filename_ANN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "filename_ANN.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/obabilonia/test1/blob/master/filename_ANN.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "aw7i7RN0RWDH",
        "colab_type": "code",
        "outputId": "090068b6-3827-412f-864e-c95bf4779afc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2247
        }
      },
      "cell_type": "code",
      "source": [
        "print(\"preparando ambiente\")\n",
        "!pip install pyvirtualdisplay\n",
        "!pip install unidecode\n",
        "!pip install lxml\n",
        "!wget https://github.com/denadai2/google_street_view_deep_neural/archive/master.zip\n",
        "!unzip master.zip\n",
        "!mv google_street_view_deep_neural-master/* .\n",
        "!wget https://ndownloader.figshare.com/files/11086517\n",
        "!mv 11086517 generated_files/pytorch_state.npy\n",
        "!pip install torchvision\n",
        "!pip install Pillow==4.0.0\n",
        "!pip install image\n",
        "!mkdir images\n",
        "!wget http://cs.famaf.unc.edu.ar/~ccardellino/SBWCE/SBW-vectors-300-min5.txt.bz2\n",
        "!bzip2 -d SBW-vectors-300-min5.txt.bz2"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "preparando ambiente\n",
            "Collecting pyvirtualdisplay\n",
            "  Downloading https://files.pythonhosted.org/packages/39/37/f285403a09cc261c56b6574baace1bdcf4b8c7428c8a7239cbba137bc0eb/PyVirtualDisplay-0.2.1.tar.gz\n",
            "Collecting EasyProcess (from pyvirtualdisplay)\n",
            "  Downloading https://files.pythonhosted.org/packages/0d/f1/d2de7591e7dfc164d286fa16f051e6c0cf3141825586c3b04ae7cda7ac0f/EasyProcess-0.2.3.tar.gz\n",
            "Building wheels for collected packages: pyvirtualdisplay, EasyProcess\n",
            "  Running setup.py bdist_wheel for pyvirtualdisplay ... \u001b[?25l-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /root/.cache/pip/wheels/d1/8c/16/1c64227974ae29c687e4cc30fd691d5c0fd40f54446dde99da\n",
            "  Running setup.py bdist_wheel for EasyProcess ... \u001b[?25l-\b \bdone\n",
            "\u001b[?25h  Stored in directory: /root/.cache/pip/wheels/b4/c6/e3/c163b04029d8fccfd54b809802640c1af587a01be8d7a04e1a\n",
            "Successfully built pyvirtualdisplay EasyProcess\n",
            "Installing collected packages: EasyProcess, pyvirtualdisplay\n",
            "Successfully installed EasyProcess-0.2.3 pyvirtualdisplay-0.2.1\n",
            "Collecting unidecode\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/31/39/53096f9217b057cb049fe872b7fc7ce799a1a89b76cf917d9639e7a558b5/Unidecode-1.0.23-py2.py3-none-any.whl (237kB)\n",
            "\u001b[K    100% |████████████████████████████████| 245kB 7.0MB/s \n",
            "\u001b[?25hInstalling collected packages: unidecode\n",
            "Successfully installed unidecode-1.0.23\n",
            "Collecting lxml\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/e5/14/f4343239f955442da9da1919a99f7311bc5627522741bada61b2349c8def/lxml-4.2.5-cp27-cp27mu-manylinux1_x86_64.whl (5.8MB)\n",
            "\u001b[K    100% |████████████████████████████████| 5.8MB 4.1MB/s \n",
            "\u001b[?25hInstalling collected packages: lxml\n",
            "Successfully installed lxml-4.2.5\n",
            "--2018-11-27 18:37:12--  https://github.com/denadai2/google_street_view_deep_neural/archive/master.zip\n",
            "Resolving github.com (github.com)... 192.30.253.112, 192.30.253.113\n",
            "Connecting to github.com (github.com)|192.30.253.112|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://codeload.github.com/denadai2/google_street_view_deep_neural/zip/master [following]\n",
            "--2018-11-27 18:37:12--  https://codeload.github.com/denadai2/google_street_view_deep_neural/zip/master\n",
            "Resolving codeload.github.com (codeload.github.com)... 192.30.253.121, 192.30.253.120\n",
            "Connecting to codeload.github.com (codeload.github.com)|192.30.253.121|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: unspecified [application/zip]\n",
            "Saving to: ‘master.zip’\n",
            "\n",
            "master.zip              [  <=>               ]   4.31M  8.49MB/s    in 0.5s    \n",
            "\n",
            "2018-11-27 18:37:12 (8.49 MB/s) - ‘master.zip’ saved [4517871]\n",
            "\n",
            "Archive:  master.zip\n",
            "384cdb3531c56bdb6754f67075cbefb27ffeee29\n",
            "   creating: google_street_view_deep_neural-master/\n",
            "  inflating: google_street_view_deep_neural-master/.gitignore  \n",
            "  inflating: google_street_view_deep_neural-master/ACMMM.ipynb  \n",
            "  inflating: google_street_view_deep_neural-master/LICENSE  \n",
            "  inflating: google_street_view_deep_neural-master/README.md  \n",
            "   creating: google_street_view_deep_neural-master/caffe/\n",
            "  inflating: google_street_view_deep_neural-master/caffe/deploy_gpu.prototxt  \n",
            "  inflating: google_street_view_deep_neural-master/caffe/places205CNN_finetune.prototxt  \n",
            "  inflating: google_street_view_deep_neural-master/caffe/solver.prototxt  \n",
            "   creating: google_street_view_deep_neural-master/data/\n",
            "  inflating: google_street_view_deep_neural-master/data/list_files.csv  \n",
            "   creating: google_street_view_deep_neural-master/figures/\n",
            "  inflating: google_street_view_deep_neural-master/figures/fit.pdf  \n",
            "  inflating: google_street_view_deep_neural-master/figures/prediction.pdf  \n",
            "  inflating: google_street_view_deep_neural-master/figures/prediction_wnames.pdf  \n",
            "  inflating: google_street_view_deep_neural-master/figures/safety_prediction.pdf  \n",
            "  inflating: google_street_view_deep_neural-master/figures/women.pdf  \n",
            "   creating: google_street_view_deep_neural-master/generated_files/\n",
            "  inflating: google_street_view_deep_neural-master/generated_files/places205CNN_mean_filtered.npy  \n",
            "  inflating: google_street_view_deep_neural-master/security_perception_prediction.png  \n",
            "  inflating: google_street_view_deep_neural-master/streetview_image.jpg  \n",
            "--2018-11-27 18:37:24--  https://ndownloader.figshare.com/files/11086517\n",
            "Resolving ndownloader.figshare.com (ndownloader.figshare.com)... 52.213.114.32, 18.203.143.159, 54.194.37.96, ...\n",
            "Connecting to ndownloader.figshare.com (ndownloader.figshare.com)|52.213.114.32|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://s3-eu-west-1.amazonaws.com/pfigshare-u-files/11086517/pytorch_state.npy [following]\n",
            "--2018-11-27 18:37:24--  https://s3-eu-west-1.amazonaws.com/pfigshare-u-files/11086517/pytorch_state.npy\n",
            "Resolving s3-eu-west-1.amazonaws.com (s3-eu-west-1.amazonaws.com)... 52.218.53.26\n",
            "Connecting to s3-eu-west-1.amazonaws.com (s3-eu-west-1.amazonaws.com)|52.218.53.26|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 227491170 (217M) [binary/octet-stream]\n",
            "Saving to: ‘11086517’\n",
            "\n",
            "11086517            100%[===================>] 216.95M  3.38MB/s    in 90s     \n",
            "\n",
            "2018-11-27 18:38:55 (2.41 MB/s) - ‘11086517’ saved [227491170/227491170]\n",
            "\n",
            "Collecting torchvision\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/ca/0d/f00b2885711e08bd71242ebe7b96561e6f6d01fdb4b9dcf4d37e2e13c5e1/torchvision-0.2.1-py2.py3-none-any.whl (54kB)\n",
            "\u001b[K    100% |████████████████████████████████| 61kB 2.3MB/s \n",
            "\u001b[?25hRequirement already satisfied: six in /usr/local/lib/python2.7/dist-packages (from torchvision) (1.11.0)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python2.7/dist-packages (from torchvision) (1.14.6)\n",
            "Collecting torch (from torchvision)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/06/a7/6a173738dd6be014ebf9ba6f0b441d91b113b1506a98e10da4ff60994b54/torch-0.4.1-cp27-cp27mu-manylinux1_x86_64.whl (519.5MB)\n",
            "\u001b[K    100% |████████████████████████████████| 519.5MB 23kB/s \n",
            "tcmalloc: large alloc 1073750016 bytes == 0x5579d4c2c000 @  0x7f75b5f372a4 0x55797ac63f18 0x55797ad57a85 0x55797ac774ca 0x55797ac7c232 0x55797ac74d0a 0x55797ac7c5fe 0x55797ac74d0a 0x55797ac7c5fe 0x55797ac74d0a 0x55797ac7c5fe 0x55797ac74d0a 0x55797ac7cc38 0x55797ac74d0a 0x55797ac7c5fe 0x55797ac74d0a 0x55797ac7c5fe 0x55797ac7c232 0x55797ac7c232 0x55797ac74d0a 0x55797ac7cc38 0x55797ac7c232 0x55797ac74d0a 0x55797ac7cc38 0x55797ac74d0a 0x55797ac7cc38 0x55797ac74d0a 0x55797ac7c5fe 0x55797ac74d0a 0x55797ac74629 0x55797aca561f\n",
            "\u001b[?25hCollecting pillow>=4.1.1 (from torchvision)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/9a/f6/3b3c82c5c75cae471e02fb584136168d732e17ae9db2d21c5dc82f9790f8/Pillow-5.3.0-cp27-cp27mu-manylinux1_x86_64.whl (2.0MB)\n",
            "\u001b[K    100% |████████████████████████████████| 2.0MB 3.1MB/s \n",
            "\u001b[?25hInstalling collected packages: torch, pillow, torchvision\n",
            "  Found existing installation: Pillow 4.0.0\n",
            "    Uninstalling Pillow-4.0.0:\n",
            "      Successfully uninstalled Pillow-4.0.0\n",
            "Successfully installed pillow-5.3.0 torch-0.4.1 torchvision-0.2.1\n",
            "Collecting Pillow==4.0.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/89/99/0e3522a9764fe371bf9f7729404b1ef7d9c4fc49cbe5f1761c6e07812345/Pillow-4.0.0-cp27-cp27mu-manylinux1_x86_64.whl (5.6MB)\n",
            "\u001b[K    100% |████████████████████████████████| 5.6MB 4.6MB/s \n",
            "\u001b[?25hRequirement already satisfied: olefile in /usr/local/lib/python2.7/dist-packages (from Pillow==4.0.0) (0.46)\n",
            "\u001b[31mtorchvision 0.2.1 has requirement pillow>=4.1.1, but you'll have pillow 4.0.0 which is incompatible.\u001b[0m\n",
            "Installing collected packages: Pillow\n",
            "  Found existing installation: Pillow 5.3.0\n",
            "    Uninstalling Pillow-5.3.0:\n",
            "      Successfully uninstalled Pillow-5.3.0\n",
            "Successfully installed Pillow-4.0.0\n",
            "Collecting image\n",
            "  Downloading https://files.pythonhosted.org/packages/0c/ec/51969468a8b87f631cc0e60a6bf1e5f6eec8ef3fd2ee45dc760d5a93b82a/image-1.5.27-py2.py3-none-any.whl\n",
            "Collecting django (from image)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/44/e7/872bbf76aa16b7a061698d75325dac023285db33db4bda8ba8fe5d3bb356/Django-1.11.16-py2.py3-none-any.whl (7.0MB)\n",
            "\u001b[K    100% |████████████████████████████████| 7.0MB 3.5MB/s \n",
            "\u001b[?25hRequirement already satisfied: pillow in /usr/local/lib/python2.7/dist-packages (from image) (4.0.0)\n",
            "Requirement already satisfied: pytz in /usr/local/lib/python2.7/dist-packages (from django->image) (2018.7)\n",
            "Requirement already satisfied: olefile in /usr/local/lib/python2.7/dist-packages (from pillow->image) (0.46)\n",
            "Installing collected packages: django, image\n",
            "Successfully installed django-1.11.16 image-1.5.27\n",
            "--2018-11-27 18:41:02--  http://cs.famaf.unc.edu.ar/~ccardellino/SBWCE/SBW-vectors-300-min5.txt.bz2\n",
            "Resolving cs.famaf.unc.edu.ar (cs.famaf.unc.edu.ar)... 200.16.17.55\n",
            "Connecting to cs.famaf.unc.edu.ar (cs.famaf.unc.edu.ar)|200.16.17.55|:80... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://cs.famaf.unc.edu.ar/~ccardellino/SBWCE/SBW-vectors-300-min5.txt.bz2 [following]\n",
            "--2018-11-27 18:41:02--  https://cs.famaf.unc.edu.ar/~ccardellino/SBWCE/SBW-vectors-300-min5.txt.bz2\n",
            "Connecting to cs.famaf.unc.edu.ar (cs.famaf.unc.edu.ar)|200.16.17.55|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 818175453 (780M) [application/x-bzip2]\n",
            "Saving to: ‘SBW-vectors-300-min5.txt.bz2’\n",
            "\n",
            "SBW-vectors-300-min 100%[===================>] 780.27M  17.7MB/s    in 57s     \n",
            "\n",
            "2018-11-27 18:42:00 (13.7 MB/s) - ‘SBW-vectors-300-min5.txt.bz2’ saved [818175453/818175453]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "i8iPJsAFRWDN",
        "colab_type": "code",
        "outputId": "3ae06dd3-d328-4bcd-d926-eaf1adc35603",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "print(\"comandos auxiliares\")\n",
        "#%save -f test.ipynb\n",
        "#%%writefile ./filename2.ipynb\n",
        "#hola\n",
        "#%lsmagic\n",
        "#page = requests.get(\"https://www.dataquest.io/blog/web-scraping-tutorial-python/\")\n",
        "#page.content\n",
        "#!zip -r images.zip images"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "comandos auxiliares\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "TWHW3M3kRWDR",
        "colab_type": "code",
        "outputId": "36cf543d-4c0d-4376-f552-68a7c6d922a7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "!rm Oscar.py"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "rm: cannot remove 'Oscar.py': No such file or directory\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "v0CNT51cRWDU",
        "colab_type": "code",
        "outputId": "688d7f5f-2c82-4de4-a964-14da81f03bb5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "cell_type": "code",
      "source": [
        "!wget -O Oscar3.py https://raw.githubusercontent.com/obabilonia/test1/master/Oscar.py"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2018-11-27 18:52:26--  https://raw.githubusercontent.com/obabilonia/test1/master/Oscar.py\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 5175 (5.1K) [text/plain]\n",
            "Saving to: ‘Oscar3.py’\n",
            "\n",
            "\rOscar3.py             0%[                    ]       0  --.-KB/s               \rOscar3.py           100%[===================>]   5.05K  --.-KB/s    in 0s      \n",
            "\n",
            "2018-11-27 18:52:26 (55.7 MB/s) - ‘Oscar3.py’ saved [5175/5175]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "7vevW_rNRWDX",
        "colab_type": "code",
        "outputId": "07a52cac-1735-443e-aa11-51a52cba26db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        }
      },
      "cell_type": "code",
      "source": [
        "!head Oscar3.py"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "import re\n",
            "import nltk\n",
            "import pandas as pd\n",
            "import numpy as np\n",
            "from unidecode import unidecode\n",
            "\n",
            "def O_print_full(x):\n",
            "    pd.set_option('display.max_rows', len(x))\n",
            "    pd.set_option('display.max_columns', None)\n",
            "    pd.set_option('display.width', 2000)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "_lXxLPO_RWDa",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import Oscar3"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Dz63ssF7RWDd",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import lxml\n",
        "import requests\n",
        "import pandas as pd\n",
        "import os\n",
        "#from selenium import webdriver\n",
        "#from selenium.webdriver.support.ui import Select\n",
        "import time\n",
        "from PIL import Image\n",
        "import subprocess\n",
        "#import commands\n",
        "import pickle\n",
        "#from selenium.webdriver.chrome.options import Options\n",
        "from pyvirtualdisplay import Display\n",
        "from IPython.display import clear_output\n",
        "import matplotlib.pyplot as plt\n",
        "from unidecode import unidecode\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "FEmm2D-ERWDs",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "from PIL import Image\n",
        "%matplotlib inline\n",
        "from matplotlib.pyplot import imshow\n",
        "\n",
        "from torch.utils.data import Dataset, DataLoader\n",
        "from torch.autograd import Variable\n",
        "from torchvision import transforms, utils\n",
        "\n",
        "from PIL import Image\n",
        "import requests\n",
        "from io import BytesIO\n",
        "from matplotlib.pyplot import imshow"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "w0ahufOVRWDh",
        "colab_type": "code",
        "outputId": "84d79b9f-c09e-4389-c547-2ac5a4b346ae",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "print(\"Get Descripcion\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Get Descripcion\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "utG4x7k2RWDk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "try: \n",
        "    from BeautifulSoup import BeautifulSoup\n",
        "except ImportError:\n",
        "    from bs4 import BeautifulSoup\n",
        "\n",
        "def get_description(fuente):\n",
        "  soup = BeautifulSoup(fuente)\n",
        "  dic = {}\n",
        "  try:\n",
        "    dic[\"ubicacion\"] = unidecode(soup.find('div', attrs={'class':'section-map-title'}).get_text()).replace(\"\\n\",\" \").replace(\"\\t\",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \")\n",
        "  except:\n",
        "    dic[\"ubicacion\"] = 0\n",
        "  try:\n",
        "    dic[\"main_descripcion\"] = unidecode(soup.find('div', attrs={'class':'description-content-main-group attribute-content'}).get_text()).replace(\"\\n\",\" \").replace(\"\\t\",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \")\n",
        "  except:\n",
        "    dic[\"main_descripcion\"] = 0\n",
        "  try:\n",
        "    dic[\"sec_descripcion\"] = unidecode(soup.find('div', attrs={'class':'description-content-secondary-group attribute-content'}).get_text()).replace(\"\\n\",\" \").replace(\"\\t\",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \")\n",
        "  except:\n",
        "    dic[\"sec_descripcion\"] = 0\n",
        "  try:\n",
        "    dic[\"vendedor\"] = unidecode(soup.find('span', attrs={'class':'profile-info-data profile-info-name-data'}).get_text()).replace(\"\\n\",\" \").replace(\"\\t\",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \")\n",
        "  except:\n",
        "    dic[\"vendedor\"] = 0\n",
        "  try:\n",
        "    dic[\"title\"] = unidecode(soup.find('p', attrs={'class':'description-content-title'}).get_text()).replace(\"\\n\",\" \").replace(\"\\t\",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \")\n",
        "  except:\n",
        "    dic[\"title\"] = 0\n",
        "  try:\n",
        "    dic[\"texto\"] = unidecode(soup.find('pre', attrs={'class':'preformated-text'}).get_text()).replace(\"\\n\",\" \").replace(\"\\t\",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \").replace(\"  \",\" \")\n",
        "  except:\n",
        "    dic[\"texto\"]=0\n",
        "  return dic\n",
        "#get_description(fuente)\n",
        "#soup = BeautifulSoup(fuente)\n",
        "#soup.find('div', attrs={'class':'description-content-secondary-group attribute-content'})"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "VGvyM1HBRWDp",
        "colab_type": "code",
        "outputId": "389688c5-721e-40b0-f718-610df299490a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "print(\"Safety Feature\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Safety Feature\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "jHksVpw9RWDv",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class KitModel(nn.Module):\n",
        "\n",
        "    def __init__(self):\n",
        "        super(KitModel, self).__init__()\n",
        "\n",
        "        self.conv1 = nn.Conv2d(3, 96, (11, 11), stride=4, padding=0)\n",
        "        self.conv2 = nn.Conv2d(in_channels=96, out_channels=256, kernel_size=5, stride=1, groups=2, padding=2)\n",
        "        self.conv3 = nn.Conv2d(in_channels=256, out_channels=384, kernel_size=3, stride=1, groups=1, padding=1)\n",
        "        self.conv4 = nn.Conv2d(in_channels=384, out_channels=384, kernel_size=3, stride=1, groups=2, padding=1)\n",
        "        self.conv5 = nn.Conv2d(in_channels=384, out_channels=256, kernel_size=3, stride=1, groups=2, padding=1)\n",
        "        self.fc6_1 = nn.Linear(in_features = 9216, out_features = 4096)\n",
        "        self.fc7_1 = nn.Linear(in_features = 4096, out_features = 4096)\n",
        "        self.ip_1 = nn.Linear(in_features = 4096, out_features = 1)\n",
        "        \n",
        "        self.relu = nn.ReLU()\n",
        "        self.drop = nn.Dropout()\n",
        "        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2)\n",
        "\n",
        "    def forward(self, x):\n",
        "        conv1           = self.conv1(x)\n",
        "        relu1           = self.relu(conv1)\n",
        "        pool1           = self.maxpool(relu1)\n",
        "        norm1           = self.LRN(size = 5, alpha=0.0001, beta=0.75)(pool1)\n",
        "        \n",
        "        conv2           = self.conv2(norm1)\n",
        "        relu2           = self.relu(conv2)\n",
        "        pool2           = self.maxpool(relu2)\n",
        "        norm2           = self.LRN(size = 5, alpha=0.0001, beta=0.75)(pool2)\n",
        "        \n",
        "        conv3           = self.conv3(norm2)\n",
        "        relu3           = self.relu(conv3)\n",
        "        conv4           = self.conv4(relu3)\n",
        "        relu4           = self.relu(conv4)\n",
        "        conv5           = self.conv5(relu4)\n",
        "        relu5           = self.relu(conv5)\n",
        "        pool5           = self.maxpool(relu5)\n",
        "        \n",
        "        fc6_0           = pool5.view(pool5.size(0), -1)\n",
        "        \n",
        "        fc6_1           = self.fc6_1(fc6_0)\n",
        "        relu6           = self.relu(fc6_1)\n",
        "        drop6           = self.drop(relu6)\n",
        "        fc7_1           = self.fc7_1(drop6)\n",
        "        relu7           = self.relu(fc7_1)\n",
        "        ip_0            = self.drop(relu7)\n",
        "        ip_1            = self.ip_1(ip_0)\n",
        "        \n",
        "        return ip_1\n",
        "    \n",
        "    class LRN(nn.Module):\n",
        "        def __init__(self, size=1, alpha=1.0, beta=0.75, ACROSS_CHANNELS=True):\n",
        "            super(KitModel.LRN, self).__init__()\n",
        "            self.ACROSS_CHANNELS = ACROSS_CHANNELS\n",
        "            if self.ACROSS_CHANNELS:\n",
        "                self.average=nn.AvgPool3d(kernel_size=(size, 1, 1),\n",
        "                        stride=1,\n",
        "                        padding=(int((size-1.0)/2), 0, 0))\n",
        "            else:\n",
        "                self.average=nn.AvgPool2d(kernel_size=size,\n",
        "                        stride=1,\n",
        "                        padding=int((size-1.0)/2))\n",
        "            self.alpha = alpha\n",
        "            self.beta = beta\n",
        "\n",
        "        def forward(self, x):\n",
        "            if self.ACROSS_CHANNELS:\n",
        "                div = x.pow(2).unsqueeze(1)\n",
        "                div = self.average(div).squeeze(1)\n",
        "                div = div.mul(self.alpha).add(1.0).pow(self.beta)\n",
        "            else:\n",
        "                div = x.pow(2)\n",
        "                div = self.average(div)\n",
        "                div = div.mul(self.alpha).add(1.0).pow(self.beta)\n",
        "            x = x.div(div)\n",
        "            return x"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "SGb3J8xhRWDy",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class PandasDataset(Dataset):\n",
        "    def __init__(self, list_images, list_targets, transform=None):\n",
        "        self.list_images = list_images\n",
        "        self.list_targets = list_targets\n",
        "        # add transforms as well\n",
        "        self.transform = transform\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        \n",
        "        image = Image.open(self.list_images[idx]).convert('RGB')\n",
        "        image = image.resize((227,227), Image.BILINEAR) \n",
        "        image = np.array(image, dtype='f4')\n",
        "        # Convert RGB to BGR \n",
        "        image = image[:, :, ::-1]\n",
        "        \n",
        "        image = image.astype('float32')\n",
        "        \n",
        "        # add transforms\n",
        "        if self.transform:\n",
        "            image = self.transform(image)\n",
        "            \n",
        "        return image, self.list_targets[idx]\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.list_images)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "C3KstLmRRWD1",
        "colab_type": "code",
        "outputId": "c55ea483-28f4-4af7-9e9b-abf52b9d830c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        }
      },
      "cell_type": "code",
      "source": [
        "model = KitModel()\n",
        "\n",
        "model.load_state_dict(torch.load('generated_files/pytorch_state.npy'))\n",
        "model.train(False)\n",
        "model.eval()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "KitModel(\n",
              "  (conv1): Conv2d(3, 96, kernel_size=(11, 11), stride=(4, 4))\n",
              "  (conv2): Conv2d(96, 256, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2), groups=2)\n",
              "  (conv3): Conv2d(256, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
              "  (conv4): Conv2d(384, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=2)\n",
              "  (conv5): Conv2d(384, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), groups=2)\n",
              "  (fc6_1): Linear(in_features=9216, out_features=4096, bias=True)\n",
              "  (fc7_1): Linear(in_features=4096, out_features=4096, bias=True)\n",
              "  (ip_1): Linear(in_features=4096, out_features=1, bias=True)\n",
              "  (relu): ReLU()\n",
              "  (drop): Dropout(p=0.5)\n",
              "  (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "metadata": {
        "id": "p-9bWUU0RWD3",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "batch_size = 30\n",
        "\n",
        "file_list = [\n",
        "    'streetview_image.jpg',\n",
        "]\n",
        "# I'm interested only in testing the predictions, so label=0\n",
        "labels = [\n",
        "    0\n",
        "]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "1AsChDarRWD6",
        "colab_type": "code",
        "outputId": "f909aa7e-f4c1-4eef-9489-6e35d3f9d835",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        }
      },
      "cell_type": "code",
      "source": [
        "image = Image.open(file_list[0]).convert('RGB')\n",
        "imshow(np.array(image))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f162f877ad0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVAAAAD8CAYAAAAhQfz4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCPH/ADb+lJAaMG/4AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f162f8fef90>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "mWE-MCxrRWD9",
        "colab_type": "code",
        "outputId": "7c3df762-d361-4dc3-f3ef-b5cba2655b3c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "model = model.double()\n",
        "\n",
        "means = np.load('generated_files/places205CNN_mean_filtered.npy')\n",
        "\n",
        "transformations = transforms.Compose([lambda x: x - means, # Subtracts image means\n",
        "                                      transforms.ToTensor(), \n",
        "                                      lambda x: x*255] # Restore the input range to [0, 255]\n",
        "                                    )\n",
        "\n",
        "\n",
        "dataset = PandasDataset(file_list, labels, transformations)\n",
        "\n",
        "\n",
        "load = DataLoader(dataset, batch_size=batch_size, shuffle=False, num_workers=10)\n",
        "\n",
        "preds = np.zeros((len(file_list), 1))\n",
        "\n",
        "for i, data in enumerate(load):\n",
        "    inputs, labels = data\n",
        "    \n",
        "    n = len(inputs)\n",
        "    ifrom = i*batch_size\n",
        "    ito = i*batch_size+n\n",
        "    \n",
        "    inputs, labels = Variable(inputs), Variable(labels)\n",
        "    \n",
        "    outputs = model(inputs)\n",
        "    preds[ifrom:ito] = outputs.data.numpy()\n",
        "    \n",
        "print(\"Predicted:\", preds)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "('Predicted:', array([[4.96546034]]))\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "NvCgJ7SXRWEB",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def get_fuente(url):\n",
        "  try:\n",
        "    return requests.get(url).content\n",
        "  except:\n",
        "    get_fuente(url)\n",
        "    \n",
        "def get_Image(url):\n",
        "  try:\n",
        "    image = Image.open(BytesIO(get_fuente(url)))\n",
        "    image.save(\"test.png\")\n",
        "    return image\n",
        "  except:\n",
        "    get_Image(url)\n",
        "\n",
        "def persection(lng = 6.1870028, lat = -75.6496416, heading= 0):\n",
        "  url = \"http://maps.googleapis.com/maps/api/streetview?size=400x300&location=\"+str(lng)+\",%20\"+str(lat)+\"&heading=\"+str(heading)+\"&sensor=false&key=AIzaSyA4XFoQFphSaUrcGmvT06NUcexvxbpBc2Y\"\n",
        "  #response = requests.get(url)\n",
        "  image = get_Image(url)\n",
        "  #image.save(\"test.png\")\n",
        "  \n",
        "  file_list = ['test.png',]\n",
        "# I'm interested only in testing the predictions, so label=0\n",
        "  labels = [0]\n",
        "\n",
        "\n",
        "  dataset = PandasDataset(file_list, labels, transformations)\n",
        "\n",
        "\n",
        "  load = DataLoader(dataset, batch_size=batch_size, shuffle=False, num_workers=10)\n",
        "\n",
        "  preds = np.zeros((len(file_list), 1))\n",
        "\n",
        "  for i, data in enumerate(load):\n",
        "      inputs, labels = data\n",
        "\n",
        "      n = len(inputs)\n",
        "      ifrom = i*batch_size\n",
        "      ito = i*batch_size+n\n",
        "\n",
        "      inputs, labels = Variable(inputs), Variable(labels)\n",
        "\n",
        "      outputs = model(inputs)\n",
        "      preds[ifrom:ito] = outputs.data.numpy()\n",
        "\n",
        "  return preds[0][0]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "LR56jmTDRWEI",
        "colab_type": "code",
        "outputId": "2db5ee2d-c40b-4a29-d7e4-afae4c2aa76b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "persection(6.2738166,-75.58028)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1.9719990748805465"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "metadata": {
        "id": "5jEw9XEkRWEO",
        "colab_type": "code",
        "outputId": "1b6a557f-0800-45a6-9bb9-9eff386d1325",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "print(\"Adquisicion\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Adquisicion\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "YJqz9jEpRWER",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df = pd.DataFrame(\n",
        "      {   \"link\": [0],\n",
        "          \"precio\": [1],\n",
        "      })"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "W8lwB5WVRWEU",
        "colab_type": "code",
        "outputId": "fd298f82-9053-42e2-85e8-4eded4979815",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "df[\"link\"][0]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "metadata": {
        "id": "04Mem440RWEX",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def filtros(df):\n",
        "  print(df[\"lat\"][0], df[\"lng\"][0], df[\"percepcion_0\"][0])\n",
        "  print(float(df[\"lat\"][0]) <= 6.1 or float(df[\"lat\"][0]) > 8. )\n",
        "  \n",
        "  if df[\"lat\"][0] == 0 or \\\n",
        "     df[\"lng\"][0] == 0 or \\\n",
        "     (df[\"percepcion_0\"][0] == 2.0001800290571454 and \\\n",
        "      df[\"percepcion_90\"][0] == 2.0001800290571454 and \\\n",
        "      df[\"percepcion_180\"][0] == 2.0001800290571454 and \\\n",
        "      df[\"percepcion_270\"][0] == 2.0001800290571454) or \\\n",
        "      float(df[\"lat\"][0]) <= 6.1 or \\\n",
        "      float(df[\"lat\"][0]) > 8. or \\\n",
        "      float(df[\"lng\"][0]) > -75.52 or \\\n",
        "      float(df[\"lng\"][0]) < -75.68:   \n",
        "    return df.drop(0)\n",
        "  else:\n",
        "    return df\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "E3ehgrYXRWEa",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df.iloc[[11]].reset_index()[\"percepcion_0\"][0] == 2.0001800290571454"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "oRyrfePJRWEe",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#filtros(df.iloc[[11]].reset_index())"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "jZPghPaHRWEh",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "YF2WuL2RRWEk",
        "colab_type": "code",
        "outputId": "6b3d35e2-48e7-4f85-f7b5-c7738b6b0bdf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "links = []\n",
        "precios = []\n",
        "df = pd.DataFrame(\n",
        "      {   \"link\": [],\n",
        "          \"precio\": [],\n",
        "          \"area\": [],\n",
        "          \"habitaciones\": [],\n",
        "          \"banos\": [],\n",
        "          \"mapa\": [],\n",
        "          \"lat\":[],\n",
        "          \"lng\":[],\n",
        "          \"descripcion\": [],\n",
        "          \"desc_texto\": [],\n",
        "          \"desc_title\": [],\n",
        "          \"desc_vend\": [],\n",
        "          \"desc_sec\": [],\n",
        "          \"desc_main\": [],\n",
        "          \"desc_ubica\": [],\n",
        "          \"imagens\":[],\n",
        "          \"percepcion_0\":[],\n",
        "          \"percepcion_90\":[],\n",
        "          \"percepcion_180\":[],\n",
        "          \"percepcion_270\":[],\n",
        "      })\n",
        "\n",
        "os.system(\"rm -r images/*\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "256"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        }
      ]
    },
    {
      "metadata": {
        "id": "MPSoRfXkRWEo",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def get_feature(url, precios):\n",
        "    #br2.get(url)\n",
        "    print(url)\n",
        "    #requests.get(\"https://apartamento.mercadolibre.com.co/MCO-437779427-apartamento-en-venta-_JM\").content[0:50]\n",
        "#    page2 = requests.get(url)\n",
        "#    time.sleep(10)\n",
        "    #fuente = br2.page_source.encode('utf-8')\n",
        "    fuente = str(get_fuente(url))\n",
        "    try:\n",
        "        atrivutos = str(fuente).split(\"vip-product-info__attributes-list\")[1].split(\"</section>\")[0]\n",
        "    except:\n",
        "        #br2.get(url)\n",
        "        page2 = requests.get(url)\n",
        "        time.sleep(10)\n",
        "        #fuente = br2.page_source.encode('utf-8')\n",
        "        fuente = str(page2.content)\n",
        "        try:\n",
        "            atrivutos = str(fuente).split(\"vip-product-info__attributes-list\")[1].split(\"</section>\")[0]\n",
        "        except:\n",
        "            atrivutos = 0\n",
        "    try:\n",
        "        area = atrivutos.split('vip-product-info__attribute-value\">')[1].split(\" \")[0]\n",
        "    except:\n",
        "        area = 0\n",
        "    try:\n",
        "        habitaciones = atrivutos.split('vip-product-info__attribute-value\">')[2].split(\"<\")[0]\n",
        "    except:\n",
        "        habitaciones = 0\n",
        "    try:\n",
        "        banos = atrivutos.split('vip-product-info__attribute-value\">')[3].split(\"<\")[0]\n",
        "    except:\n",
        "        banos = 0\n",
        "    try:\n",
        "        mapa = str(fuente).split('https://maps.googleapis.com/maps/api/staticmap?center=')[1].split(\"&\")[0].split(\"%2C\")\n",
        "    except:\n",
        "        mapa = ['0', '0']\n",
        "    try:\n",
        "        #descripcion = fuente.split('vip-section-description container\">')[1].split(\"</section>\")[0]\n",
        "        descripcion = get_description(fuente)\n",
        "    except:\n",
        "        descripcion = 0\n",
        "    try:\n",
        "        imagens0 = str(fuente).split(\"playedVideoTime = 0\")[1].split(\"items =\")[1].split(\"]\")[0].split(\"[\")[1]\n",
        "        imagens = [i[\"src\"] for i in eval(imagens0)]\n",
        "    except:\n",
        "        imagens = 0\n",
        "    percepcion_0 = persection(mapa[0],mapa[1], 0)\n",
        "    percepcion_90 = persection(mapa[0],mapa[1], 90)\n",
        "    percepcion_180 = persection(mapa[0],mapa[1], 180)\n",
        "    percepcion_270 = persection(mapa[0],mapa[1], 270)\n",
        "    \n",
        "    #print(area, habitaciones, banos, mapa)\n",
        "    #print(descripcion)\n",
        "    df = pd.DataFrame(\n",
        "      {   \"link\": url,\n",
        "          \"precio\": precios,\n",
        "          \"area\": area,\n",
        "          \"habitaciones\": habitaciones,\n",
        "          \"banos\": banos,\n",
        "          \"mapa\": [mapa],\n",
        "          \"lat\":mapa[0],\n",
        "          \"lng\":mapa[1],\n",
        "          \"descripcion\": [descripcion],\n",
        "          \"desc_texto\": descripcion[\"texto\"],\n",
        "          \"desc_title\": descripcion[\"title\"],\n",
        "          \"desc_vend\": descripcion[\"vendedor\"],\n",
        "          \"desc_sec\": descripcion[\"sec_descripcion\"],\n",
        "          \"desc_main\": descripcion[\"main_descripcion\"],\n",
        "          \"desc_ubica\": descripcion[\"ubicacion\"],\n",
        "          \"imagens\":[imagens],\n",
        "          \"percepcion_0\":percepcion_0,\n",
        "          \"percepcion_90\":percepcion_90,\n",
        "          \"percepcion_180\":percepcion_180,\n",
        "          \"percepcion_270\":percepcion_270,\n",
        "      })\n",
        "    \n",
        "    df = filtros(df)\n",
        "    return df\n",
        "\n",
        "\n",
        "url = \"https://inmuebles.mercadolibre.com.co/apartamentos/venta/antioquia/medellin/_Desde_1_OrderId_PRICE\"\n",
        "\n",
        "ind = 0\n",
        "for k in range(100):\n",
        "    break\n",
        "    #page = requests.get(url)\n",
        "    #br.get(url)\n",
        "    time.sleep(5)\n",
        "    fuente = str(get_fuente(url))\n",
        "    \n",
        "    #print len(str(br.page_source.encode('utf-8')).split(\"searchResults\"))\n",
        "    if len(fuente.split(\"searchResults\")) == 1:\n",
        "        print(\"termino\")\n",
        "        break\n",
        "\n",
        "    fuente = fuente.split(\"searchResults\")[1].split(\"item-url\")\n",
        "    fuente.pop(0)\n",
        "    ind = ind+len(fuente)\n",
        "\n",
        "    for link in fuente:\n",
        "        links.append(link.split(\" \")[0].split('\"')[1])\n",
        "        precios.append(int(link.split('price__fraction\">')[1].split('<')[0].replace('.', '')))\n",
        "        \n",
        "        df_a = get_feature(links[-1], precios[-1])\n",
        "\n",
        "        \n",
        "        clear_output(wait=True)\n",
        "        \n",
        "        df = df.append(df_a).reset_index(drop=True)\n",
        "        print(df[[\"area\",\"habitaciones\",\"mapa\",\"percepcion_0\",\"precio\"]].tail())\n",
        "        os.system(\"mkdir images/\"+str(df.shape[0] -1 ))\n",
        "        if df[\"imagens\"][df.shape[0]-1] != 0:\n",
        "          [ os.system(\"wget -O images/\"+str(df.shape[0] -1)+\"/\"+j.split(\"/\")[-1]+\" \"+j) for j in df[\"imagens\"][df.shape[0]-1] ]\n",
        "        elif df[\"imagens\"][df.shape[0]-1] == 0:\n",
        "          os.system(\"wget -O images/\"+str(df.shape[0] -1)+\"/No_image.jpg https://http2.mlstatic.com/resources/frontend/statics/img-not-available/1.0.0/V.jpg\")\n",
        "\n",
        "    print(\"_________________________________________________________\")\n",
        "    url = \"https://listado.mercadolibre.com.co/inmuebles/apartamentos/venta/antioquia/medellin/_Desde_\"+str(ind+1)+\"_OrderId_PRICE\"\n",
        "    print(url)\n",
        "    #df.to_csv(\"df.csv\", sep=';')\n",
        "    #!./dropbox_uploader.sh upload df.csv /\n",
        "    print(\"_________________________________________________________\")\n",
        "    #      df[\"lat\"][0] <= 6.1 or \\\n",
        "#      df[\"lat\"][0] > 8 or \\\n",
        "#      df[\"lng\"][0] > -75.52 or \\\n",
        "#      df[\"lng\"][0] < -75.68:   "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "6ZQVai8QRWEr",
        "colab_type": "code",
        "outputId": "5872fefb-8ad6-4486-946e-0f3ede9ddf3a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        }
      },
      "cell_type": "code",
      "source": [
        "!curl \"https://raw.githubusercontent.com/andreafabrizi/Dropbox-Uploader/master/dropbox_uploader.sh\" -o dropbox_uploader.sh\n",
        "\n",
        "\n",
        "!chmod +x dropbox_uploader.sh\n",
        "!./dropbox_uploader.sh\n",
        "#iJMCDCbjLNIAAAAAAAANHFZu-tNNmw6FXMABCQuh-2f03xDCxv33TFd9yO0XQlqE\n",
        "#%notebook ./filename.ipynb\n",
        "#!./dropbox_uploader.sh upload filename.ipynb /"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n",
            "                                 Dload  Upload   Total   Spent    Left  Speed\n",
            "\r  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0\r100 49655  100 49655    0     0   418k      0 --:--:-- --:--:-- --:--:--  418k\n",
            "\n",
            " This is the first time you run this script, please follow the instructions:\n",
            "\n",
            " 1) Open the following URL in your Browser, and log in using your account: https://www.dropbox.com/developers/apps\n",
            " 2) Click on \"Create App\", then select \"Dropbox API app\"\n",
            " 3) Now go on with the configuration, choosing the app permissions and access restrictions to your DropBox folder\n",
            " 4) Enter the \"App Name\" that you prefer (e.g. MyUploader830410070244)\n",
            "\n",
            " Now, click on the \"Create App\" button.\n",
            "\n",
            " When your new App is successfully created, please click on the Generate button\n",
            " under the 'Generated access token' section, then copy and paste the new access token here:\n",
            "\n",
            " # Access token: iJMCDCbjLNIAAAAAAAANHFZu-tNNmw6FXMABCQuh-2f03xDCxv33TFd9yO0XQlqE\n",
            "\n",
            " > The access token is iJMCDCbjLNIAAAAAAAANHFZu-tNNmw6FXMABCQuh-2f03xDCxv33TFd9yO0XQlqE. Looks ok? [y/N]: y\n",
            "   The configuration has been saved.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "5OmqlhvkRWEu",
        "colab_type": "code",
        "outputId": "5a9586d8-2412-4e9e-f58b-883cbc9bb7a0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "!./dropbox_uploader.sh download df.csv ."
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " > Downloading \"/df.csv\" to \"/content/df.csv\"... DONE\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "ywrrTkXiRWEx",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df = pd.read_csv(\"df.csv\", delimiter=\";\", index_col=\"Unnamed: 0\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "GCu5jXz6RWEz",
        "colab_type": "code",
        "outputId": "c54de494-af1d-45bc-d261-9460343adaf3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        }
      },
      "cell_type": "code",
      "source": [
        "df.tail()"
      ],
      "execution_count": 12,
      "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>area</th>\n",
              "      <th>banos</th>\n",
              "      <th>desc_main</th>\n",
              "      <th>desc_sec</th>\n",
              "      <th>desc_texto</th>\n",
              "      <th>desc_title</th>\n",
              "      <th>desc_ubica</th>\n",
              "      <th>desc_vend</th>\n",
              "      <th>descripcion</th>\n",
              "      <th>habitaciones</th>\n",
              "      <th>imagens</th>\n",
              "      <th>lat</th>\n",
              "      <th>link</th>\n",
              "      <th>lng</th>\n",
              "      <th>mapa</th>\n",
              "      <th>percepcion_0</th>\n",
              "      <th>percepcion_180</th>\n",
              "      <th>percepcion_270</th>\n",
              "      <th>percepcion_90</th>\n",
              "      <th>precio</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>936</th>\n",
              "      <td>77</td>\n",
              "      <td>2</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 77 Op...</td>\n",
              "      <td>0</td>\n",
              "      <td>APTO PISO SEPTIMO: Consta de 77 m2, 3 habitac...</td>\n",
              "      <td>Apto 77 M2 3 Hab Cerca De Estacion San Javier</td>\n",
              "      <td>Ubicacion Medellin 3053181060, San Javier, Me...</td>\n",
              "      <td>Asiestucasa</td>\n",
              "      <td>{'title': 'Apto 77 M2 3 Hab Cerca De Estacion ...</td>\n",
              "      <td>3</td>\n",
              "      <td>['https://http2.mlstatic.com/none-D_NQ_NP_8829...</td>\n",
              "      <td>6.255651</td>\n",
              "      <td>https://apartamento.mercadolibre.com.co/MCO-48...</td>\n",
              "      <td>-75.618627</td>\n",
              "      <td>['6.2556509', '-75.6186265']</td>\n",
              "      <td>4.852417</td>\n",
              "      <td>4.220643</td>\n",
              "      <td>3.951810</td>\n",
              "      <td>4.010629</td>\n",
              "      <td>200000000.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>937</th>\n",
              "      <td>55</td>\n",
              "      <td>2</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 55 Op...</td>\n",
              "      <td>Otros Condicion del item: Usado</td>\n",
              "      <td>Informacion detallada. Hermoso Apartamento ub...</td>\n",
              "      <td>Apartamento En Belen Los Alpes</td>\n",
              "      <td>Ubicacion Belen Los Alpes, Medellin, Antioquia</td>\n",
              "      <td>Luz Elena Meneses Moreno</td>\n",
              "      <td>{'title': 'Apartamento En Belen Los Alpes', 'v...</td>\n",
              "      <td>2</td>\n",
              "      <td>['https://http2.mlstatic.com/none-D_NQ_NP_8414...</td>\n",
              "      <td>6.228799</td>\n",
              "      <td>https://apartamento.mercadolibre.com.co/MCO-48...</td>\n",
              "      <td>-75.606185</td>\n",
              "      <td>['6.2287988', '-75.6061855']</td>\n",
              "      <td>4.482049</td>\n",
              "      <td>4.327452</td>\n",
              "      <td>5.193517</td>\n",
              "      <td>5.267939</td>\n",
              "      <td>200000000.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>938</th>\n",
              "      <td>67</td>\n",
              "      <td>2</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 67 Op...</td>\n",
              "      <td>0</td>\n",
              "      <td>Se vende apartamento en Rodeo Alto, 3 habitac...</td>\n",
              "      <td>Se Vende Apartamento En Rodeo Alto</td>\n",
              "      <td>Ubicacion Medellin, Antioquia</td>\n",
              "      <td>Prado Gestion Inmobiliaria S.a.s</td>\n",
              "      <td>{'title': 'Se Vende Apartamento En Rodeo Alto'...</td>\n",
              "      <td>3</td>\n",
              "      <td>['https://http2.mlstatic.com/none-D_NQ_NP_9144...</td>\n",
              "      <td>6.244203</td>\n",
              "      <td>https://apartamento.mercadolibre.com.co/MCO-48...</td>\n",
              "      <td>-75.581212</td>\n",
              "      <td>['6.244203', '-75.5812119']</td>\n",
              "      <td>4.042085</td>\n",
              "      <td>4.859144</td>\n",
              "      <td>4.474911</td>\n",
              "      <td>4.472118</td>\n",
              "      <td>200000000.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>939</th>\n",
              "      <td>67</td>\n",
              "      <td>2</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 67 Op...</td>\n",
              "      <td>Caracteristicas adicionales Balcon Cocina Int...</td>\n",
              "      <td>SE VENDE APARTAMETO EN LOS COLORES en Centro ...</td>\n",
              "      <td>Se Vende Apartameto En Los Colores</td>\n",
              "      <td>Ubicacion Centro Occidente, Medellin, Antioquia</td>\n",
              "      <td>Prado Gestion Inmobiliaria S.a.s</td>\n",
              "      <td>{'title': 'Se Vende Apartameto En Los Colores'...</td>\n",
              "      <td>3</td>\n",
              "      <td>['https://http2.mlstatic.com/none-D_NQ_NP_7736...</td>\n",
              "      <td>6.253453</td>\n",
              "      <td>https://apartamento.mercadolibre.com.co/MCO-48...</td>\n",
              "      <td>-75.624792</td>\n",
              "      <td>['6.2534528', '-75.6247916']</td>\n",
              "      <td>5.527111</td>\n",
              "      <td>4.960363</td>\n",
              "      <td>4.864713</td>\n",
              "      <td>5.027576</td>\n",
              "      <td>200000000.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>940</th>\n",
              "      <td>67</td>\n",
              "      <td>2</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 67 Op...</td>\n",
              "      <td>Caracteristicas adicionales Balcon Calefaccio...</td>\n",
              "      <td>HERMOSO APARTAMENTO EN BELEN LA MOTA CERCA A ...</td>\n",
              "      <td>Hermoso Apartamento En Belen La Mota</td>\n",
              "      <td>Ubicacion La Mota, Medellin, Antioquia</td>\n",
              "      <td>Su Propiedad</td>\n",
              "      <td>{'title': 'Hermoso Apartamento En Belen La Mot...</td>\n",
              "      <td>2</td>\n",
              "      <td>['https://http2.mlstatic.com/none-D_NQ_NP_9028...</td>\n",
              "      <td>6.208338</td>\n",
              "      <td>https://apartamento.mercadolibre.com.co/MCO-48...</td>\n",
              "      <td>-75.599514</td>\n",
              "      <td>['6.2083376', '-75.5995139']</td>\n",
              "      <td>4.486433</td>\n",
              "      <td>5.168179</td>\n",
              "      <td>6.034755</td>\n",
              "      <td>4.307685</td>\n",
              "      <td>200000000.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    area banos                                          desc_main  \\\n",
              "936   77     2   Inmueble: Apartamento Metros de const.: 77 Op...   \n",
              "937   55     2   Inmueble: Apartamento Metros de const.: 55 Op...   \n",
              "938   67     2   Inmueble: Apartamento Metros de const.: 67 Op...   \n",
              "939   67     2   Inmueble: Apartamento Metros de const.: 67 Op...   \n",
              "940   67     2   Inmueble: Apartamento Metros de const.: 67 Op...   \n",
              "\n",
              "                                              desc_sec  \\\n",
              "936                                                  0   \n",
              "937                   Otros Condicion del item: Usado    \n",
              "938                                                  0   \n",
              "939   Caracteristicas adicionales Balcon Cocina Int...   \n",
              "940   Caracteristicas adicionales Balcon Calefaccio...   \n",
              "\n",
              "                                            desc_texto  \\\n",
              "936   APTO PISO SEPTIMO: Consta de 77 m2, 3 habitac...   \n",
              "937   Informacion detallada. Hermoso Apartamento ub...   \n",
              "938   Se vende apartamento en Rodeo Alto, 3 habitac...   \n",
              "939   SE VENDE APARTAMETO EN LOS COLORES en Centro ...   \n",
              "940   HERMOSO APARTAMENTO EN BELEN LA MOTA CERCA A ...   \n",
              "\n",
              "                                        desc_title  \\\n",
              "936  Apto 77 M2 3 Hab Cerca De Estacion San Javier   \n",
              "937                 Apartamento En Belen Los Alpes   \n",
              "938             Se Vende Apartamento En Rodeo Alto   \n",
              "939             Se Vende Apartameto En Los Colores   \n",
              "940           Hermoso Apartamento En Belen La Mota   \n",
              "\n",
              "                                            desc_ubica  \\\n",
              "936   Ubicacion Medellin 3053181060, San Javier, Me...   \n",
              "937    Ubicacion Belen Los Alpes, Medellin, Antioquia    \n",
              "938                     Ubicacion Medellin, Antioquia    \n",
              "939   Ubicacion Centro Occidente, Medellin, Antioquia    \n",
              "940            Ubicacion La Mota, Medellin, Antioquia    \n",
              "\n",
              "                              desc_vend  \\\n",
              "936                        Asiestucasa    \n",
              "937           Luz Elena Meneses Moreno    \n",
              "938   Prado Gestion Inmobiliaria S.a.s    \n",
              "939   Prado Gestion Inmobiliaria S.a.s    \n",
              "940                       Su Propiedad    \n",
              "\n",
              "                                           descripcion habitaciones  \\\n",
              "936  {'title': 'Apto 77 M2 3 Hab Cerca De Estacion ...            3   \n",
              "937  {'title': 'Apartamento En Belen Los Alpes', 'v...            2   \n",
              "938  {'title': 'Se Vende Apartamento En Rodeo Alto'...            3   \n",
              "939  {'title': 'Se Vende Apartameto En Los Colores'...            3   \n",
              "940  {'title': 'Hermoso Apartamento En Belen La Mot...            2   \n",
              "\n",
              "                                               imagens       lat  \\\n",
              "936  ['https://http2.mlstatic.com/none-D_NQ_NP_8829...  6.255651   \n",
              "937  ['https://http2.mlstatic.com/none-D_NQ_NP_8414...  6.228799   \n",
              "938  ['https://http2.mlstatic.com/none-D_NQ_NP_9144...  6.244203   \n",
              "939  ['https://http2.mlstatic.com/none-D_NQ_NP_7736...  6.253453   \n",
              "940  ['https://http2.mlstatic.com/none-D_NQ_NP_9028...  6.208338   \n",
              "\n",
              "                                                  link        lng  \\\n",
              "936  https://apartamento.mercadolibre.com.co/MCO-48... -75.618627   \n",
              "937  https://apartamento.mercadolibre.com.co/MCO-48... -75.606185   \n",
              "938  https://apartamento.mercadolibre.com.co/MCO-48... -75.581212   \n",
              "939  https://apartamento.mercadolibre.com.co/MCO-48... -75.624792   \n",
              "940  https://apartamento.mercadolibre.com.co/MCO-48... -75.599514   \n",
              "\n",
              "                             mapa  percepcion_0  percepcion_180  \\\n",
              "936  ['6.2556509', '-75.6186265']      4.852417        4.220643   \n",
              "937  ['6.2287988', '-75.6061855']      4.482049        4.327452   \n",
              "938   ['6.244203', '-75.5812119']      4.042085        4.859144   \n",
              "939  ['6.2534528', '-75.6247916']      5.527111        4.960363   \n",
              "940  ['6.2083376', '-75.5995139']      4.486433        5.168179   \n",
              "\n",
              "     percepcion_270  percepcion_90       precio  \n",
              "936        3.951810       4.010629  200000000.0  \n",
              "937        5.193517       5.267939  200000000.0  \n",
              "938        4.474911       4.472118  200000000.0  \n",
              "939        4.864713       5.027576  200000000.0  \n",
              "940        6.034755       4.307685  200000000.0  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "metadata": {
        "id": "mhZ5c6V1RWE2",
        "colab_type": "code",
        "outputId": "cc46dfe5-e02e-443c-e0d1-e31a5bf45b8c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "print(\"Get Distances - Interes Points\")"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Get Distances - Interes Points\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "U5MaTND0RWE5",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df[\"lat\"] = df[\"lat\"].astype(float)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "oAwa1bjHRWE8",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "\n",
        "df[\"lng\"] = df[\"mapa\"].map(lambda x: eval(eval(x)[0]))\n",
        "df[\"lat\"] = df[\"mapa\"].map(lambda x: eval(eval(x)[1]))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "s-DhwnIrRWE_",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def get_Estrato(x):\n",
        "  try:\n",
        "    y= x.split(\"Estrato:\")[1].split(\" \")\n",
        "    return int(y[1])\n",
        "  except:\n",
        "    return np.nan\n",
        "    \n",
        "df[\"Estrato\"] = df[\"desc_main\"].map(lambda x: get_Estrato(x))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "vnb5dTmMRWFC",
        "colab_type": "code",
        "outputId": "537361cb-8eb3-4225-a846-020d85aad03c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 80
        }
      },
      "cell_type": "code",
      "source": [
        "Oscar3.O_check_base(df[[\"Estrato\"]])"
      ],
      "execution_count": 17,
      "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>columns</th>\n",
              "      <th>types</th>\n",
              "      <th>missing</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Estrato</td>\n",
              "      <td>float64</td>\n",
              "      <td>263.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   columns    types  missing\n",
              "0  Estrato  float64    263.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "metadata": {
        "id": "fiHTAZ7-RWFF",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df[\"Estrato\"] = df[[\"Estrato\"]].fillna(df[\"Estrato\"].mode()[0])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "GXhkTpvWRWFI",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df = df[df[\"area\"] != \"3</span>\\n\\t\\t\\t</li>\\n\"]\n",
        "df = df[df[\"area\"] != '1</span>\\n\\t\\t\\t</li>\\n']\n",
        "#df[df[\"banos\"] == 'M\\xc3\\xa1s de 10']\n",
        "#Oscar3.O_print_full(df[df[\"banos\"] == 'M\\xc3\\xa1s de 10'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "jvEutKJvRWFK",
        "colab_type": "code",
        "outputId": "afef4434-8c65-43c6-92a6-6c0291e556aa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        }
      },
      "cell_type": "code",
      "source": [
        "Oscar3.O_print_full(df[df[\"habitaciones\"] == 'M\\xc3\\xa1s de 10'])"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "    area banos                                                                                                                                                                               desc_main                                                                                                                                                                                                                                                                                                                    desc_sec                                                                                                                                                                                                                                                                                                                                                                                                       desc_texto                       desc_title                                         desc_ubica            desc_vend  \\\n",
            "5    1    0      Inmueble: Apartamento Metros de const.: 1 Operacion: Venta Habitaciones: Mas de 10 Banos: No tiene Anos de antiguedad: 1 Metros de terreno: 1                                           Otros Condicion del item: Usado                                                                                                                                                                                                                                                                                            0                                                                                                                                                                                                                                                                                                                                                                                                                Apartamentos En Venta             Ubicacion Medellin, Antioquia                     0                     \n",
            "604  63   2      Inmueble: Apartamento Metros de const.: 63 Operacion: Venta Valor administracion ($): 148000 Habitaciones: Mas de 10 Banos: 2 Anos de antiguedad: 0 Estrato: 4 Metros de terreno: 63    Caracteristicas adicionales Balcon Calefaccion o Chimenea Cocina Integral Garaje Piscina Vista panoramica Cercanias a menos de dos cuadras Centro comercial Colegio Parque Universidad Instalaciones y comodidades Circuito Cerrado de TV Conjunto Cerrado Ascensores Parque infantil Salon de Fiesta Vigilancia privada    Apartamento en venta en rodeo alto con area de 63m2, 3 alcobas , 2 banos, sala comedor, cocina integral cerrada, balcon con vista panoramica, calentador de paso, parqueadero cubierto. Unidad cerrada con vigilancia las 24 horas, pscina, placas deportivas, amplias zonas verdes, sendero ecologico, parqueadero de visitantes. Valor de venta: 165.000.000. NEGOCIABLES ! Urve inmobiliaria Colombia SAS.   Apartamento En Venta Rodeo Alto   Ubicacion Belen Rodeo Alto, Medellin, Antioquia    Patricia Valencia    \n",
            "\n",
            "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      descripcion habitaciones                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         imagens                  lat                                                                                       link                  lng                          mapa  \\\n",
            "5    {'title': 'Apartamentos En Venta', 'vendedor': 0, 'texto': 0, 'ubicacion': ' Ubicacion Medellin, Antioquia ', 'main_descripcion': ' Inmueble: Apartamento Metros de const.: 1 Operacion: Venta Habitaciones: Mas de 10 Banos: No tiene Anos de antiguedad: 1 Metros de terreno: 1 ', 'sec_descripcion': ' Otros Condicion del item: Usado '}                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  Más de 10    0                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            -75.58  https://apartamento.mercadolibre.com.co/MCO-494405318-apartamentos-en-venta-_JM                           6.24  ['6.2447031', '-75.5807149']   \n",
            "604  {'title': 'Apartamento En Venta Rodeo Alto', 'vendedor': ' Patricia Valencia ', 'texto': ' Apartamento en venta en rodeo alto con area de 63m2, 3 alcobas , 2 banos, sala comedor, cocina integral cerrada, balcon con vista panoramica, calentador de paso, parqueadero cubierto. Unidad cerrada con vigilancia las 24 horas, pscina, placas deportivas, amplias zonas verdes, sendero ecologico, parqueadero de visitantes. Valor de venta: 165.000.000. NEGOCIABLES ! Urve inmobiliaria Colombia SAS. ', 'ubicacion': ' Ubicacion Belen Rodeo Alto, Medellin, Antioquia ', 'main_descripcion': ' Inmueble: Apartamento Metros de const.: 63 Operacion: Venta Valor administracion ($): 148000 Habitaciones: Mas de 10 Banos: 2 Anos de antiguedad: 0 Estrato: 4 Metros de terreno: 63 ', 'sec_descripcion': ' Caracteristicas adicionales Balcon Calefaccion o Chimenea Cocina Integral Garaje Piscina Vista panoramica Cercanias a menos de dos cuadras Centro comercial Colegio Parque Universidad Instalaciones y comodidades Circuito Cerrado de TV Conjunto Cerrado Ascensores Parque infantil Salon de Fiesta Vigilancia privada '}  Más de 10    ['https://http2.mlstatic.com/none-D_NQ_NP_715780-MCO27884834507_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_756261-MCO27884834510_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_823084-MCO27884834506_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_974547-MCO27884834504_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_831514-MCO27884834505_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_853051-MCO27884834502_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_655567-MCO27884834508_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_982623-MCO27884834509_082018-F.jpg', 'https://http2.mlstatic.com/none-D_NQ_NP_615278-MCO27884834503_082018-F.jpg']               -75.61  https://apartamento.mercadolibre.com.co/MCO-479630765-apartamento-en-venta-rodeo-alto-_JM                 6.20  ['6.2036942', '-75.6050723']   \n",
            "\n",
            "            percepcion_0       percepcion_180       percepcion_270        percepcion_90               precio              Estrato  \n",
            "5                   5.18                 5.37                 5.30                 4.40         1,000,000.00                 3.00  \n",
            "604                 5.06                 5.37                 5.43                 5.10       165,000,000.00                 4.00  \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "tN5ZNjnoRWFQ",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#Oscar3.O_print_full(df.iloc[603])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "4J0CH8YcRWFX",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df.drop(index=5, inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "a5-vjumJRWFb",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df.at[407,\"banos\"] = 2"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fFFLZI-zRWFg",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df.at[604,\"habitaciones\"] = 3"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "cIIB7Z8tRWFk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df.iloc[603]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "i6ABtLbqRWFq",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df.iloc[407]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "roSSrD2FRWFt",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df[\"area\"] = df[\"area\"].astype(int)\n",
        "df[\"banos\"] = df[\"banos\"].astype(int)\n",
        "df[\"habitaciones\"] = df[\"habitaciones\"].astype(int)\n",
        "df[\"Estrato\"] = df[\"Estrato\"].astype(int)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Dkq9NMhDRWFw",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df[[\"desc_main\"]].head()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Q-xiEMCBRWF0",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df[\"lat\"]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Tl7vt7-gRWF4",
        "colab_type": "code",
        "outputId": "4dada38f-86f4-4623-f7c5-7207001a8d72",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 700
        }
      },
      "cell_type": "code",
      "source": [
        "Oscar3.O_check_base(df)"
      ],
      "execution_count": 30,
      "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>columns</th>\n",
              "      <th>types</th>\n",
              "      <th>missing</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>area</td>\n",
              "      <td>int64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>banos</td>\n",
              "      <td>int64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>desc_main</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>desc_sec</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>desc_texto</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>desc_title</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>desc_ubica</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>desc_vend</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>descripcion</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>habitaciones</td>\n",
              "      <td>int64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>imagens</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>lat</td>\n",
              "      <td>float64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>link</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>lng</td>\n",
              "      <td>float64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>mapa</td>\n",
              "      <td>object</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>percepcion_0</td>\n",
              "      <td>float64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>percepcion_180</td>\n",
              "      <td>float64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>percepcion_270</td>\n",
              "      <td>float64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>percepcion_90</td>\n",
              "      <td>float64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>precio</td>\n",
              "      <td>float64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>Estrato</td>\n",
              "      <td>int64</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           columns    types  missing\n",
              "0             area    int64      0.0\n",
              "1            banos    int64      0.0\n",
              "2        desc_main   object      0.0\n",
              "3         desc_sec   object      0.0\n",
              "4       desc_texto   object      0.0\n",
              "5       desc_title   object      0.0\n",
              "6       desc_ubica   object      0.0\n",
              "7        desc_vend   object      0.0\n",
              "8      descripcion   object      0.0\n",
              "9     habitaciones    int64      0.0\n",
              "10         imagens   object      0.0\n",
              "11             lat  float64      0.0\n",
              "12            link   object      0.0\n",
              "13             lng  float64      0.0\n",
              "14            mapa   object      0.0\n",
              "15    percepcion_0  float64      0.0\n",
              "16  percepcion_180  float64      0.0\n",
              "17  percepcion_270  float64      0.0\n",
              "18   percepcion_90  float64      0.0\n",
              "19          precio  float64      0.0\n",
              "20         Estrato    int64      0.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "metadata": {
        "id": "WiYumkC6RWF9",
        "colab_type": "code",
        "outputId": "8d40b62c-32f6-41fb-d0ed-d7fb2ad00f6f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        }
      },
      "cell_type": "code",
      "source": [
        "print(df.columns.tolist())"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['area', 'banos', 'desc_main', 'desc_sec', 'desc_texto', 'desc_title', 'desc_ubica', 'desc_vend', 'descripcion', 'habitaciones', 'imagens', 'lat', 'link', 'lng', 'mapa', 'percepcion_0', 'percepcion_180', 'percepcion_270', 'percepcion_90', 'precio', 'Estrato']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "-T8wVwxZRWGA",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df_0 = df\n",
        "df[\"description\"] = df_0['desc_main'] + df_0['desc_sec'] + df_0['desc_texto'] + df_0['desc_title'] + df_0['desc_ubica'] + df_0['desc_vend']\n",
        "df = df[['area', 'banos', 'habitaciones', 'lat', 'lng', 'percepcion_0', 'percepcion_180', 'percepcion_270', 'percepcion_90', 'Estrato', 'description', 'precio']]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fwnhuBFWRWGE",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "df = df[df[\"precio\"] >0.8e8]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "YJmYs1xMRWGK",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df = df.drop(columns=['desc_main', 'desc_sec', 'desc_texto', 'desc_title', 'desc_ubica', 'desc_vend', 'descripcion','imagens','mapa','link'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "72KwWODjRWGM",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#lista"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "jDhXmZ93RWGQ",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "ind  = df.columns.tolist().index(\"precio\")\n",
        "lista = df.columns.tolist()\n",
        "V_objetivo = lista.pop(ind)\n",
        "X = df[lista]\n",
        "y = df[[V_objetivo]]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "bB8N1KEERWGT",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#X.head().columns.tolist()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "g5mwf-1uRWGX",
        "colab_type": "code",
        "outputId": "243f14f5-6747-4105-a43a-4173740a36b4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "cell_type": "code",
      "source": [
        "X.head()"
      ],
      "execution_count": 38,
      "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>area</th>\n",
              "      <th>banos</th>\n",
              "      <th>habitaciones</th>\n",
              "      <th>lat</th>\n",
              "      <th>lng</th>\n",
              "      <th>percepcion_0</th>\n",
              "      <th>percepcion_180</th>\n",
              "      <th>percepcion_270</th>\n",
              "      <th>percepcion_90</th>\n",
              "      <th>Estrato</th>\n",
              "      <th>description</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>53</th>\n",
              "      <td>41</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>-75.598370</td>\n",
              "      <td>6.278447</td>\n",
              "      <td>4.707252</td>\n",
              "      <td>4.382922</td>\n",
              "      <td>5.228550</td>\n",
              "      <td>4.345707</td>\n",
              "      <td>2</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 41 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>54</th>\n",
              "      <td>36</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>-75.595000</td>\n",
              "      <td>6.282000</td>\n",
              "      <td>5.046718</td>\n",
              "      <td>5.681197</td>\n",
              "      <td>4.264784</td>\n",
              "      <td>5.246507</td>\n",
              "      <td>3</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 36 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>55</th>\n",
              "      <td>43</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>-75.658270</td>\n",
              "      <td>6.186444</td>\n",
              "      <td>5.616747</td>\n",
              "      <td>5.264588</td>\n",
              "      <td>4.858189</td>\n",
              "      <td>5.142254</td>\n",
              "      <td>3</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 43 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>56</th>\n",
              "      <td>60</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>-75.595636</td>\n",
              "      <td>6.200800</td>\n",
              "      <td>5.077016</td>\n",
              "      <td>5.385752</td>\n",
              "      <td>5.238197</td>\n",
              "      <td>5.441324</td>\n",
              "      <td>2</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 60 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>57</th>\n",
              "      <td>48</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>-75.598985</td>\n",
              "      <td>6.260916</td>\n",
              "      <td>4.849460</td>\n",
              "      <td>5.030130</td>\n",
              "      <td>4.857601</td>\n",
              "      <td>5.665002</td>\n",
              "      <td>4</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 48 Op...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    area  banos  habitaciones        lat       lng  percepcion_0  \\\n",
              "53    41      1             2 -75.598370  6.278447      4.707252   \n",
              "54    36      1             3 -75.595000  6.282000      5.046718   \n",
              "55    43      1             2 -75.658270  6.186444      5.616747   \n",
              "56    60      1             2 -75.595636  6.200800      5.077016   \n",
              "57    48      1             2 -75.598985  6.260916      4.849460   \n",
              "\n",
              "    percepcion_180  percepcion_270  percepcion_90  Estrato  \\\n",
              "53        4.382922        5.228550       4.345707        2   \n",
              "54        5.681197        4.264784       5.246507        3   \n",
              "55        5.264588        4.858189       5.142254        3   \n",
              "56        5.385752        5.238197       5.441324        2   \n",
              "57        5.030130        4.857601       5.665002        4   \n",
              "\n",
              "                                          description  \n",
              "53   Inmueble: Apartamento Metros de const.: 41 Op...  \n",
              "54   Inmueble: Apartamento Metros de const.: 36 Op...  \n",
              "55   Inmueble: Apartamento Metros de const.: 43 Op...  \n",
              "56   Inmueble: Apartamento Metros de const.: 60 Op...  \n",
              "57   Inmueble: Apartamento Metros de const.: 48 Op...  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 38
        }
      ]
    },
    {
      "metadata": {
        "id": "CgS3q0PhRWGb",
        "colab_type": "code",
        "outputId": "defd50b8-f4b1-4182-e3d1-98ae6d07e3fb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 327
        }
      },
      "cell_type": "code",
      "source": [
        "y.hist()"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7ff9d1d29d50>]],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAEUCAYAAADUVaY3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7ff9d1cc3290>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "LcRLw4zfRWGf",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import RandomForestRegressor\n",
        "from sklearn.datasets import make_regression\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "7PCMnJFHRWGh",
        "colab_type": "code",
        "outputId": "daf5bab2-ccb6-4d34-a9f4-9deeae52a5da",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 327
        }
      },
      "cell_type": "code",
      "source": [
        "y_train.hist()"
      ],
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7ff9ceec9890>]],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAEUCAYAAADUVaY3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUJekghjqklQQQ12SCvL/YMrRiofdeKAAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7ff9ceec9190>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "XI6t2pmuRWGj",
        "colab_type": "code",
        "outputId": "550fb969-8346-4db3-f76d-4ecaa2981a05",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 327
        }
      },
      "cell_type": "code",
      "source": [
        "y_test.hist()"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7ff9cec03510>]],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAW8AAAEUCAYAAADk2bcWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSQUyvCWpQP8Pe7vdDesMLs0AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7ff9ceba35d0>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "OJOoHdrGRWGm",
        "colab_type": "code",
        "outputId": "dd8bbd52-2b19-4dcc-b9f5-1ad64eb3e424",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "cell_type": "code",
      "source": [
        "X_train.columns[0 : -1 ]"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index([u'area', u'banos', u'habitaciones', u'lat', u'lng', u'percepcion_0',\n",
              "       u'percepcion_180', u'percepcion_270', u'percepcion_90', u'Estrato'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 43
        }
      ]
    },
    {
      "metadata": {
        "id": "tZSiaCk1RWGp",
        "colab_type": "code",
        "outputId": "eb4a8d66-eb49-4244-c81c-97f92eccf5f7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "cell_type": "code",
      "source": [
        "X_train.head()"
      ],
      "execution_count": 44,
      "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>area</th>\n",
              "      <th>banos</th>\n",
              "      <th>habitaciones</th>\n",
              "      <th>lat</th>\n",
              "      <th>lng</th>\n",
              "      <th>percepcion_0</th>\n",
              "      <th>percepcion_180</th>\n",
              "      <th>percepcion_270</th>\n",
              "      <th>percepcion_90</th>\n",
              "      <th>Estrato</th>\n",
              "      <th>description</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>309</th>\n",
              "      <td>45</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>-75.560286</td>\n",
              "      <td>6.252380</td>\n",
              "      <td>4.318297</td>\n",
              "      <td>5.006413</td>\n",
              "      <td>4.393741</td>\n",
              "      <td>4.716533</td>\n",
              "      <td>4</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 45 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>228</th>\n",
              "      <td>45</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>-75.599427</td>\n",
              "      <td>6.279888</td>\n",
              "      <td>4.623398</td>\n",
              "      <td>5.794126</td>\n",
              "      <td>5.650282</td>\n",
              "      <td>5.263768</td>\n",
              "      <td>3</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 45 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110</th>\n",
              "      <td>38</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>-75.656639</td>\n",
              "      <td>6.186241</td>\n",
              "      <td>4.402307</td>\n",
              "      <td>5.063399</td>\n",
              "      <td>4.032859</td>\n",
              "      <td>4.199199</td>\n",
              "      <td>3</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 38 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>564</th>\n",
              "      <td>57</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>-75.556054</td>\n",
              "      <td>6.237129</td>\n",
              "      <td>4.398416</td>\n",
              "      <td>5.371661</td>\n",
              "      <td>5.485881</td>\n",
              "      <td>5.190678</td>\n",
              "      <td>4</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 57 Op...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>121</th>\n",
              "      <td>28</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>-75.574635</td>\n",
              "      <td>6.227697</td>\n",
              "      <td>4.541640</td>\n",
              "      <td>4.799234</td>\n",
              "      <td>4.683369</td>\n",
              "      <td>4.625693</td>\n",
              "      <td>3</td>\n",
              "      <td>Inmueble: Apartamento Metros de const.: 28 Op...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     area  banos  habitaciones        lat       lng  percepcion_0  \\\n",
              "309    45      2             1 -75.560286  6.252380      4.318297   \n",
              "228    45      1             2 -75.599427  6.279888      4.623398   \n",
              "110    38      1             2 -75.656639  6.186241      4.402307   \n",
              "564    57      2             3 -75.556054  6.237129      4.398416   \n",
              "121    28      1             1 -75.574635  6.227697      4.541640   \n",
              "\n",
              "     percepcion_180  percepcion_270  percepcion_90  Estrato  \\\n",
              "309        5.006413        4.393741       4.716533        4   \n",
              "228        5.794126        5.650282       5.263768        3   \n",
              "110        5.063399        4.032859       4.199199        3   \n",
              "564        5.371661        5.485881       5.190678        4   \n",
              "121        4.799234        4.683369       4.625693        3   \n",
              "\n",
              "                                           description  \n",
              "309   Inmueble: Apartamento Metros de const.: 45 Op...  \n",
              "228   Inmueble: Apartamento Metros de const.: 45 Op...  \n",
              "110   Inmueble: Apartamento Metros de const.: 38 Op...  \n",
              "564   Inmueble: Apartamento Metros de const.: 57 Op...  \n",
              "121   Inmueble: Apartamento Metros de const.: 28 Op...  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "metadata": {
        "id": "oj8wMGC9RWGt",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#df_0['desc_main'] + df_0['desc_sec'] + df_0['desc_texto'] + df_0['desc_title'] + df_0['desc_ubica'] + df_0['desc_vend']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Go9q3jALRWGv",
        "colab_type": "code",
        "outputId": "4bbb168f-5d5f-45b7-f71f-fe2f08d26f0d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "cell_type": "code",
      "source": [
        "try:\n",
        "    from keras.preprocessing.text import Tokenizer\n",
        "    from keras.preprocessing.sequence import pad_sequences\n",
        "    from keras.layers import Input, Embedding, LSTM, Dense, SimpleRNN, concatenate, Dropout\n",
        "    from keras.models import Model\n",
        "    from keras.optimizers import RMSprop\n",
        "    from keras.callbacks import EarlyStopping \n",
        "except ModuleNotFoundError:\n",
        "    from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "    from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "    from tensorflow.keras.layers import Input, Embedding, LSTM, Dense, SimpleRNN, concatenate\n",
        "    from tensorflow.keras.models import Model\n",
        "    from tensorflow.keras.optimizers import RMSprop\n",
        "    from tensorflow.keras.callbacks import EarlyStopping \n",
        "    \n",
        "    \n",
        "MAX_LEN = 100\n",
        "MAX_WORDS = 25000\n",
        "INDEX = 3\n",
        "\n",
        "tokenizer = Tokenizer(num_words = MAX_WORDS)\n",
        "tokenizer.fit_on_texts(X_train['description'])\n",
        "print('Found {} unique tokens.'.format(len(tokenizer.word_index)))\n",
        "\n",
        "sequences = tokenizer.texts_to_sequences(X_train['description'])\n",
        "X_train_main = pad_sequences(sequences, maxlen = MAX_LEN)\n",
        "print('\\nTrain Dataset:')\n",
        "print('\\tShape of the main input: {}'.format(X_train_main.shape))\n",
        "X_train_aux = X_train[X_train.columns[0 : -1 ]]\n",
        "print('\\tShape of the aux input: {}'.format(X_train_aux.shape))\n",
        "print('\\tShape of output: {}'.format(y_train.shape))\n",
        "\n",
        "sequences = tokenizer.texts_to_sequences(X_test['description'])\n",
        "X_test_main = pad_sequences(sequences, maxlen = MAX_LEN)\n",
        "#sequences = tokenizer.texts_to_sequences(X_test['description'])\n",
        "#np.vstack((X_test_main, pad_sequences(sequences, maxlen = MAX_LEN)))\n",
        "print('\\nValidation Dataset:')\n",
        "print('\\tShape of the main input: {}'.format(X_test_main.shape))\n",
        "X_test_aux = X_test[X_test.columns[0 : -1]]\n",
        "print('\\tShape of the aux input: {}'.format(X_test_aux.shape))\n",
        "print('\\tShape of output: {}'.format(y_test.shape))"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Found 3102 unique tokens.\n",
            "\n",
            "Train Dataset:\n",
            "\tShape of the main input: (707, 100)\n",
            "\tShape of the aux input: (707, 10)\n",
            "\tShape of output: (707, 1)\n",
            "\n",
            "Validation Dataset:\n",
            "\tShape of the main input: (177, 100)\n",
            "\tShape of the aux input: (177, 10)\n",
            "\tShape of output: (177, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "CjgpgEnhRWGx",
        "colab_type": "code",
        "outputId": "0184306f-2757-4fe6-bf0f-f5f0fc848637",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "EMBEDDING_DIM = 300\n",
        "\n",
        "embeddings_index = {}\n",
        "#with open('../datasets/{}/cbow_s{}.txt'.format(EMBEDDING, EMBEDDING_DIM), mode = 'r', encoding = 'UTF-8') as f:\n",
        "with open('./SBW-vectors-300-min5.txt', mode = 'r') as f:\n",
        "    for i, line in enumerate(f):\n",
        "        values = line.split(' ')\n",
        "        word = values[0]\n",
        "        try:\n",
        "            coefs = np.asarray(values[1 : ], dtype = 'float32')\n",
        "            embeddings_index[word] = coefs\n",
        "        except:\n",
        "            print('[ERROR] {}: {}'.format(i, values[1 : ]))\n",
        "    f.close()\n",
        "print('Found %s word vectors.' % len(embeddings_index))\n",
        "\n",
        "embedding_weights = np.zeros((MAX_WORDS, EMBEDDING_DIM))\n",
        "for word, i in tokenizer.word_index.items():\n",
        "    embedding_vector = embeddings_index.get(word)\n",
        "    if i < MAX_WORDS and embedding_vector is not None:\n",
        "            embedding_weights[i] = embedding_vector"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 1000654 word vectors.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "ZXQYvq7LZ35b",
        "colab_type": "code",
        "outputId": "a5f77d16-7631-4c95-c816-0d44d4b94e7d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "X_train_main.shape[1]"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "100"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "metadata": {
        "id": "6oOzwWBQRWGz",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#X_train_main"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "lFeO4FIrRWG2",
        "colab_type": "code",
        "outputId": "5537d621-3821-4aa2-c991-4d2d6af6ce29",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        }
      },
      "cell_type": "code",
      "source": [
        "LSTM_SIZE = 50\n",
        "DENSE1_SIZE = 256*3\n",
        "DENSE2_SIZE = 256*3\n",
        "LEARNING_RATE = 0.01\n",
        "RHO = 0.05\n",
        "EPSILON = None\n",
        "DECAY = 0.0\n",
        "\n",
        "main_input = Input(shape = (X_train_main.shape[1], ), dtype = 'int32', name = 'main_input')\n",
        "\n",
        "x = Embedding(MAX_WORDS, EMBEDDING_DIM, \n",
        "              input_length = MAX_LEN, \n",
        "              weights = [embedding_weights], \n",
        "              name = 'embedding',\n",
        "              trainable = True)(main_input)\n",
        "\n",
        "lstm_out = LSTM(LSTM_SIZE, name = 'lstm')(x)\n",
        "\n",
        "aux_input = Input(shape = (X_train_aux.shape[1], ), name = 'aux_input')\n",
        "\n",
        "x = concatenate([lstm_out, aux_input])\n",
        "x = Dense(DENSE1_SIZE, kernel_initializer = 'uniform', activation = 'relu', name = 'dense1')(x)\n",
        "x = Dropout(.3)(x)\n",
        "x = Dense(DENSE2_SIZE, kernel_initializer = 'uniform', activation='relu', name = 'dense2')(x)\n",
        "x = Dropout(.3)(x)\n",
        "#x = Dense(DENSE2_SIZE, kernel_initializer = 'uniform', activation='relu', name = 'dense3')(x)\n",
        "#x = Dropout(.3)(x)\n",
        "\n",
        "main_output = Dense(1, activation = 'linear', kernel_initializer = 'normal', name = 'main_output')(x)\n",
        "\n",
        "model = Model(inputs = [main_input, aux_input], outputs = [main_output])\n",
        "optimizer = RMSprop(lr = LEARNING_RATE, rho = RHO, epsilon = EPSILON, decay = DECAY)\n",
        "model.compile(optimizer = optimizer, loss = 'mse', metrics = ['mse'])\n",
        "model.summary()"
      ],
      "execution_count": 130,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "main_input (InputLayer)         (None, 100)          0                                            \n",
            "__________________________________________________________________________________________________\n",
            "embedding (Embedding)           (None, 100, 300)     7500000     main_input[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "lstm (LSTM)                     (None, 50)           70200       embedding[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "aux_input (InputLayer)          (None, 10)           0                                            \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_3 (Concatenate)     (None, 60)           0           lstm[0][0]                       \n",
            "                                                                 aux_input[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dense1 (Dense)                  (None, 768)          46848       concatenate_3[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "dropout_3 (Dropout)             (None, 768)          0           dense1[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "dense2 (Dense)                  (None, 768)          590592      dropout_3[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dropout_4 (Dropout)             (None, 768)          0           dense2[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "main_output (Dense)             (None, 1)            769         dropout_4[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 8,208,409\n",
            "Trainable params: 8,208,409\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "FfiONBuRRWG5",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#X_test.head()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "2H5GqqYHRWG6",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#y_test.head()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "UuwH7-ogRWG9",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import math\n",
        "#df_results = pd.read_csv(DF_RESULTS_FILENAME, index_col = 0)\n",
        "    \n",
        "PATIENCE = 1000\n",
        "EPOCHS = 1000\n",
        "BATCH_SIZE = int(math.ceil(y_train.shape[0] / 4))\n",
        "\n",
        "\n",
        "history = model.fit([X_train_main, X_train_aux], \n",
        "                    y_train,\n",
        "                    epochs = EPOCHS, \n",
        "                    batch_size = BATCH_SIZE,\n",
        "                    validation_data = ([X_test_main, X_test_aux], y_test),\n",
        "                    verbose = 0,\n",
        "                    callbacks = [EarlyStopping(monitor = 'mean_squared_error', \n",
        "                                               restore_best_weights = True,\n",
        "                                               patience = PATIENCE)])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "jPlPEN9brXgd",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "5l3TPb_ORWHH",
        "colab_type": "code",
        "outputId": "33cf8275-116b-4835-a8ca-ad9266d45eea",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 543
        }
      },
      "cell_type": "code",
      "source": [
        "train_mse = history.history['mean_squared_error']\n",
        "validation_mse = history.history['val_mean_squared_error']\n",
        "train_loss = history.history['loss']\n",
        "validation_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(train_mse))\n",
        "\n",
        "plt.plot(epochs, train_mse, 'b-', label = 'Training MSE')\n",
        "plt.plot(epochs, validation_mse, 'r-', label = 'Validation MSE')\n",
        "plt.title('Training and validation MSE')\n",
        "plt.legend()\n",
        "\n",
        "plt.figure()\n",
        "plt.plot(epochs, train_loss, 'b-', label = 'Training loss')\n",
        "plt.plot(epochs, validation_loss, 'r-', label = 'Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAEHCAYAAACzy817AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\noBIAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7ff9b296b850>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAEHCAYAAACzy817AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7ff9b42d1a10>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "3T8QnghuRWHA",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#y_test.head()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "XrKmUPC5RWHD",
        "colab_type": "code",
        "outputId": "18a85a58-70d5-4735-dda3-5fb950add823",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import mean_squared_error\n",
        "y_pred = model.predict([X_test_main, X_test_aux]) \n",
        "\n",
        "print(\"Number of predictions:\", len(y_pred))\n",
        "\n",
        "mse = mean_squared_error(y_test, y_pred)\n",
        "print(\"MSE: \", mse)\n",
        "rmse = math.sqrt(mse)\n",
        "print(\"RMSE: \", rmse) \n",
        "y_pred = np.array([x[0] for x in y_pred]) \n",
        "\n",
        "results = pd.DataFrame({\"y\":y_test[\"precio\"], 'pred': y_pred}, columns = [ \"y\", 'pred']) \n",
        "#result[\"y\"] = y_test[\"precio\"]\n",
        "results['error'] = abs(results['pred'].astype(float) - results['y'].astype(float))\n",
        "print(\"MAPE: \", 100 * np.mean(results['error'] / y_test[\"precio\"]))\n",
        "results.sort_values(by = 'error', ascending = True)\n",
        "\n",
        "results.head()"
      ],
      "execution_count": 146,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "('Number of predictions:', 177)\n",
            "('MSE: ', 771771035398991.1)\n",
            "('RMSE: ', 27780767.365193337)\n",
            "('MAPE: ', 16.151893091416152)\n"
          ],
          "name": "stdout"
        },
        {
          "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>y</th>\n",
              "      <th>pred</th>\n",
              "      <th>error</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>812</th>\n",
              "      <td>185000000.0</td>\n",
              "      <td>156539152.0</td>\n",
              "      <td>28460848.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>643</th>\n",
              "      <td>170000000.0</td>\n",
              "      <td>156092448.0</td>\n",
              "      <td>13907552.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>440</th>\n",
              "      <td>145000000.0</td>\n",
              "      <td>148781584.0</td>\n",
              "      <td>3781584.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>254</th>\n",
              "      <td>120000000.0</td>\n",
              "      <td>121290640.0</td>\n",
              "      <td>1290640.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>883</th>\n",
              "      <td>195000000.0</td>\n",
              "      <td>155405504.0</td>\n",
              "      <td>39594496.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               y         pred       error\n",
              "812  185000000.0  156539152.0  28460848.0\n",
              "643  170000000.0  156092448.0  13907552.0\n",
              "440  145000000.0  148781584.0   3781584.0\n",
              "254  120000000.0  121290640.0   1290640.0\n",
              "883  195000000.0  155405504.0  39594496.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 146
        }
      ]
    },
    {
      "metadata": {
        "id": "8untzcXqYgfM",
        "colab_type": "code",
        "outputId": "9fa13e11-7cb1-4496-a901-63d9e094a05b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        }
      },
      "cell_type": "code",
      "source": [
        "y_pred = model.predict([X_train_main, X_train_aux]) \n",
        "\n",
        "print(\"Number of predictions:\", len(y_pred))\n",
        "\n",
        "mse = mean_squared_error(y_train, y_pred)\n",
        "print(\"MSE: \", mse)\n",
        "rmse = math.sqrt(mse)\n",
        "print(\"RMSE: \", rmse) \n",
        "y_pred = np.array([x[0] for x in y_pred]) \n",
        "\n",
        "results = pd.DataFrame({\"y\":y_train[\"precio\"], 'pred': y_pred}, columns = [ \"y\", 'pred']) \n",
        "#result[\"y\"] = y_test[\"precio\"]\n",
        "results['error'] = abs(results['pred'].astype(float) - results['y'].astype(float))\n",
        "print(\"MAPE: \", 100 * np.mean(results['error'] / y_train[\"precio\"]))\n",
        "results.sort_values(by = 'error', ascending = True)\n",
        "\n",
        "results.head()"
      ],
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "('Number of predictions:', 707)\n",
            "('MSE: ', 630391799885849.8)\n",
            "('RMSE: ', 25107604.423477955)\n",
            "('MAPE: ', 14.837862742259489)\n"
          ],
          "name": "stdout"
        },
        {
          "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>y</th>\n",
              "      <th>pred</th>\n",
              "      <th>error</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>309</th>\n",
              "      <td>130000000.0</td>\n",
              "      <td>159560544.0</td>\n",
              "      <td>29560544.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>228</th>\n",
              "      <td>115000000.0</td>\n",
              "      <td>122409800.0</td>\n",
              "      <td>7409800.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110</th>\n",
              "      <td>91000000.0</td>\n",
              "      <td>115234984.0</td>\n",
              "      <td>24234984.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>564</th>\n",
              "      <td>160000000.0</td>\n",
              "      <td>170075264.0</td>\n",
              "      <td>10075264.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>121</th>\n",
              "      <td>94161000.0</td>\n",
              "      <td>113182464.0</td>\n",
              "      <td>19021464.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               y         pred       error\n",
              "309  130000000.0  159560544.0  29560544.0\n",
              "228  115000000.0  122409800.0   7409800.0\n",
              "110   91000000.0  115234984.0  24234984.0\n",
              "564  160000000.0  170075264.0  10075264.0\n",
              "121   94161000.0  113182464.0  19021464.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 147
        }
      ]
    },
    {
      "metadata": {
        "id": "NbFrxe6dRWHJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "fbbc9f26-2543-4b2c-ef21-c9222a3733c1"
      },
      "cell_type": "code",
      "source": [
        "import numpy\n",
        "import pandas\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense\n",
        "from keras.wrappers.scikit_learn import KerasRegressor\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.model_selection import KFold\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.pipeline import Pipeline\n",
        "\n",
        "LSTM_SIZE = 40\n",
        "DENSE1_SIZE = 256\n",
        "DENSE2_SIZE = 256\n",
        "LEARNING_RATE = 0.01\n",
        "RHO = 0.05\n",
        "EPSILON = None\n",
        "DECAY = 0.0\n",
        "\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Dense(DENSE1_SIZE, input_dim=X_train_aux.shape[1], kernel_initializer='normal', activation='relu'))\n",
        "model.add(Dense(1, kernel_initializer='normal'))\n",
        "# Compile model\n",
        "optimizer = RMSprop(lr = LEARNING_RATE, rho = RHO, epsilon = EPSILON, decay = DECAY)\n",
        "model.compile(loss='mean_squared_error', optimizer=optimizer)\n",
        "model.summary()"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_1 (Dense)              (None, 256)               2816      \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 257       \n",
            "=================================================================\n",
            "Total params: 3,073\n",
            "Trainable params: 3,073\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "c9MR9oUnxOI6",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import math\n",
        "#df_results = pd.read_csv(DF_RESULTS_FILENAME, index_col = 0)\n",
        "    \n",
        "PATIENCE = 10\n",
        "EPOCHS = 5000\n",
        "BATCH_SIZE = int(math.ceil(y_train.shape[0] / 4))\n",
        "\n",
        "\n",
        "history = model.fit(X_train_aux, \n",
        "                    y_train,\n",
        "                    epochs = EPOCHS, \n",
        "                    batch_size = BATCH_SIZE,\n",
        "                    validation_data = (X_test_aux, y_test),\n",
        "                    verbose = 0,\n",
        "                    callbacks = [EarlyStopping(monitor = 'mean_squared_error', \n",
        "                                               restore_best_weights = True,\n",
        "                                               patience = PATIENCE)])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "j2P_0gOvxtsw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 280
        },
        "outputId": "72ff30fe-9832-42ea-8547-696f5ddc6656"
      },
      "cell_type": "code",
      "source": [
        "\n",
        "train_loss = history.history['loss']\n",
        "validation_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(train_loss))\n",
        "\n",
        "\n",
        "plt.figure()\n",
        "plt.plot(epochs, train_loss, 'b-', label = 'Training loss')\n",
        "plt.plot(epochs, validation_loss, 'r-', label = 'Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 126,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAEHCAYAAACzy817AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nD5717RnQibCjAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7ff9b31e0210>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "3mj-3gRgyU2z",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        },
        "outputId": "e45e1bc1-0e7a-48bd-dc5e-3093f9412a0a"
      },
      "cell_type": "code",
      "source": [
        "y_pred = model.predict(X_test_aux) \n",
        "\n",
        "print(\"Number of predictions:\", len(y_pred))\n",
        "\n",
        "mse = mean_squared_error(y_test, y_pred)\n",
        "print(\"MSE: \", mse)\n",
        "rmse = math.sqrt(mse)\n",
        "print(\"RMSE: \", rmse) \n",
        "y_pred = np.array([x[0] for x in y_pred]) \n",
        "\n",
        "results = pd.DataFrame({\"y\":y_test[\"precio\"], 'pred': y_pred}, columns = [ \"y\", 'pred']) \n",
        "#result[\"y\"] = y_test[\"precio\"]\n",
        "results['error'] = abs(results['pred'].astype(float) - results['y'].astype(float))\n",
        "print(\"MAPE: \", 100 * np.mean(results['error'] / y_test[\"precio\"]))\n",
        "results.sort_values(by = 'error', ascending = True)\n",
        "\n",
        "results.head()"
      ],
      "execution_count": 127,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "('Number of predictions:', 177)\n",
            "('MSE: ', 730652252913385.2)\n",
            "('RMSE: ', 27030579.95887963)\n",
            "('MAPE: ', 16.487652478069283)\n"
          ],
          "name": "stdout"
        },
        {
          "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>y</th>\n",
              "      <th>pred</th>\n",
              "      <th>error</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>812</th>\n",
              "      <td>185000000.0</td>\n",
              "      <td>157910736.0</td>\n",
              "      <td>27089264.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>643</th>\n",
              "      <td>170000000.0</td>\n",
              "      <td>158896912.0</td>\n",
              "      <td>11103088.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>440</th>\n",
              "      <td>145000000.0</td>\n",
              "      <td>150570704.0</td>\n",
              "      <td>5570704.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>254</th>\n",
              "      <td>120000000.0</td>\n",
              "      <td>127545656.0</td>\n",
              "      <td>7545656.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>883</th>\n",
              "      <td>195000000.0</td>\n",
              "      <td>156948544.0</td>\n",
              "      <td>38051456.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               y         pred       error\n",
              "812  185000000.0  157910736.0  27089264.0\n",
              "643  170000000.0  158896912.0  11103088.0\n",
              "440  145000000.0  150570704.0   5570704.0\n",
              "254  120000000.0  127545656.0   7545656.0\n",
              "883  195000000.0  156948544.0  38051456.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 127
        }
      ]
    },
    {
      "metadata": {
        "id": "gTP41PPPyg3o",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        },
        "outputId": "3603e6dd-5536-4227-e1ab-dae1df3306fd"
      },
      "cell_type": "code",
      "source": [
        "y_pred = model.predict(X_train_aux) \n",
        "\n",
        "print(\"Number of predictions:\", len(y_pred))\n",
        "\n",
        "mse = mean_squared_error(y_train, y_pred)\n",
        "print(\"MSE: \", mse)\n",
        "rmse = math.sqrt(mse)\n",
        "print(\"RMSE: \", rmse) \n",
        "y_pred = np.array([x[0] for x in y_pred]) \n",
        "\n",
        "results = pd.DataFrame({\"y\":y_train[\"precio\"], 'pred': y_pred}, columns = [ \"y\", 'pred']) \n",
        "#result[\"y\"] = y_test[\"precio\"]\n",
        "results['error'] = abs(results['pred'].astype(float) - results['y'].astype(float))\n",
        "print(\"MAPE: \", 100 * np.mean(results['error'] / y_train[\"precio\"]))\n",
        "results.sort_values(by = 'error', ascending = True)\n",
        "\n",
        "results.head()"
      ],
      "execution_count": 128,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "('Number of predictions:', 707)\n",
            "('MSE: ', 656617166521781.9)\n",
            "('RMSE: ', 25624542.2695076)\n",
            "('MAPE: ', 15.645622795169778)\n"
          ],
          "name": "stdout"
        },
        {
          "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>y</th>\n",
              "      <th>pred</th>\n",
              "      <th>error</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>309</th>\n",
              "      <td>130000000.0</td>\n",
              "      <td>154139632.0</td>\n",
              "      <td>24139632.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>228</th>\n",
              "      <td>115000000.0</td>\n",
              "      <td>128497168.0</td>\n",
              "      <td>13497168.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110</th>\n",
              "      <td>91000000.0</td>\n",
              "      <td>122358728.0</td>\n",
              "      <td>31358728.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>564</th>\n",
              "      <td>160000000.0</td>\n",
              "      <td>165614656.0</td>\n",
              "      <td>5614656.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>121</th>\n",
              "      <td>94161000.0</td>\n",
              "      <td>118637672.0</td>\n",
              "      <td>24476672.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               y         pred       error\n",
              "309  130000000.0  154139632.0  24139632.0\n",
              "228  115000000.0  128497168.0  13497168.0\n",
              "110   91000000.0  122358728.0  31358728.0\n",
              "564  160000000.0  165614656.0   5614656.0\n",
              "121   94161000.0  118637672.0  24476672.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 128
        }
      ]
    },
    {
      "metadata": {
        "id": "cj_I74OiRWHQ",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "%notebook ./filename.ipynb\n",
        "!./dropbox_uploader.sh upload filename.ipynb /"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}