2-Funcionalidades-basicas-de-pandas/3_1_Save_Load_formats.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "3_1 Save Load formats.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/francomanca93/analisis-de-datos/blob/funcionalidades-basicas/2-Funcionalidades-basicas-de-pandas/3_1_Save_Load_formats.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qMCNKRVw_g4U",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 124
        },
        "outputId": "b8035a78-979f-4244-eb6c-cacc3f938318"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HKsNnjnR_kh7",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "35a8f243-e7fb-4796-bb67-cb340e718052"
      },
      "source": [
        "%cd '/content/drive/My Drive/Colab Notebooks'\n",
        "!ls"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/Colab Notebooks\n",
            "backup\tcurso_pandas  db  pd_1_0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZRZUwdPi_maN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "73c63489-e9a6-4a2a-faa6-e1158178cf5b"
      },
      "source": [
        "dir_pandas = '/content/drive/My Drive/Colab Notebooks/db'\n",
        "dir_pandas"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'/content/drive/My Drive/Colab Notebooks/db'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7Qq79oLT7gjT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F8x6PJFH7oPZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "bc5b7e55-9027-48d9-f17a-ab8e9e6b1f7f"
      },
      "source": [
        "pd.__version__"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'0.25.3'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Uo1Qsk5S7qha",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import seaborn as sns"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RBwK0NMA8G9q",
        "colab_type": "text"
      },
      "source": [
        "Utilizaremos:\n",
        "\n",
        "* [uuid](https://es.wikipedia.org/wiki/Identificador_%C3%BAnico_universal) : Identificador unico personal. uuid4 crea un identificador unico universar **aleatorio**. [Doc oficial](https://docs.python.org/3/library/uuid.html#uuid.UUID)\n",
        "\n",
        "* os: Módulo que provee d forma portable el uso de funcionalidades dependientes del sistema operativo. Ejemplo: os.stat(filename) devuelve información del archivo. [Doc oficial](https://docs.python.org/3.8/library/os.html).\n",
        "\n",
        "* time: Módulo que provee varias funciones relacionadas con el tiempo. [Doc oficial](https://docs.python.org/3/library/time.html).\n",
        "\n",
        "* tracemalloc: Es una herramienta de depuración para rastrear bloques de memoria asignador por Python. [Doc oficial](https://docs.python.org/3/library/tracemalloc.html)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cusgp0MpXUdD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from uuid import uuid4\n",
        "import os\n",
        "import time\n",
        "import tracemalloc"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q-b5aCX4UshZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def generate_dataset(n_rows, num_count, cat_count, max_nan=0.1, max_cat_size=100):\n",
        "    \"\"\"Randomly generate datasets with numerical and categorical features.\n",
        "    \n",
        "    The numerical features are taken from the normal distribution X ~ N(0, 1).\n",
        "    The categorical features are generated as random uuid4 strings with \n",
        "    cardinality C where 2 <= C <= max_cat_size.\n",
        "    \n",
        "    Also, a max_nan proportion of both numerical and categorical features is replaces\n",
        "    with NaN values.\n",
        "    \"\"\"\n",
        "    dataset, types = {}, {}\n",
        "    \n",
        "    def generate_categories():\n",
        "        \n",
        "        category_size = np.random.randint(2, max_cat_size)\n",
        "        return [str(uuid4()) for _ in range(category_size)]\n",
        "    \n",
        "    for col in range(num_count):\n",
        "        name = f'n{col}'\n",
        "        values = np.random.normal(0, 1, n_rows)\n",
        "        nan_cnt = np.random.randint(1, int(max_nan*n_rows))\n",
        "        index = np.random.choice(n_rows, nan_cnt, replace=False)\n",
        "        values[index] = np.nan\n",
        "        dataset[name] = values\n",
        "        types[name] = 'float32'\n",
        "        \n",
        "    for col in range(cat_count):\n",
        "        name = f'c{col}'\n",
        "        cats = generate_categories()\n",
        "        values = np.array(np.random.choice(cats, n_rows, replace=True), dtype=object)\n",
        "        nan_cnt = np.random.randint(1, int(max_nan*n_rows))\n",
        "        index = np.random.choice(n_rows, nan_cnt, replace=False)\n",
        "        values[index] = np.nan\n",
        "        dataset[name] = values\n",
        "        types[name] = 'object'\n",
        "    \n",
        "    return pd.DataFrame(dataset), types"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f8lwGbE_VO_I",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def size_of(filename, unit=1024**2):\n",
        "    return round(os.stat(filename).st_size / unit, 2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "duHIcZ2LXdVT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_save_load(df, fmt):\n",
        "    save = getattr(df, f'to_{fmt}')\n",
        "    load = getattr(pd, f'read_{fmt}')\n",
        "    return save, load"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5vIqJa0YXdXt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def benchmark(list_of_formats,\n",
        "              data_size=1000000,\n",
        "              n_num=15,\n",
        "              n_cat=15,\n",
        "              n_rounds=20,\n",
        "              as_category=False):\n",
        "    \"\"\"Runs dataset saving/loading benchamrk using formts from the list_of_formats.\n",
        "    \n",
        "    Each round a new random dataset is generated with data_size observations. \n",
        "    The measurements for each of the rounds are concatenated together and returned\n",
        "    as a single data frame.\n",
        "    \n",
        "    Parameters:\n",
        "        list_of_formats: A list of tuples in the format (<format_name>, [<params_dict>]). \n",
        "            The <format_name> should be one of the pandas supported formats.\n",
        "        data_size: A number of samples in the generated dataset.\n",
        "        n_num: A number of numerical columns in the generated dataset.\n",
        "        n_cat: A number of categorical columns in the generated dataset.\n",
        "        n_rounds: A number of randomly generated datasets to test the formats.\n",
        "        as_category: If True, then categorical columns will be converted into \n",
        "            pandas.Category type before saving.\n",
        "            \n",
        "    \"\"\"\n",
        "    runs = []\n",
        "    n = 1\n",
        "    cnt = 0\n",
        "\n",
        "    for j in range(n_rounds):\n",
        "        print('Benchmarking round # {}'.format(j))\n",
        "        #print('generating dataset...')\n",
        "        dataset, _ = generate_dataset(data_size, n_num, n_cat)\n",
        "        \n",
        "        if as_category:\n",
        "            #print('converting categorical columns into pandas.Category')\n",
        "            cat_cols = dataset.select_dtypes(include=object).columns\n",
        "            dataset[cat_cols] = dataset[cat_cols].fillna('none').astype('category')\n",
        "        \n",
        "        for case in list_of_formats:\n",
        "              fmt, params = case if len(case) == 2 else (case[0], {})\n",
        "            \n",
        "              print('testing format: {}'.format(fmt))\n",
        "              filename = 'random.{}'.format(fmt)\n",
        "              save, load = get_save_load(dataset, fmt)\n",
        "              results = {}\n",
        "              \n",
        "              results['format'] = fmt\n",
        "              results['filename'] = filename\n",
        "\n",
        "              tracemalloc.start()\n",
        "              start = time.time()\n",
        "              for i in range(n):\n",
        "                  save(dir_pandas + filename, **params)\n",
        "              end = (time.time()- start) / n\n",
        "              current, peak = tracemalloc.get_traced_memory()\n",
        "              tracemalloc.stop()\n",
        "\n",
        "              results['size_mb'] = size_of(dir_pandas + filename)\n",
        "              results['save_ram_mb'] = current / 10**6\n",
        "              results['save_ram_peak_mb'] = peak / 10**6\n",
        "              results['save_time'] = end\n",
        "              \n",
        "              tracemalloc.start()\n",
        "              start = time.time()\n",
        "              for i in range(n):\n",
        "                  _ = load(filename)\n",
        "              end = (time.time()- start) / n\n",
        "              current, peak = tracemalloc.get_traced_memory()\n",
        "              tracemalloc.stop()\n",
        "\n",
        "              results['load_ram_mb'] = current / 10**6\n",
        "              results['load_ram_peak_mb'] = peak / 10**6\n",
        "              results['load_time'] = end\n",
        "\n",
        "              results['run_no'] = j\n",
        "\n",
        "              cnt += 1\n",
        "              run = pd.DataFrame(results,[cnt])\n",
        "              runs.append(run)\n",
        "            \n",
        "    benchmark = pd.concat(runs, axis=0)\n",
        "    benchmark.reset_index(inplace=True, drop=True)\n",
        "    return benchmark\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mcKNtj8Q0EAk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "formats = [\n",
        "    ('json',),\n",
        "    ('csv', {'index': None}),\n",
        "    ('hdf', {'key': 'data', 'format': 'table'}),\n",
        "    ('pickle',),\n",
        "    #('msgpack',),\n",
        "    ('feather',),\n",
        "    ('parquet', {'engine': 'pyarrow'})\n",
        "]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f7NJGrfkKufk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "sns.set(rc={'figure.figsize':(11.7,8.27)})"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CEo7zdFEO28R",
        "colab_type": "text"
      },
      "source": [
        "# Strings"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jc-23ktz0bK4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 139
        },
        "outputId": "02b6a17f-34de-4cf4-d2fe-30e5a5d16c5e"
      },
      "source": [
        "results_io = benchmark(formats,\n",
        "              data_size=100000,\n",
        "              n_num=15,\n",
        "              n_cat=15,\n",
        "              n_rounds=1,\n",
        "              as_category=False)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Benchmarking round # 0\n",
            "testing format: json\n",
            "testing format: csv\n",
            "testing format: hdf\n",
            "testing format: pickle\n",
            "testing format: feather\n",
            "testing format: parquet\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SYiwEhEkEchM",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "outputId": "79592d22-9653-4d62-f315-9982d9e821fc"
      },
      "source": [
        "avg = results_io.groupby('format').mean().reset_index().drop(columns=['run_no'])\n",
        "avg"
      ],
      "execution_count": null,
      "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>format</th>\n",
              "      <th>size_mb</th>\n",
              "      <th>save_ram_mb</th>\n",
              "      <th>save_ram_peak_mb</th>\n",
              "      <th>save_time</th>\n",
              "      <th>load_ram_mb</th>\n",
              "      <th>load_ram_peak_mb</th>\n",
              "      <th>load_time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>csv</td>\n",
              "      <td>76.58</td>\n",
              "      <td>0.020431</td>\n",
              "      <td>13.496845</td>\n",
              "      <td>6.708381</td>\n",
              "      <td>24.283428</td>\n",
              "      <td>48.300520</td>\n",
              "      <td>1.245796</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>feather</td>\n",
              "      <td>65.95</td>\n",
              "      <td>0.061709</td>\n",
              "      <td>0.103927</td>\n",
              "      <td>0.408491</td>\n",
              "      <td>24.073244</td>\n",
              "      <td>24.081095</td>\n",
              "      <td>0.243123</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>hdf</td>\n",
              "      <td>66.90</td>\n",
              "      <td>12.029854</td>\n",
              "      <td>199.646839</td>\n",
              "      <td>2.340985</td>\n",
              "      <td>335.242716</td>\n",
              "      <td>378.933460</td>\n",
              "      <td>1.941241</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>json</td>\n",
              "      <td>94.00</td>\n",
              "      <td>0.814209</td>\n",
              "      <td>233.581957</td>\n",
              "      <td>1.431729</td>\n",
              "      <td>144.733795</td>\n",
              "      <td>1065.407778</td>\n",
              "      <td>4.480983</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>parquet</td>\n",
              "      <td>15.04</td>\n",
              "      <td>0.021925</td>\n",
              "      <td>0.885827</td>\n",
              "      <td>0.346414</td>\n",
              "      <td>24.094453</td>\n",
              "      <td>24.110855</td>\n",
              "      <td>0.195544</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>pickle</td>\n",
              "      <td>64.68</td>\n",
              "      <td>0.013195</td>\n",
              "      <td>215.861380</td>\n",
              "      <td>0.749980</td>\n",
              "      <td>146.032663</td>\n",
              "      <td>176.244113</td>\n",
              "      <td>0.970435</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    format  size_mb  save_ram_mb  ...  load_ram_mb  load_ram_peak_mb  load_time\n",
              "0      csv    76.58     0.020431  ...    24.283428         48.300520   1.245796\n",
              "1  feather    65.95     0.061709  ...    24.073244         24.081095   0.243123\n",
              "2      hdf    66.90    12.029854  ...   335.242716        378.933460   1.941241\n",
              "3     json    94.00     0.814209  ...   144.733795       1065.407778   4.480983\n",
              "4  parquet    15.04     0.021925  ...    24.094453         24.110855   0.195544\n",
              "5   pickle    64.68     0.013195  ...   146.032663        176.244113   0.970435\n",
              "\n",
              "[6 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-OkZEM27FCUH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "outputId": "c750ea37-1922-4fd6-9390-753773a7820c"
      },
      "source": [
        "io_speed = avg[~avg['format'].isin([])][['format', 'save_time', 'load_time']].melt(id_vars='format')\n",
        "ax = sns.barplot(x='format', y='value', hue='variable', data=io_speed, palette=\"magma\")\n",
        "_ = ax.set_xlabel('File Format')\n",
        "_ = ax.set_ylabel('Seconds')\n",
        "_ = ax.set_title('Time to Save/Load a Data Frame')\n",
        "ax.grid(True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAArwAAAIECAYAAAD7I/1bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAQaQReAAAARNo7VmX5lPdyU1IAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 842.4x595.44 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SHy28H1HFTdm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "outputId": "a72cabab-c508-4b86-9c3d-643ae9e4b47a"
      },
      "source": [
        "mem_size = avg[~avg['format'].isin([])][['format', 'save_ram_peak_mb', 'load_ram_peak_mb']].melt(id_vars='format')\n",
        "ax = sns.barplot(x='format', y='value', hue='variable', data=mem_size, palette=\"magma\")\n",
        "_ = ax.set_xlabel('File Format')\n",
        "_ = ax.set_ylabel('Memory, Mb')\n",
        "ax.set_title('Memory Peaks Consumption Growth When Saving/Loading')\n",
        "ax.grid(True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtEAAAIECAYAAADfOnKbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAACATf8HqDxdU57R9o8AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 842.4x595.44 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "udJtzeEFFVWY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "outputId": "81089e6f-1237-42bf-8252-6f44e171885a"
      },
      "source": [
        "ax = sns.barplot(x='format', y='size_mb', data=avg, palette=\"magma\")\n",
        "_ = ax.set_ylabel('Memory, Mb')\n",
        "_ = ax.set_xlabel('File Format')\n",
        "_ = ax.set_title('Serialized File Size')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIECAYAAAD8YSNOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 842.4x595.44 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2UVOxdGQO0BD",
        "colab_type": "text"
      },
      "source": [
        "# Categorias"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9HnDk9el9y7W",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 139
        },
        "outputId": "f99758bd-d84e-4889-9af7-3bef1032991b"
      },
      "source": [
        "results_io_cat = benchmark(formats,\n",
        "              data_size=100000,\n",
        "              n_num=15,\n",
        "              n_cat=15,\n",
        "              n_rounds=1,\n",
        "              as_category=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Benchmarking round # 0\n",
            "testing format: json\n",
            "testing format: csv\n",
            "testing format: hdf\n",
            "testing format: pickle\n",
            "testing format: feather\n",
            "testing format: parquet\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "41o6H1ZxOTzU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "outputId": "9d2153d2-70c3-4ee4-ab88-ed2c4c2cdf83"
      },
      "source": [
        "avg = results_io_cat.groupby('format').mean().reset_index().drop(columns=['run_no'])\n",
        "avg"
      ],
      "execution_count": null,
      "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>format</th>\n",
              "      <th>size_mb</th>\n",
              "      <th>save_ram_mb</th>\n",
              "      <th>save_ram_peak_mb</th>\n",
              "      <th>save_time</th>\n",
              "      <th>load_ram_mb</th>\n",
              "      <th>load_ram_peak_mb</th>\n",
              "      <th>load_time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>csv</td>\n",
              "      <td>77.38</td>\n",
              "      <td>0.028869</td>\n",
              "      <td>13.507343</td>\n",
              "      <td>6.529927</td>\n",
              "      <td>24.310116</td>\n",
              "      <td>48.327208</td>\n",
              "      <td>1.260878</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>feather</td>\n",
              "      <td>13.09</td>\n",
              "      <td>0.068804</td>\n",
              "      <td>0.303033</td>\n",
              "      <td>0.083436</td>\n",
              "      <td>12.113894</td>\n",
              "      <td>12.121842</td>\n",
              "      <td>0.041463</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>hdf</td>\n",
              "      <td>18.25</td>\n",
              "      <td>0.495662</td>\n",
              "      <td>32.310204</td>\n",
              "      <td>1.071665</td>\n",
              "      <td>29.672178</td>\n",
              "      <td>43.564825</td>\n",
              "      <td>0.465293</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>json</td>\n",
              "      <td>94.57</td>\n",
              "      <td>0.824381</td>\n",
              "      <td>234.204345</td>\n",
              "      <td>1.355046</td>\n",
              "      <td>149.845604</td>\n",
              "      <td>1074.165543</td>\n",
              "      <td>4.107476</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>parquet</td>\n",
              "      <td>15.09</td>\n",
              "      <td>0.024861</td>\n",
              "      <td>0.893393</td>\n",
              "      <td>0.313584</td>\n",
              "      <td>24.098545</td>\n",
              "      <td>24.115486</td>\n",
              "      <td>0.141382</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>pickle</td>\n",
              "      <td>12.91</td>\n",
              "      <td>0.020097</td>\n",
              "      <td>34.067053</td>\n",
              "      <td>0.054402</td>\n",
              "      <td>13.608012</td>\n",
              "      <td>25.693748</td>\n",
              "      <td>0.050316</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    format  size_mb  save_ram_mb  ...  load_ram_mb  load_ram_peak_mb  load_time\n",
              "0      csv    77.38     0.028869  ...    24.310116         48.327208   1.260878\n",
              "1  feather    13.09     0.068804  ...    12.113894         12.121842   0.041463\n",
              "2      hdf    18.25     0.495662  ...    29.672178         43.564825   0.465293\n",
              "3     json    94.57     0.824381  ...   149.845604       1074.165543   4.107476\n",
              "4  parquet    15.09     0.024861  ...    24.098545         24.115486   0.141382\n",
              "5   pickle    12.91     0.020097  ...    13.608012         25.693748   0.050316\n",
              "\n",
              "[6 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BTzjl9NrOV_J",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 548
        },
        "outputId": "d48892cc-1155-46e8-ead9-627289e5c9f1"
      },
      "source": [
        "io_speed = avg[~avg['format'].isin(['json','csv'])][['format', 'save_time', 'load_time']].melt(id_vars='format')\n",
        "ax = sns.barplot(x='format', y='value', hue='variable', data=io_speed, palette=\"magma\")\n",
        "_ = ax.set_xlabel('File Format')\n",
        "_ = ax.set_ylabel('Seconds')\n",
        "_ = ax.set_title('Time to Save/Load a Data Frame\\n(categories)')\n",
        "ax.grid(True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAscAAAITCAYAAADrWbEXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRBZMjomIiIiILJgcExERERFZ/A989/7FymCMdQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 842.4x595.44 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DvesOpYfOWH_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 548
        },
        "outputId": "cb41a41b-4ef7-45a7-8714-f92dc4d84b61"
      },
      "source": [
        "mem_size = avg[~avg['format'].isin(['json','csv'])][['format', 'save_ram_peak_mb', 'load_ram_peak_mb']].melt(id_vars='format')\n",
        "ax = sns.barplot(x='format', y='value', hue='variable', data=mem_size, palette=\"magma\")\n",
        "_ = ax.set_xlabel('File Format')\n",
        "_ = ax.set_ylabel('Memory, Mb')\n",
        "ax.set_title('Memory Consumption Growth When Saving/Loading\\n(categories)')\n",
        "ax.grid(True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAITCAYAAADishFtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ncFSeplIAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 842.4x595.44 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bFgcgDF1IprQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 548
        },
        "outputId": "03218245-a126-4111-8ceb-93fb89e7c974"
      },
      "source": [
        "ax = sns.barplot(x='format', y='size_mb', data=avg, palette=\"magma\")\n",
        "_ = ax.set_ylabel('Memory, Mb')\n",
        "_ = ax.set_xlabel('File Format')\n",
        "_ = ax.set_title('Serialized File Size\\n(categories)')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAITCAYAAADishFtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 842.4x595.44 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VcUufqkmOnL2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vbZFR4yqQvf6",
        "colab_type": "text"
      },
      "source": [
        "Curso: Análisis de datos con Pandas\t\t\t\t\n",
        "Clase: I/O Tools\t\t\t\t\n",
        "Nivel: Intermedio\t\t\t\t\n",
        "Keywords: Load - Save - Time - Size - DataFrame - Input - Output - Tools\t\t\t\t\n",
        "\t\t\t\t\n",
        "Abstract:\t\t\t\t\n",
        "En esta clase te enseñaremos sobre diversos formatos de almacenamiento de datos y cómo elegir el formato eficiente que se adapte a tus necesidades.\t\t\t\t\n",
        "\t\t\t\t\n",
        "\t\t\t\t\n",
        "\t\t\t\t\n",
        "\t\t\t\t"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iQtn6gDoQwym",
        "colab_type": "text"
      },
      "source": [
        "Conclusiones:\n",
        "\n",
        "CSV y formatos String : Son simples, requieren alto costo computacional y algo lentos.\n",
        "\n",
        "HDF : Gran soporte, adecuado para grandes cantidades de datos, rápido a costo de alto costo computacional.\n",
        "\n",
        "Parquet : Puede igualar a hdf e inclusive trabajar por chunks y en paralelo.\n",
        "\n",
        "Pickle : Es práctico pero lento con grandes cantidades de datos."
      ]
    }
  ]
}