLi/4BXgUXAbimlS1NK82rxeY8Aq3r4nwE8XK399Pwqgf2BBflpg2HAURHRMe8QeBQwLJ+2MCL2z68KOL3auiRJUg1srE/Ad4FlwI+BH1W7Ci/Ijui3W9+CEXEXWce+LhExjayX/2XAXyPiK8BUPr750GPAcWT3IVhKdsqBlNLciPgFWRAB+PmqToLA18muQGhF1iHQToGSJG2CyDrkF8eQIUPSiBEjSl2GJEkNIiJGppSGrGtajYcNliRJWxdDgCRJBWUIkCSpoAwBkiQVlCFAkqSCMgRIklRQhgBJkgrKECBJUkEZAiRJKihDgCRJBWUIkCSpoAwBkiQVlCFAkqSCMgRIklRQhgBJkgrKECBJUkEZAiRJKihDgCRJBWUIkCSpoAwBkiQVlCFAkqSCMgRIklRQhgBJkgrKECBJUkEZAiRJKihDQEP4+9/hr38tdRWSJK2mWakLKIRu3aCiotRVSJK0GkNAfXrnHXjySTjnnHVPLy+Hli0hokHLkiQJDAH1KyWoqlr/9B/+EPbfH6ZNg69/HVq1arjaJEmFZ5+A+tS/P3z1q+uffv758MlPQlkZNPGfQpLUsDwSUN8qK2H+fOjade1pAwZkz+ef37A1SZKEIaB+PPkkzJwJu+8OkybBiBHQr1/2/qCDSl2dJEmApwPq3rJl8NBDMHs2jB+f/bX//vvQo0d2lYAkSVsIQ0BdGzYMnngiOxLwyCNZp7+vfAWGDoUddih1dZIkfcTTAXXpoYeyXv6dO2edAo88Eg47LLsM8NFHs1MBnTqVukpJkgBDQN1YuTIbE+Bvf8vev/ACtGsH3/1udnrgE5+AUaOgd29DgCRpi+HpgNp68UXYdVfYZ5/sNEDbtvDAA9m0X/8a2rSBK6+En/wEBg+u2Tq3267+6pUkKWcIqI2JE7Nz/bNmZZcALlsGM2bA009n01u0yE4JXHVVzdcZkfUj6NKlfmqWJClnCKiNCRNgm23goovgM5/JXvfqBTvuCJde+vF8LVvWfJ2TJ2fPc+bUaamSJK3JPgG1MXRodrj/zDOzQYF22QXat8+OCmzuIf0BA7LhhiVJqmeGgNq4804466zssH8EHHpodhh/2TI47rhSVydJ0gYZAmqjR4/sy79HDzjwQDj7bAcEkiQ1GoaA2vjGN7Ib/1x7bXYjIEmSGhFDQG3cdx988EE2IJAkSY2MIaA2dtwxe0iS1Ah5iaAkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQXVrNQFaCNuuil7PuUUaNOmtLVIkrYqHgnY0n3603DAAfDlL8Pbb5e6GknSVqQkISAi3omIMRExKiJG5G2dIuKpiHgzf+6Yt0dEXBMRkyNidETsVW09Z+TzvxkRZ5RiW+pd//6w227wq19Bv36lrkaStBUp5ZGAI1JKe6aUhuTvLwKeTikNBJ7O3wMcCwzMH+cAN0AWGoBLgP2AfYFLVgWHrdIOO0BEqauQJG1FtqTTAUOBO/LXdwAnVGv/U8q8DHSIiB7A0cBTKaW5KaV5wFPAMQ1ddK38+tcwatTH719+GZYv//j9hAnwn/80fF2SpEIoVQhIwJMRMTIizsnbuqWUZuSvPwC65a97Ae9VW3Za3ra+9i3f++/DK6/AgAFw//0wezYsWAB33AFjxnw83+jRMHJk6eqUJG3VSnV1wMEppekRsS3wVERMqD4xpZQiItXVh+VB4xyAPn361NVqN91TT8E++2R//Y8fD9/4Brz7Lvz5zzBtGtxwQxYOxo2DXXaBL36xdLVKkrZ6JTkSkFKanj/PAh4kO6c/Mz/MT/48K599OrBdtcV7523ra1/X592UUhqSUhrStWvXutyUTfPSS/DGG3DccfDd78If/wiTJsGyZfCd72TzjB8PU6aUrkZJUmE0eAiIiNYR0XbVa+AoYCzwCLCqh/8ZwMP560eA0/OrBPYHFuSnDYYBR0VEx7xD4FF525Zp/Hj4n/+BAw/M3peXw/XXQ6dO0KtX9gBo3x523LF0dUqSCqMUpwO6AQ9G1tO9GXBnSumJiHgV+GtEfAWYCqw6Fv4YcBwwGVgKnAmQUpobEb8AXs3n+3lKaW7DbcYmmD4dfv7zrOPfOedkh/xPOQUqKmDQILjrLhg6FNq1g1mzoHfvUlcsSSqASKnOTr03CkOGDEkjRoxo2A/9y1/ggw9g4EBo3Rp+8hNICSZPhnnzspEAb7stO1IgSVIdioiR1S7HX43DBtenOXOyS/wmToTbb4du3bJR/yoqoFkzaN4cmjTJwsHs2aWuVpJUMIaA+jB7dnZ536OPwuOPZ1/6ixdnRwOaNIGWLWG77WDq1KyPwE47ZacJJElqQFvSYEFbjwkTYMQIOPbY7Eu+ZUvYfntYsSJ7dO+ezReRnRa48cb1jwY4dy4M23L7O0qSGi9DQF2qqsqeu3TJLvO77rpsMKC33souBVzl+OPhE5/IjgocfnjWllJ2yeDs2VkgWBUKJk2CV19FkqS65umAujJ5MvziF9llfzNnwj33QGUlLF2aTV+8OHuuqsq+7Jctg099KgsB//pXdqRg1ixYsmT19e6/f/aQJKmOGQLqyvbbw3nnwXPPZVcDHHNMFgRWKSuDlSuzv/4/+cls0KC99so6CH71q1nnwIceyuYt2BUbkqTSMATUlcWLs7/kzz4bFi1a+y/6ZcuyQ/xlZdkh/jZtYNdds+XatYMDDihN3ZKkwrJPQF0ZNgwuvDAbCXDNALBKStmQwR98kN0waOZMOPRQ+Pa34fvfb9h6JUmF55GAutKuXXYb4AUL1j9Pz57Z+f3p07POgZ06ZZ0GH3wwG0Ogb1/o2hX+/W/o18/hgyVJ9coQUFfuv3/DN/7p1w8WLsyuGJg2DcaOze4U+LnPZZ0DDzkkG0lwjz1g/vzsToOf/CQMWecgT5Ik1ZohoLZSgrPOyob93ZC5c6F/f9h3X9htt2wwoSFD4NxzP57npz+FFi2yvgP33pt1JJQkqZ7YJ6C2KiuzS/42pEmT7NGuXdZf4JZbsisDunXLwsEqZWUfjw/whS/AfvvVW9mSJBkCautLX/p4kKD1ifi4w2DLltnlgGeeCWPGZKcG1mXlSi8VlCTVK0NAbfz2t9lh+4353OeyPgHTpmX9Au67LwsDP/kJ7L77upf5xS+y+SRJqif2CaiNCy/c+Dzdu8Npp2V3DGzXLusYWBOnnJKdLpAkqZ54JKA2vvnNDU9v1gwuuSQ7tD94MJx00vpvFLSmQYOgffva1yhJ0noYAmrjd7/b8PTHH4e//S3r5LfDDtnAQJIkbSE8HVDXmjTJbiG8117ZDYKaNYNWrbJr/iVJ2oIYAmqhCoj8AWRf+o8/nn3xr3L44R/fLliSpC2IIaAWVpKdT2nauze8+27Nz/dLkrQFMATUQnOv45ckNWJ2DJQkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYASZIKyhAgSVJBGQIkSSooQ4AkSQVlCJAkqaAMAZIkFZQhQJKkgjIESJJUUIYAbRFefO9FVlStKHUZklQohgCV3LIVy7j19VuZPHdyqUuRpEJpVuoCpLJmZdz637eWugxJKhyPBGiLccOrN/DkW0+WugxJKgyPBGiLsXu33enWplupy5CkwjAEaItxUJ+DSl2CJBWKpwMa2h13wOjRpa5CkiRDQJ35/e9hxIiNz1dWBs2br90+Zw5cdhmUl9d9bZIkrYMhoK706wft2sGPfwwVFeuf76ST4BOfWLu9rAw6doRmnqGRJDUMQ0Bd+cxnoGVLePhh+OIXYcwYuOIKWLmyZsu3bQvnnrvuowRrGDdrHMMmD6tlwZKkovPPzrrUpw/85z/w739D9+7Qvj1E1PnHzKuYx/RF0+t8vZKkYvFIQF1r0gQOPRS6doWzzsrer+kf/4AJE2DJkvWvZ8wY+Na3YMXaQ+ke3Odgvjz4y3VY9OZ7YvITXP7C5Wu13/767dz7xr0lqEiSVFMeCahr990HPXrAQeu43O3ll2HIEHjsMVi4EI46Kjt1sC6DBsF//dcW30dgcPfBtC9rv1b7nt33pGWzliWoqP48+86zDOg0gN7tepe6FEmqE1v2N0xjc9NNWR+Avn0/bkspOyWweDFcdVXWb+CKK7I+AC1arL78vHkwahQccUTWUfBTn2rY+jdDtzbd1jnAz+Aeg0tQTf16bcZrVKUqQ4CkrYanA+pSq1bZX/f77JO9X7IETjgBHnoI2rSBO++EE0+ETp2yL/k1+wuMHg0vvNDwdTeQD5d+yMzFM0tdxmb79gHf5oj+R5S6DEmqM4aAunTaaTBgwMfvFy6Enj1h//2z902bwnHHrbufAMBhh2WXGDZij7/5OEsrl65z2j3j7uHOMXc2cEWSpPUxBNSlVR39liyBmTOzvgE/+AH8/e9Zf4D/+z944431Lj565miWrVi2VvtdY+7i/jfur6+qN9mS5Uu4eeTNa9Valap45p1nmDJvymrtt712G7tdvxsn7XIS5+97fkOWKknaAENAXZkxA/73f2HWrGxAoNNPz/oHPP003HgjXHttNqLgwoXrXcWvn/81971xH/946x9UpaqP2gd2Gsi2rbdtgI2omWUrlzFn6RwqqyqZvWQ233riW0xfOJ0PFn/A5Z++nF233RXIrhwYM3MMgzoPoss2XWjVvBXNm258HATVzoqqFdw55k4qVmxg0CpJwhBQew88AC++mP3V/9OfwnnnwauvwpQpsOuucMEF2XgBvXplIwm2a7fWKi595lKueukqqqhi+qLp/OiZH/HajNcAWFq5lBenvcgDEx5o6C1bTUqJV6e/SkqJZSuW0bFlR1ZWreSF915gz+578qvnf8XVL1/90fw3vHoDt4y8hQXLFrB0xVJO2e2Uta4WqFxZyfKVyxt6U2pkxPsjWFlVw4GetjDlleW8MfsNFlQsKHUpkrZwhoDaat4cbrklO+z/8MPZX/vbbANTppDefJPKFSuoemtyFgoOOQTGjVtrFVMXTOVXz/+KU3c5le8f9H3+dca/GNAx61vw3NTneHfBu3z3gO+utsycpXOoXFm5SaVOWziNB8c/uFmb+cTkJzj+zuO5ceSNjHh/BJe9cBnjZo/j9Rmvs2zFMhYtW8RxOxzH8GnDuWb4NUz6cBJ79dyLg/sczHNTn+Ppt59e6y/Tm0bexHWvXAdkYWdL+dJdvnI5175yLRM/nFjqUjZL27K2/PLIX3pbZkkbZQiorc98BoYPh+uvhwcfhEWL4O23oaqKqhYtWNS2OZXbtIT+/bMRBd9/P7tE8PnnP1rFX8b8hb7t+1JRVcH8ivkMvWsoe96wJ1PnT2X7jtvTtkVberbtudrHXvHSFTwwftOODsxeMpupC6Zu8ib+fvjvadakGQM6DuBXz/2KqlTFgooFXPHiFXRp3YVbXruF8spy5i+bT4+2Pejdrjcrqlbwt0l/4+EJDzN0x6Hc/fm7adW81WrrPXm3kzl191MBuPyFyxu00+Di5Yu54LELeGf+O2tNa9G0BbcNvY2du+7cYPXUt7+M/guPv/l4qcuQtIVxnIDamj07+yu/ogJee221SU1PO41Ov/pVdrsPuQYAABWFSURBVJSgffuPBwb6xz9ghx0+mu+mz9zE+wvfZ+8eexMEKSW2abENFz99Mdu23pYZi2YwYfYEdt724y+lCw+8kHZl2amFpZVLuf312ylfUc5Ze51Fh5Yd1lnq4B6DN+v6/YcnPMyMXjN4/NTHOf2h05kwewLlleU8NvkxFi1bxMwlM2kSTZi2cBon7HQC7cvac8trt7BDpx0Y+f5I2vZvy9NTnmbMjDF0atOJI/odwXbtt6NTq04AfPvhb3PCniewT899Nrm2zdW6eWuOH3T8eq/5bxJbVz7u37E/2zTfptRlSNrCGAJqY+XKbFyAynUclu/SJQsId98N++zDu8cexOQp/2Ty3Mmc86lzPppt2ORhfLL/Jznw1gNZvmI5lxxxCU+d8RQ3j7yZB8Y/wKBOg/jX2//in2//k+5tuzNs8jBO2vWkj75AAZatWMaH5R/SsllLgo/HHrh55M0M6Tmk1gP3HDvwWN788E1+N/x3PPvOszzx5hNs02IbKldU8sw7zwDQpkUblixbwu9f+T1tmrehvLKc5cuX84vnf8GTbz3J8PeHA3DNsdesdQnh1aOu5upRV5MuSbWqc1NEBMfscEyDfV6pHbjdgaUuQdIWyBBQG9/+djbC37osWkR67TUYNYo46CBuHzSDnV56kz4HHQnAyqqVvLvgXX7zwm+47rjrmF8xnytevoIBnQcwpOcQdu+2O7t03YUXp71I9zbdaV3Wmj//589MmT+FE3c9cbUv+46tOvLTw366VgmdWnWidYvWm715k+dO5tGJj/KJLp/gxhE38ta8t0hkX9RLli1h2zbbMmvJLLq36c7SyqW8Ne8t+nfsz8vvvcw2zbehXat2DOgwgOMGHEfLpi15d8G7nLv3uTw04SF26LQDTZs0/eiztttmu82us8jmlc/jxfde5PhBx5e6FEmNkCGgNvr0WbutSZNsdMAIZjep4O1P7sV+v7+ZPuPv4u0RN7DHgOyQ/tf+9jX6dujL6x+8zvef+j6DewymTbM2PDT+IW59/VYG9xjMCTuewN1j7+ZLu32J56c+z/NTn+eYHY6p8aHqz+38uc3etPkV81m6fCkfLP6At+e/zeR5k1ebPnTHoYybM442LdqwsGIhZU3LuH/8/Ry03UGctsdpdGrViclzJ9OuZTuefe9Z9uq1Fz878mfMWTqH1z94naMHHE37ltk9BxryCMDWZsq8KYycMdIQIGmzRErF+gU8ZMiQNGLEiLpZ2fvvZ5f+VdeuHXzpSzBoEMtuv5VF3Tvy2JAOXN37PcbOGsvFh1zMRQdfxL4370v7Fu2ZtmgaFSsqWLx8MVWpip8e/lOG9BzCs+88y/yK+cxaMovmTZrTr0M/+nfszym7nbJWB7s1zVoyi86tOq/2l3ZNVa6s5PzHzmfW4llMmTeFKfOnsLhy8VrztW/Rnj267cGkuZOYt3QeTaIJ5VXlHNbnMPq078Or01+lrFkZJ+56Il8e/GUSib+M/gvtytpx2h6nNejNhe4eezetW7TmM4M+w00jbyIIzt777E1eT3llOX8Y+QfO3PPMjwKMJG3pImJkSmnIuqZtXb2fGtrXvrb6+whWLF5I+R23svDNcbx12nHMXDyTbf72OI+e8ijvf+99xs4ay2OTH2N+xXxenP4i7y96n7lL57JkxRLKV5Zz3SvX8eD4Bxk3exzHDzye646/jq/s9RUqVlYwdcFUWjVvxQvvvvDRlQH3j7+fsbPGAvDQhIe4+uWr+c0Lv+HRSY9u1ia9t+A9/vXOv3hjzhuMnj16nQEAYMHyBbz43ovMXDKTZWkZ5VXltI22/OSwnzB1/lTeXvA27cracepup9KtTTe6t+nOibueyPadtufsR89u0MsB+3XoR6+2vZi9ZDavzXiN/XvvT/XwO+nDScxYNAOAN2a/wYaCcRDEmvd8kKRGytMBtfHIIx+/7tkTunRh+YSxfNi7Mx8+fS+/v2A/un//GMYtnEyz91/lhJ1O4Bv7fYPz/n4ec5bMIZGoTFmnwiBo3qQ5lSsreeLNJ+jdvjeDewzmuuHXsXv33bn99dsZ3D3r4Ne8aXPKmpYB2RC+qwbceWvuW7wx5w0u/9Tl671CYGMemvgQxw86nmtevmaj865gxWrvF6VFnPPoOSyqWMTR2x9N520689oHr9GnQ3bapHe73vRs25NebXtt1lGKzbV/7/1ZWbWSe8bdw8Q5E7l33L3cMOIGPrvTZ/n0gE9z19i7WFC+gF9+8pf8/NmfM6TnELbvsD3/s/P/rLaeaQuncXi/wz+6KkOSGjtDQF347Gfh2GOZft1v6N55W7rNWMA/v3MU371zLO/1X8RpP7qcP057lIrlFVz+4uWMm7P2gEEAXVp24dTdT+Wxtx5j6fKl9L+6P0Q2bHBlVSWVKytZvHwx+/ba96NlTt/j9I9en7/v+SxctpDO23TerM1IKTFh9gRufv1mOi+BC1+Ayw+GuTW8sqxpNOWDxR+wdMVSFi5fSMvmLfnU9qvfDrlJNGGnLjttVn2bY/rC6UxdMJW7x95NVapiSM8hjJ41mpN3PZntO27PjSNu5NP9P82v//1rbn/9dq4++mqmLZpG8yZrD2886oNRzK+Yzx7d92iw+qurSlWklBo0QEnauhkCaiMl+O1v4Xe/Y+5rL9Lu3Zk0SfCdY4I/lz3LHsdty2mTy+kdMG3+NH77wm9ZyceHwY+aDC/3hoUtsy/Qfp368YfX/sDi5Yvp2aYn5SvLadmkJS2btuSQPofQp32fDV7rXdasjK7Num7WppRXljNsxL28fd/NMADmt4QRvWBBWc2Wb0YzfnjID3lvwXsAfO+A79G5dWfatGizWfXUlbGzxjJ65mi6te5Gn3Z9+Mfb/2Bl1Ur26L4HLZq2oHJlJXv22JND+x7KrCWzWJFWMKTnOk+d8YVdvtDA1Wfmlc9j7KyxjJs9jooVFXxr/2+VpA597Jrh13BEvyPYrdtupS5FqhVDQG316wfNmtF+5hzebwkT28DteybazJ7LL0Z25IOff4s/vfsI979x/2oBAODAd2FeS3i1N7Rq1oqp86ZSXlnOoE6DGNR5EE0/aMqKqhXs1XMvJsyZwG7ddqu3QWyufOlKHv3Tjzl2KvxjAKxsCvftsvHlypqUsaxqGSkSL017iaE7DuXBCQ8ybfE0du5W+hH3jt7haJZULuHZd57ln2//k4lzJnL8wOPZsfOORAQX7HcBABcedGGJK12/1z94neemPscZe5xBi6YtSl2OgB5tetg5VPXinfnvsG3rbRtscK9G3zEwIo6JiIkRMTkiLmrQD587F373O5bPnMHUzk1pCcxuB4e/Dd96CZafexb/c/DZHLjdgVSkte/odumRWQDo27YvAzoMYOaSmVSmSqYtnMaomaM4cvsj2anrTgzsNJD9eu/HWXudtdrdBTfH01Oe5u+T/r5W+wX7XcDw7bKaNsV+vfajRZMWnDDoBPbqvhflK8q55TO3cNSAo2pVZ239+T9/5qyHz+LaV67l8L6Hc9KuJ9GxVUcGdh7I1IVTmb10dknr2xQDOg5gzMwx3DPuHnq27cldY+5iXvm8j6a/t+A9Fi5b/90pVfe+sMsX6NN+HZcINzL/fvffTJgzodRlqJqbRt7EY28+1mCf16iPBEREU+A64NPANODViHgkpfRGgxSQElUjR9KsfBlNFy1j6InwSm84c2wzDmjVj/Pn38UpLwRXvXTVBlfz7qJ3mbooG9O/KU3p1rYbrZq34ow9z+DW12/lw/IP6dm2JxWVFZz/+Pn85NCfMLDzwM0quUXTFusMEtcPv36z1vf6zNc5b9/zuPLoK5lfMZ+ypmUbvYSxIXTZpgsT5kxg6sKpdG7ZmUlzJ3Hb0NtoV9aOypWVjeqWxq2at+LI/kdy8q4nU5WqeGveW8wtn0vHVh0B+OOoP9K3Q9/V+odINTF+9nj6tu/boP106tNNI2/iqAFH0a9Dv1KXstkuOeySBj3i16hDALAvMDmlNAUgIu4GhgINEgJWHHM0TcvLScCHzeCzE+GlvjDlhEM55IynuXfOJE6/73Rmls9c5/LbttyWQ/oewv0T7/+orVOrTnzngO8wc/FMDu17KIf1O4yqVMVZj5zFxN4TOX/f8xnQacA61zds8jDemvsWX9/36+ut+ZC+h6yz/eJnLl7vMi1owXJWv+Vvz9Y96dSqE73b9+br+2Sft7lXJNSlypWVfPvxbzNsyjAO7XsoD4x/gB8f/GNO3v3kj+ZpLAGgcmUlt79+O61btObdBe/StElTmjZpyo8P/fFq8/3g4B/QrElj/6+sUtic8TK2ZKtGNG3MyprVsCNWHWnsvzl6Ae9Vez8N2K+hPrxyxEiaALNawN17QodlcPKgz/PFwdmd8QZ1GcTwmcPXu/wfhv6B4wYex+wFszn2zmPp0rYLMxbNYJ+e+7BPr49vptMkmlDWpIwf/vOHfG3vr7GyaiUHbHfAWusb2HkgbcvaAnD9K9czuMfgdc63pkE/G7Teabt03YXffOo3HNH/CH77wm/p0bYHzZs05/D+h7OgYgEdWnagb4e+G/2MhrLj73fk7QVv07VV12zAoqZlDH9/OIf1P6zUpW2yPW7cg3nl83j45Icpa1b20b/tmuwn0HCufOlKdui0A/+943+XuhStw7l7n1vqEhqdxh4CaiQizgHOAeizrqF+N9Pv9odD3oWjz4AlZTDv+/NoU9ZmvX+VdWzWkZ998me0a9GOF6e/yAk7nQBAr069eObL2Y14/j7p7+u84c+FB13IxA8nUrGi4qNxAda0fcft2b7j9kB217htW29bo+1ozer3F2hBC8acN4aKFRXs1m03IoIVVSvo06EPn9/58yXv8b8h1x57LRf+40JGfXXUR50oG+sdAT/3ic/RtkVb9u2172qXhap0Dut7GD3a9ih1GVKdadTDBkfEAcClKaWj8/cXA6SUfr2+Zep02GBJkrZwW/Owwa8CAyOif0S0AE4CHtnIMpIkiUZ+OiCltCIizgeGAU2B21JK6x6OT5IkraZRhwCAlNJjQMNdVClJ0laisZ8OkCRJm8kQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIKKlFKpa2hQETEbmFqHq+wCzKnD9Wlt7uP65z6uX+7f+uc+Xr++KaWu65pQuBBQ1yJiREppSKnr2Jq5j+uf+7h+uX/rn/t483g6QJKkgjIESJJUUIaA2rup1AUUgPu4/rmP65f7t/65jzeDfQIkSSoojwRIklRQhoDNFBHHRMTEiJgcEReVup7GJiLeiYgxETEqIkbkbZ0i4qmIeDN/7pi3R0Rck+/r0RGxV7X1nJHP/2ZEnFGq7dkSRMRtETErIsZWa6uzfRoRe+f/ZpPzZaNht7D01rOPL42I6fnP8qiIOK7atIvz/TUxIo6u1r7O3x8R0T8ihuft90REi4bbutKLiO0i4pmIeCMixkXEN/N2f47rS0rJxyY+gKbAW8D2QAvgP8DOpa6rMT2Ad4Aua7RdDlyUv74I+E3++jjgcSCA/YHheXsnYEr+3DF/3bHU21bCfXoosBcwtj72KfBKPm/kyx5b6m3eQvbxpcD31jHvzvnvhjKgf/47o+mGfn8AfwVOyl/fCHyt1NvcwPu3B7BX/rotMCnfj/4c19PDIwGbZ19gckppSkppOXA3MLTENW0NhgJ35K/vAE6o1v6nlHkZ6BARPYCjgadSSnNTSvOAp4BjGrroLUVK6Tlg7hrNdbJP82ntUkovp+w36Z+qrasw1rOP12cocHdKaVlK6W1gMtnvjnX+/sj/Ij0SuC9fvvq/VyGklGaklF7LXy8CxgO98Oe43hgCNk8v4L1q76flbaq5BDwZESMj4py8rVtKaUb++gOgW/56ffvbf4eNq6t92it/vWa7Mufnh6NvW3Womk3fx52B+SmlFWu0F1JE9AMGA8Px57jeGAJUKgenlPYCjgXOi4hDq0/MU7qXrtQh92m9uQEYAOwJzACuKG05jV9EtAHuB76VUlpYfZo/x3XLELB5pgPbVXvfO29TDaWUpufPs4AHyQ6RzswP15E/z8pnX9/+9t9h4+pqn07PX6/ZXngppZkppZUppSrgZrKfZdj0ffwh2eHsZmu0F0pENCcLAH9JKT2QN/tzXE8MAZvnVWBg3pO3BXAS8EiJa2o0IqJ1RLRd9Ro4ChhLtg9X9eI9A3g4f/0IcHreE3h/YEF+aHAYcFREdMwPwR6Vt+ljdbJP82kLI2L//Nz16dXWVWirvpxynyX7WYZsH58UEWUR0R8YSNYpbZ2/P/K/cJ8BPp8vX/3fqxDyn61bgfEppSurTfLnuL6UumdiY32Q9UqdRNbL90elrqcxPch6Rf8nf4xbtf/Izok+DbwJ/APolLcHcF2+r8cAQ6qt68tkHa4mA2eWettKvF/vIjscXUl2rvMrdblPgSFkX3BvAdeSDzZWpMd69vH/y/fhaLIvpR7V5v9Rvr8mUq0X+vp+f+T/N17J9/29QFmpt7mB9+/BZIf6RwOj8sdx/hzX38MRAyVJKihPB0iSVFCGAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQtEERsTK/O97YiLg3IrbZxOV7RsR9+es917jL3n+Hd+GUSsZLBCVtUEQsTim1yV//BRiZVh/IZVPW9b9k13KfX4clStpMHgmQtCmeB3bI7+/+UH7TnJcjYneAiDgsP2owKiJej4i2EdEvP4rQAvg5cGI+/cSI+N+IuDYi2kfE1Ihokq+ndUS8FxHNI2JARDyR32zq+YjYqYTbL21VDAGSaiQf0/5YspHZfga8nlLaHfgh2S1ZAb4HnJdS2hM4BChftXzKbpv7U+CelNKeKaV7qk1bQDY63GF503+RDfNaCdwEXJBS2jtf//X1t5VSsTTb+CySCq5VRIzKXz9PNrb7cOBzACmlf0ZE54hoB7wAXJmfNnggpTQtG6K9Ru4BTiQbP/8k4Pr8bnIHAvdWW09ZHWyTJAwBkjauPP/L/iPr+2JPKV0WEX8nG+/9hYg4Gqio4ec8AvwqIjoBewP/BFoD89f8fEl1w9MBkjbH88CpABFxODAnpbQwIgaklMaklH5Ddre8Nc/fLwLarmuFKaXF+TK/A/6WstvzLgTejogv5J8VEbFHvWyRVECGAEmb41Jg74gYDVzGx7d5/VbeCXA02Z32Hl9juWeAnVd1DFzHeu8BvpQ/r3Iq8JWIWHXXyaF1txlSsXmJoCRJBeWRAEmSCsoQIElSQRkCJEkqKEOAJEkFZQiQJKmgDAGSJBWUIUCSpIIyBEiSVFD/H1HGTLLlK6xUAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Plot the model alongside the data**\n"
      ],
      "metadata": {
        "id": "hrliBvXxx_jF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Equation for the separation plane\n",
        "# It give a value in the negative axe as a function of a positive value\n",
        "# f(pos, neg, W) = w0 + w1 * pos + w2 * neg = 0\n",
        "# s(pos, W) = (w0 - w1 * pos) / w2\n",
        "def neg(theta, pos):\n",
        "    return (-theta[0] - pos * theta[1]) / theta[2]\n",
        "\n",
        "# Equation for the direction of the sentiments change\n",
        "# We don't care about the magnitude of the change. We are only interested \n",
        "# in the direction. So this direction is just a perpendicular function to the \n",
        "# separation plane\n",
        "# df(pos, W) = pos * w2 / w1\n",
        "def direction(theta, pos):\n",
        "    return    pos * theta[2] / theta[1]"
      ],
      "metadata": {
        "id": "GqJFPlyNyHWr"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the samples using columns 1 and 2 of the matrix\n",
        "fig, ax = plt.subplots(figsize = (8, 8))\n",
        "\n",
        "colors = ['red', 'green']\n",
        "\n",
        "# Color base on the sentiment Y\n",
        "ax.scatter(X[:,1], X[:,2], c=[colors[int(k)] for k in Y], s = 0.1)  # Plot a dot for each pair of words\n",
        "plt.xlabel(\"Positive\")\n",
        "plt.ylabel(\"Negative\")\n",
        "\n",
        "# Now lets represent the logistic regression model in this chart. \n",
        "maxpos = np.max(X[:,1])\n",
        "\n",
        "offset = 5000 # The pos value for the direction vectors origin\n",
        "\n",
        "# Plot a gray line that divides the 2 areas.\n",
        "ax.plot([0,  maxpos], [neg(theta, 0),   neg(theta, maxpos)], color = 'gray') \n",
        "\n",
        "# Plot a green line pointing to the positive direction\n",
        "ax.arrow(offset, neg(theta, offset), offset, direction(theta, offset), head_width=500, head_length=500, fc='g', ec='g')\n",
        "# Plot a red line pointing to the negative direction\n",
        "ax.arrow(offset, neg(theta, offset), -offset, -direction(theta, offset), head_width=500, head_length=500, fc='r', ec='r')\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 497
        },
        "id": "ShTljgS1yVq5",
        "outputId": "fc381d27-2e53-4ed9-e2dc-fe064cadf14d"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **References**"
      ],
      "metadata": {
        "id": "inLdMeliYnZN"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "[1- Visualizing tweets and the Logistic Regression model](https://github.com/ibrahimjelliti/Deeplearning.ai-Natural-Language-Processing-Specialization/blob/master/1%20-%20Natural%20Language%20Processing%20with%20Classification%20and%20Vector%20Spaces/Labs/Week%201/NLP_C1_W1_lecture_nb_03.ipynb)"
      ],
      "metadata": {
        "id": "qnOr5bMxYto4"
      }
    }
  ]
}