Clase02.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Clase02.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyMdToniBiMU1Iqru1c5C9De",
      "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/estebanvz/alura/blob/clase02/Clase02.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Importación de librerias"
      ],
      "metadata": {
        "id": "AQfytatk2P4u"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "0f8v_g-gZLj4"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import matplotlib.pyplot as plt"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Cargando el dataset "
      ],
      "metadata": {
        "id": "iSb_kqtQ2WqV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "dataset = keras.datasets.fashion_mnist\n",
        "(( imagenes_entrenamiento, etiquetas_entrenamiento ),(imagenes_test, etiquetas_test)) =dataset.load_data()"
      ],
      "metadata": {
        "id": "CKU2rD32Zlb1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "87bdbe4b-9f9b-40b6-b41e-1649aa954ec3"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
            "32768/29515 [=================================] - 0s 0us/step\n",
            "40960/29515 [=========================================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
            "26427392/26421880 [==============================] - 0s 0us/step\n",
            "26435584/26421880 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
            "16384/5148 [===============================================================================================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
            "4423680/4422102 [==============================] - 0s 0us/step\n",
            "4431872/4422102 [==============================] - 0s 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Exploración de datos"
      ],
      "metadata": {
        "id": "a2LbAHpj2imr"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![image.png](data:image/png;base64,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.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.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.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.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.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.com/1vPz8/bWXpRfQMG0RDT5mQDQgMFgF4nweLvJ71mvd87hs3bty6dWvevHkYhlVUVNy/fz8iIiImJkb9G6SjoyMlJWXt2rVCoVA9VZ8Y0yXoU4s15LHw0bmfffbZtm3bcBzHMGzLli0ffPCBvb09QsjCalgDeR06AA0dcCBpyBGA99nKH5klLPf69esFAgFC6MKFC1VVVRs2bGhtbc3MzGxra3N0dJw1a5ZQKKR8Zi1fvry1tfX333+XSqWenp6JiYkCgaCuru7atWsdHR1CoTA6OjogIKCjo+PKlStNTU1CoTA2NtbJyYkuwcPDw8q5m6Ke5f+oXF1dW1paEEI8Ho+y3w4ODgRBmNKQ4VTW8g9lONGDtlgbAXifre2JqOhj0RVqo0ePrqioaG9vLy4udnZ2fvrppy9cuJCTk7N06dJZs2ZduXJl9erVtra2OTk5EydOnDBhQkpKSn5+/tSpU3NyclxcXJYvX15WVlZfXx8QEHD58mU2m71hw4aCgoL09PS1a9dSEhwcHFQaOcxu58yZg2HYoDRKJpMhhPbs2VNSUjIoClhzpYP1UKyZCeg2RAnMmTPHkprjOP7NN9/Qa+TxeC+++CI9Rls4Ly/v5s2bGzduZLPZ2vLoGU8O3GIYtm7dOhsbG4SQicLpxe/evRscHMzj8fRURnc2i1pusovGYrE8PDzu3r37yy+/SKVSPp+PEGKxetbKsXsvV1fX+/fvP3jwAMdxsViMEHJzc3vw4MHjx49Hjx49efJkgiAaGxsxDDt06BC5Eaijo4OSoLvBwyD10qVLFm6Fm5vbo0ePVPrcP//8s4XVsPLqYATCyh8QqGflBMaOHRsWFkYqSf6eW1jh0tJSFouF4/jDhw9DQ0NNrz0sLGz8+PFsNrujoyMnJycwMHBIWu7q6mry2InMzMyIiIhp06ZdvHixqamJDqiysrKwsHDBggW+vr6kYUYIzZw508/Pr66urri4uLCw8Pnnn0cITZw4ccaMGVRZFTlUPARMJPDZZ59JJBL6ODk5z22iWCgOBIAAEKATEAqFo0aNomKqq6t//fVX0ha0tLQcO3Zs9uzZzs7O9JlW+sRoVlbWgwcPyJ76zz//LBQKFyxYIBKJVPKfPHmSw+EoFAonJ6dZs2ZR1SGEysrKRo0aJZfLy8rK1C13ZmZmWVmZnZ2du7t7aWnpxo0bcRzPzc0tLS1VKBTe3t5xcXG2trb79+8PCQmprKyMjo7u7Oy8efPm6tWrU1JSEEKHDx9etmxZampqWFhYXV1dZ2fntGnTmpqaysrKxowZM2/ePISQSCTKyspqaWmxtbWNjIwMDg6ma0iFLbErTCQS1dTUZGZmVldXR0REkGvnMAxraWmpr69XKBTk6ieEUEtLCzkei2FYSUmJrPciCOLy5cuPHz+Ojo6ePHmyRCKRyWSenp5VVVVdXV23b98+d+4cXYJSqaSaBwHTCWzbtk2hULz77rtN3liZwQAACClJREFUTU27d++2sNmeP38+p/fCMIwMcDic9vb2AdvFyIFjA9YCGYAAEGCKAEEQ+B8XQRCjR48WCARVVVXkaQVsNtvf35+caV27di2Px8vJyRmwavX8bDa7rq5u4sSJkydPphdvaWlpbW0N6L0aGhrI4V4qQ1VVVVFRUWRk5Ny5c2tqasipsfzea/bs2cuXL29ubs7MzMR6r8rKyvj4eB8fH7K4UCgkPxFWr17t6emJYVhtbe3ChQsdHR2zsrJ8fHyio6MrKioaGhrkcvn58+cFAsHq1avHjRt35coVkUhE6UAPWMJynzt37tdff62trY2JiZkyZYpQKAwJCSkoKLhy5cqsWbOUSuXly5e9vb3t7OwyMjJsbW29vLwuXrxYU1Mzbdq0qqqq+/fve3t737p168CBA3l5eTNmzLCxsZk9e7adnd3Ro0eLi4snTpzIYrEoCdD5pj9g08O7du169OiR5W02qfn58+cVvdfo0aOvXr1Khh0dHQdsFyMHjg1YC2QAAkCAKQL37t375o8rMzOTxWIFBARUVlaSltvPz4/H43l4eDQ3N//yyy8tLS0qxlWjGhrzOzg4jB8/XqUTUlZWxmKx/Pz8xo4dS/a/6QKbm5sRQsHBwY6Ojn5+fmRSTU2Ns7Ozj4+Pk5OTr69vbW0tGe/n5zdq1ChyXTYZozKZO2bMGKFQSFrxgIAAcqRBIpE8evTo8ePHwcHBtra2ISEhCKG6ujq6GlTYvPPcU3svqjIqENd7kbcvvfQSGVizZg0ZGDNmDJUzPDycDE+YMIGKRAg5ODgsXryYHmNnZ0dJoMdD2EQCW7duNVEC48XPnDmzfft2mUwWEBDw7bffent7V1RUrF+/nvSysmnTps2bNy9btqy1tTUpKSktLY1xBUAgEAACjBMICAiYNGkSKZZcIBYYGHjv3r3Kysrm5uYpU6Z0d3frmGnFMAzHcbK4QqFACGnLr3GyuaysDMdx6nzhsrKyiIgI3W2kL20hCIJao6pRPl0UuZgOwzAyQNp1ujR6Zo1hS/S5NVYMkUDAOAINDQ3r1q1LTk4uLi5OTEx88803EUKffPLJvHnzSkpKbt269fvvv+M4/q9//Ss2NhbMtnGQoRQQsDwBgiCUf1xdXV3k5LFQKMzJyeHz+b6+vgqFgjSQ9JlWSk+hUKhQKOrq6mpqakiPirrzUwURQk1NTR0dHVOmTHmm94qIiCBjqDyurq4IoeLi4ra2NnIYACHk6+vb2tpaW1vb2tpaVVVFDY9TpagAadRbWlp0T+a6urra2Njcv39fIpEUFBRgGKZNpnn73JTeEAACTBG4cOFCTEwMuQb1tdde27Ztm1Kp9Pb2vnjx4rx586Kioo4dO8ZUXSAHCAABixEo772o6lauXOnk5BQYGHj37l1ySpSaaa2rq5s1a9alS5cuX75MzZ0FBQWVl5efO3cuMDDQ29tbqVRqzE/JpwfIofLw8HByr5ONjU1+fn5ZWRmVx9fXNygoKC8v7+HDh56enjU1NQihsLAwiUSSkZGB4/iYMWNiY2Op/CoBajJ3/vz5Kkn0Wy6XO3/+/KysrJSUFDs7u4SEBPKLgZ6HDJvXE4t6fRAzbAhY2FfDmDFjjh07FhMT88knn/z1r391c3MjSba3txcVFbm5uX3yySfJycnUAWLHjh375ptvSA/nFmBuYRoWaBFUAQSAAEVAIpEUFxd79F5nz55ls9lLliyhUi0fgD635ZlDjSYR8PLySkpK+uWXX1SkqBwgppIKt0AACAABownY2NhIJJL09HSlUunu7j5z5kyjRTFSEOa5GcEIQixHYO7cuVlZWaQTtxs3brz99tsIIfUDxKgDxyynGdQEBIDAMCWAYVhcXNyGDRtefvnlJUuWODk5DW5Doc89uPyhdoMJeHp6Hjhw4Nlnn5VIJPb29nv37kUIvfvuu2/0XiwW64UXXoiKihozZkxdXZ2Pjw85I2VwNVAACAABIGCtBGCe21qfjNXrBTO79EcENOg0IAwEgIBZCcBouVnxgnAgAASAABAAAgwTAMvNMFAQBwSAABAAAkDArATAcpsVLwgHAkAACAABIMAwAbDcDAMFcUAACAABIAAEzEoALLdZ8YJwIAAEgAAQAAIMEwDLzTBQEAcEgAAQAAJAwKwEwHKbFS8IBwJAAAgAASDAMAGw3AwDBXFAAAgAASAABMxKACy3WfGCcCAABIAAEAACDBMAy80wUBAHBIAAEAACQMCsBMBymxUvCAcCQAAIAAEgwDABsNwMAwVxQAAIAAEgAATMSgAst1nxgnAgAASAABAAAgwTAMvNMFAQBwSAABAAAkDArATAcpsVLwgHAkAACAABIMAwAbDcDAMFcUAACAABIAAEzEoALLdZ8YJwIAAEgAAQAAIMEwDLzTBQEAcEgAAQAAJAwKwEwHKbFS8IBwJAAAgAASDAMAGw3AwDBXFAAAgAASAABMxKACy3WfGCcCAABIAAEAACDBMAy80wUBAHBIAAEAACQMCsBMBymxUvCAcCQAAIAAEgwDABsNwMAwVxQAAIAAEgAATMSgAst1nxgnAgAASAABAAAgwTAMvNMFAQBwSAABAAAkDArATAcpsVLwgHAkAACAABIMAwAbDcDAMFcUAACAABIAAEzEoALLdZ8YJwIAAEgAAQAAIMEwDLzTBQEAcEgAAQAAJAwKwEwHKbFS8IBwJAAAgAASDAMAGw3AwDBXFAAAgAASAABMxKgGNW6SB8eBPAMGx4NxBaBwSAABCwQgIYQRBWqBaoBASAABAAAkAACGgk8P8D7wa0eVipIwgAAAAASUVORK5CYII=)"
      ],
      "metadata": {
        "id": "Gm7x8yJv2tVT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def mostrar_estructura(imagenes,etiquetas,titulo):\n",
        "  print(\"--> \",titulo)\n",
        "  print(\n",
        "      \"Cantidad de imágenes: {} \\nCantidad de etiquetas: {} \\nEstructura de las imágenes: {}\\nEtiquetas: {}\"\n",
        "      .format(imagenes.shape[0],etiquetas.shape[0],imagenes.shape[1:],list(set(etiquetas)))\n",
        "  )"
      ],
      "metadata": {
        "id": "p53O6WBmueax"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "mostrar_estructura(imagenes_entrenamiento,etiquetas_entrenamiento,\"Dataset de entrenamiento\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V_xVUJ-AvZLF",
        "outputId": "c2898cac-f7da-4c9d-9efa-01978381e257"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "-->  Dataset de entrenamiento\n",
            "Cantidad de imágenes: 60000 \n",
            "Cantidad de etiquetas: 60000 \n",
            "Estructura de las imágenes: (28, 28)\n",
            "Etiquetas: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mostrar_estructura(imagenes_test,etiquetas_test,\"Dataset de test\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SkRx_TFEvgZZ",
        "outputId": "0b30598b-c673-4487-b16b-eb1c16735960"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "-->  Dataset de test\n",
            "Cantidad de imágenes: 10000 \n",
            "Cantidad de etiquetas: 10000 \n",
            "Estructura de las imágenes: (28, 28)\n",
            "Etiquetas: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Descripción de las etiquetas\n",
        "[Github fashion mnist](https://github.com/zalandoresearch/fashion-mnist)\n",
        "\n",
        "Etiquetas (labels):\n",
        "* 0 T-shirt/top -> camisa\n",
        "* 1 Trouser -> pantalón\n",
        "* 2 Pullover -> suéter\n",
        "* 3 Dress -> vestido\n",
        "* 4 Coat -> saco\n",
        "* 5 Sandal -> sandalia\n",
        "* 6 Shirt -> camisa\n",
        "* 7 Sneaker -> zapatilla\n",
        "* 8 Bag -> bolsa\n",
        "* 9 Ankle boot -> bota"
      ],
      "metadata": {
        "id": "EWj7F_kaICgl"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Visualización de datos"
      ],
      "metadata": {
        "id": "oB9YHOQg2nL2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "etiquetas = [\"camisa\",\"pantalón\",\"suéter\",\"vestido\",\"saco\",\"sandalia\",\"camisa\",\"zapatilla\",\"bolsa\",\"bota\",]"
      ],
      "metadata": {
        "id": "10o7K-2VGt6u"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for index in range(10):\n",
        "  plt.subplot( 2, 5, index + 1 )\n",
        "  plt.axis(\"off\")\n",
        "  plt.title( etiquetas[etiquetas_entrenamiento[index+30]])\n",
        "  plt.imshow(imagenes_entrenamiento[index+30],cmap=\"gray\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 223
        },
        "id": "mkXaMV8ZIfKg",
        "outputId": "abb05428-877e-4785-d49b-cb006bac62a1"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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.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.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.com/M/T7lc5sqVK2xsbHD58mWWl5d57rnnWFhYMIvVQeNIzAs6uIPBoFltO50OzWaTWq1mpLHx8XGCweCh2VbudegEa7fbA/ZKXbXb7baZdN1ul0KhwNbWllHLfT4fiUTCENBhQiWvaDRqpNR8Ps/4+DjZbNYc12q1aDQaZsCrXVqJIxAIUCqVAMyE0EVHf6e2bp2EKk3H43FEhFarRbPZJBwO0+12qdfrLC0tmT48blBTi9/vN/1Sr9fNvFDNwXY26bHaJ+pI0zESCATMOY8zfD7fgHSrjuLz58/zute9jlwux5kzZ2i1WhSLxQGtxW1+abVaA2bNhx56CL/fz9jYGCsrK/j9fqLRKGtra6ytrRmJV+fPQWGkpGsPCsdxiMfj3H//github.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.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\n",
            "text/plain": [
              "<Figure size 432x288 with 10 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Redes Neuronales (Machine Learning Clásico)"
      ],
      "metadata": {
        "id": "-sKnzOYg8ZSe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "imagenes_entrenamiento.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2J_pZPun8dm9",
        "outputId": "357be488-fe33-4dbd-b5db-da69959fe8bf"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28)"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "imagenes_entrenamiento_flat = imagenes_entrenamiento.reshape( (60000, 28 * 28) )\n",
        "imagenes_entrenamiento_flat.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MPpVJ15T8mCW",
        "outputId": "828eb15a-e653-4b78-fa8c-ef8d0ea4d33c"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 784)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "red_neuronal = keras.Sequential([\n",
        "  keras.layers.Dense(10, input_shape=(784,) )\n",
        "])\n",
        "red_neuronal.compile( loss = \"sparse_categorical_crossentropy\" )"
      ],
      "metadata": {
        "id": "8G7hUU4b9C9T"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "red_neuronal.fit( imagenes_entrenamiento_flat, etiquetas_entrenamiento )"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k3K3xYyN-FNy",
        "outputId": "4ade79c6-9e5c-40ab-e602-3c3b5a498501"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1875/1875 [==============================] - 3s 1ms/step - loss: 2.8512\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fa172bea3d0>"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "p_Mw0eKy-NrE"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}