170100016_A2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "celltoolbar": "Raw Cell Format",
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.1"
    },
    "colab": {
      "name": "170100016_A2.ipynb",
      "provenance": []
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fGdfJb1wUgPM",
        "colab_type": "text"
      },
      "source": [
        "## Assignment 2  ME 781  (Data Scale)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RjbEgfUbUgPO",
        "colab_type": "text"
      },
      "source": [
        "### Perform the following tasks in sequence:\n",
        "\n",
        "1. Create a list of 5 open data sets available for machine learning along with their links for each of the three Data Types (Discrete data Type, Sequential (temporal) data Type and Spatial data Type)\n",
        "\n",
        "\n",
        "2. Write a python code to input a dataset (csv file) of Discrete data type, automatically identify and print the\n",
        "\n",
        "    a. Number of attributes and number of entries\n",
        "    \n",
        "    b. Data scale of each attribute\n",
        "    \n",
        "    c. Data value of each attribute\n",
        "\n",
        "\n",
        "3.      Write a python code to perform statistics on each data column based on its data scale.\n",
        "\n",
        "\n",
        "4.      Can we use a box plot for visualizing ordinal data? (Yes/No). Explain why. \n",
        "\n",
        "![MicrosoftTeams-image2.png](attachment:MicrosoftTeams-image2.png)\n",
        "\n",
        "\n",
        "5.      Write a python code to visualize data in the csv file (a generalized method depending upon data scale) \n",
        "\n",
        "    Assignment instructions:\n",
        "    1. Naming convention: Please name your colab files as rollnumber_A2.ipynb. \n",
        "    2. Learn how to use markdown to add texts in colab to make your code interactive and understandable. \n",
        "    3. After completion of your assignment download the file as .ipynb and submit itusing Teams assignment tab."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UmrqJcAiUgPQ",
        "colab_type": "text"
      },
      "source": [
        "  ###  We have a software to check plagiarism in code, if anyone found copying the code, strict action will be taken against such students and can also be given FR grade directly.\n",
        "    \n",
        "  We  expect  you  to  upload  colab  notebooks  when  you  are  required  to  turn  your  programming assignments.  Please  make  sure  that  you  provide  enough  comments  in  the  code  for  the TAs to understand \n",
        "       "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "43T0XSKzUgPR",
        "colab_type": "text"
      },
      "source": [
        "### Libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cjB_iBdcUgPS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns "
      ],
      "execution_count": 173,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "udovwPBBUgPX",
        "colab_type": "text"
      },
      "source": [
        "### 1. List of 5 open data sets"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e7zY1HxoUgPY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "## Paste the URLs of open datasets here and add comments describing the dataset\n",
        "url1 = \"https://raw.githubusercontent.com/fivethirtyeight/data/master/airline-safety/airline-safety.csv\"\n",
        "url2 = \"https://raw.githubusercontent.com/fivethirtyeight/data/master/fight-songs/fight-songs.csv\"\n",
        "url3 = \"https://raw.githubusercontent.com/fivethirtyeight/data/master/world-cup-predictions/wc-20140609-140000.csv\"\n",
        "url4 = \"https://raw.githubusercontent.com/fivethirtyeight/data/master/births/US_births_1994-2003_CDC_NCHS.csv\"\n",
        "url5 = \"https://raw.githubusercontent.com/fivethirtyeight/data/master/librarians/librarians-by-msa.csv\"\n",
        "\n",
        "# github link for url1 = \"https://github.com/fivethirtyeight/data/blob/master/airline-safety/airline-safety.csv\"\n",
        "# github link for url2 = \"https://github.com/fivethirtyeight/data/blob/master/fight-songs/fight-songs.csv\"\n",
        "# github link for url3 = \"https://github.com/fivethirtyeight/data/blob/master/world-cup-predictions/wc-20140609-140000.csv\"\n",
        "# github link for url4 = \"https://github.com/fivethirtyeight/data/blob/master/births/US_births_1994-2003_CDC_NCHS.csv\"\n",
        "# github link for url5 = \"https://github.com/fivethirtyeight/data/blob/master/librarians/librarians-by-msa.csv\""
      ],
      "execution_count": 174,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-tNbb-vRUgPd",
        "colab_type": "text"
      },
      "source": [
        "### 2.   Write a python code to input a dataset (csv file) of Discrete datatype, automatically identify and print the\n",
        "\n",
        "    a. Number of attributes and number of data points\n",
        "    \n",
        "    b. Data scale of each attribute\n",
        "    \n",
        "    c. Data value of each attribute"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7oJ-Z4AhUgPe",
        "colab_type": "text"
      },
      "source": [
        "#### Note: The number of categories for categorical (nominal,ordinal) data scale for the evaluating datasets is not greater than 8."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T-1wDF_OUgPf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "3da4c232-271f-4491-c5df-002f6b62616e"
      },
      "source": [
        "def data_description(filename):\n",
        "    ## Write a well commented function that inputs dataset file location (.csv file) and outputs the following\n",
        "    ## 1. data_scale (of length = number of attributes)\n",
        "    ## 2. data_value (of length = number of attributes)\n",
        "    ## 3. imported dataset for later use\n",
        "    ## It should also print those details for each attribute(Output must be similar to what is mentioned below)\n",
        "\n",
        "    # the pd.read_csv() function takes a filepath as its first input and returns a Pandas Dataframe\n",
        "    df = pd.read_csv(filename)      # the filename should be in the string format with the url of the csv file\n",
        "    print(\"This dataset contains {} entries, each with {} attributes\".format(len(df),df.shape[1]))\n",
        "    print()\n",
        "    data_scale = []      # list to store the data scale of each attribute\n",
        "    data_value = []      # list to store the data value of each attribute\n",
        "    data_types = df.dtypes\n",
        "    for i in range(df.shape[1]):\n",
        "      if data_types[i] == 'object':\n",
        "        data_scale.append(\"nominal\")      \n",
        "        data_value.append(\"discrete\")    # Data value of categorical data must be discrete\n",
        "      elif data_types[i] == 'float':\n",
        "        # if the number of categories(unique values) are less than or equal to 8, then the data scale must be nominal or ordinal\n",
        "        if df.iloc[:,i].nunique(dropna = True) <= 8:   \n",
        "          try:\n",
        "            df.iloc[0,i] > df.iloc[1,i]    # checking if the greater than operation is valid or not\n",
        "            isValid = True         \n",
        "          except:\n",
        "            isValid = False \n",
        "          if isValid:\n",
        "            data_scale.append(\"ordinal\")   # if the greater than operation is valid, then it must be ordinal data scale\n",
        "          else:\n",
        "            data_scale.append(\"nominal\")\n",
        "        else:\n",
        "          # if the number of categories(unique values) are less than or equal to 8, then the data scale must be interval or ratio\n",
        "          data_scale.append(\"ratio\")\n",
        "        data_value.append(\"continuous\")    # if the data is floating type, then it must be continuous data \n",
        "      else:\n",
        "        if df.iloc[:,i].nunique(dropna = True) <= 8:   \n",
        "          try:\n",
        "            df.iloc[0,i] > df.iloc[1,i]    # checking if the greater than operation is valid or not\n",
        "            isValid = True         \n",
        "          except:\n",
        "            isValid = False \n",
        "          if isValid:\n",
        "            data_scale.append(\"ordinal\")   # if the greater than operation is valid, then it must be ordinal data scale\n",
        "          else:\n",
        "            data_scale.append(\"nominal\")\n",
        "        else:\n",
        "          # if the number of categories(unique values) are less than or equal to 8, then the data scale must be interval or ratio\n",
        "          data_scale.append(\"ratio\")\n",
        "        # if the range of values is less than or equal to 100, then the data values are discrete \n",
        "        if np.max(df.iloc[:,i]) - np.min(df.iloc[:,i]) <= 100:\n",
        "          data_value.append(\"discrete\")      \n",
        "        else:    # if the range of values is greater to 100, then the data values are continuous \n",
        "          data_value.append(\"continuous\")\n",
        "      print(\"{} {} : {}, {}\".format(i,df.columns[i],data_scale[i],data_value[i]))\n",
        "    # return data_scale, data_value, df\n",
        "data_description(url1)\n",
        "print()\n",
        "data_description(url2)\n",
        "print()\n",
        "data_description(url3)\n",
        "print()\n",
        "data_description(url4)\n",
        "print()\n",
        "data_description(url5)"
      ],
      "execution_count": 175,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "This dataset contains 56 entries, each with 8 attributes\n",
            "\n",
            "0 airline : nominal, discrete\n",
            "1 avail_seat_km_per_week : ratio, continuous\n",
            "2 incidents_85_99 : ratio, discrete\n",
            "3 fatal_accidents_85_99 : ratio, discrete\n",
            "4 fatalities_85_99 : ratio, continuous\n",
            "5 incidents_00_14 : ratio, discrete\n",
            "6 fatal_accidents_00_14 : ordinal, discrete\n",
            "7 fatalities_00_14 : ratio, continuous\n",
            "\n",
            "This dataset contains 65 entries, each with 23 attributes\n",
            "\n",
            "0 school : nominal, discrete\n",
            "1 conference : nominal, discrete\n",
            "2 song_name : nominal, discrete\n",
            "3 writers : nominal, discrete\n",
            "4 year : nominal, discrete\n",
            "5 student_writer : nominal, discrete\n",
            "6 official_song : nominal, discrete\n",
            "7 contest : nominal, discrete\n",
            "8 bpm : ratio, continuous\n",
            "9 sec_duration : ratio, continuous\n",
            "10 fight : nominal, discrete\n",
            "11 number_fights : ratio, discrete\n",
            "12 victory : nominal, discrete\n",
            "13 win_won : nominal, discrete\n",
            "14 victory_win_won : nominal, discrete\n",
            "15 rah : nominal, discrete\n",
            "16 nonsense : nominal, discrete\n",
            "17 colors : nominal, discrete\n",
            "18 men : nominal, discrete\n",
            "19 opponents : nominal, discrete\n",
            "20 spelling : nominal, discrete\n",
            "21 trope_count : ordinal, discrete\n",
            "22 spotify_id : nominal, discrete\n",
            "\n",
            "This dataset contains 32 entries, each with 12 attributes\n",
            "\n",
            "0 country : nominal, discrete\n",
            "1 country_id : nominal, discrete\n",
            "2 group : nominal, discrete\n",
            "3 spi : ratio, continuous\n",
            "4 spi_offense : ratio, continuous\n",
            "5 spi_defense : ratio, continuous\n",
            "6 win_group : ratio, continuous\n",
            "7 sixteen : ratio, continuous\n",
            "8 quarter : ratio, continuous\n",
            "9 semi : ratio, continuous\n",
            "10 cup : ratio, continuous\n",
            "11 win : ratio, continuous\n",
            "\n",
            "This dataset contains 3652 entries, each with 5 attributes\n",
            "\n",
            "0 year : ratio, discrete\n",
            "1 month : ratio, discrete\n",
            "2 date_of_month : ratio, discrete\n",
            "3 day_of_week : ordinal, discrete\n",
            "4 births : ratio, continuous\n",
            "\n",
            "This dataset contains 373 entries, each with 6 attributes\n",
            "\n",
            "0 prim_state : nominal, discrete\n",
            "1 area_name : nominal, discrete\n",
            "2 tot_emp : nominal, discrete\n",
            "3 emp_prse : nominal, discrete\n",
            "4 jobs_1000 : nominal, discrete\n",
            "5 loc_quotient : nominal, discrete\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ylulvGNVUgPj",
        "colab_type": "text"
      },
      "source": [
        "    Output must similar to this:\n",
        "    \n",
        "    This dataset contains 649 entries, each with 6 attributes\n",
        "\n",
        "    0 gender : nominal, discrete\n",
        "    1 age : ratio, discrete\n",
        "    2 health_rating : ordinal, discrete\n",
        "    3 StudentIQ : ratio, discrete\n",
        "    4 no_of_courses : ratio, discrete\n",
        "    5 Marks : ratio, continuous\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "erSulOVhUgPk",
        "colab_type": "text"
      },
      "source": [
        "### 3. Write a python code to perform statistics on each data column based on its data scale."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wVRActKYUgPk",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        },
        "outputId": "3b2f0156-94a9-4c11-b007-c0c8e9371331"
      },
      "source": [
        "def stats(filename):\n",
        "    ## Write a well commented function that calls the function data_description and prints\n",
        "    ## relevant statistics based on data scale.\n",
        "    ## Hint: Explore describe()\n",
        "    # the pd.read_csv() function takes a filepath as its first input and returns a Pandas Dataframe\n",
        "    df = pd.read_csv(filename)      # the filename should be in the string format with the url of the csv file\n",
        "    print(df.describe())\n",
        "    return\n",
        "stats(url2)"
      ],
      "execution_count": 183,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              bpm  sec_duration  number_fights  trope_count\n",
            "count   65.000000     65.000000      65.000000    65.000000\n",
            "mean   128.800000     71.907692       2.846154     3.615385\n",
            "std     33.152677     25.056014       3.231828     1.674182\n",
            "min     65.000000     27.000000       0.000000     0.000000\n",
            "25%     90.000000     58.000000       0.000000     3.000000\n",
            "50%    140.000000     67.000000       2.000000     4.000000\n",
            "75%    151.000000     85.000000       5.000000     5.000000\n",
            "max    180.000000    172.000000      17.000000     8.000000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EdlC8EutUgPo",
        "colab_type": "text"
      },
      "source": [
        "    Output must be some what similar to this:\n",
        "\n",
        "    gender\n",
        "    mode F\n",
        "    F    383\n",
        "    M    266\n",
        "    Name: gender, dtype: int64 \n",
        "\n",
        "    age\n",
        "    mode      17.000000\n",
        "    mean      16.744222\n",
        "    std        1.218138\n",
        "    min       15.000000\n",
        "    25%       16.000000\n",
        "    median    17.000000\n",
        "    75%       18.000000\n",
        "    max       22.000000\n",
        "    Name: age, dtype: float64 \n",
        "\n",
        "    health_rating\n",
        "    mode   5\n",
        "    median 4\n",
        "    1     90\n",
        "    2     78\n",
        "    3    124\n",
        "    4    108\n",
        "    5    249\n",
        "    Name: health_rating, dtype: int64 \n",
        "\n",
        "    StudentIQ\n",
        "    mode     104.000000\n",
        "    mean     100.229584\n",
        "    std       10.256494\n",
        "    min       75.000000\n",
        "    25%       93.000000\n",
        "    median   101.000000\n",
        "    75%      107.000000\n",
        "    max      131.000000\n",
        "    Name: StudentIQ, dtype: float64 \n",
        "\n",
        "    no_of_courses\n",
        "    mode      17.000000\n",
        "    count    649.000000\n",
        "    mean      10.596302\n",
        "    std        5.714697\n",
        "    min        1.000000\n",
        "    25%        5.000000\n",
        "    median    11.000000\n",
        "    75%       16.000000\n",
        "    max       20.000000\n",
        "    Name: no_of_courses, dtype: float64 \n",
        "\n",
        "    Marks\n",
        "    mode      52.810000\n",
        "    count    649.000000\n",
        "    mean      75.137257\n",
        "    std       14.559722\n",
        "    min       50.120000\n",
        "    25%       62.560000\n",
        "    median    75.120000\n",
        "    75%       87.670000\n",
        "    max       99.860000\n",
        "    Name: Marks, dtype: float64 "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qKaO9GzzUgPp",
        "colab_type": "text"
      },
      "source": [
        "### 4. Can we use a box plot for visualizing ordinal data ? (Yes/No). Explain why.\n",
        "![MicrosoftTeams-image2.png](attachment:MicrosoftTeams-image2.png)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zxogu01qUgPq",
        "colab_type": "text"
      },
      "source": [
        "### Write your answer here. \n",
        "    NO because ordinal data have very few categories and so they doesn't have a distribution. Hence bar graphs are used to plot them."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7PZ6zZ0JUgPr",
        "colab_type": "text"
      },
      "source": [
        "###   5.  Write a python code to visualize the data (a generalized method depending on data scale) "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mEs2XHOdUgPs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "3dfc4e8b-398d-411e-fb33-c673602575a4"
      },
      "source": [
        "def visualize_data(filename):\n",
        "    ## Write a well commented function that can call any of the functions from above \n",
        "    ## to plot relevant plots\n",
        "    ## Hint: Use sns.distplot, sns.countplot. Make a new figure for each plot (plt.figure(i))\n",
        "\n",
        "    # the pd.read_csv() function takes a filepath as its first input and returns a Pandas Dataframe\n",
        "    df = pd.read_csv(filename)      # the filename should be in the string format with the url of the csv file\n",
        "    data_types = df.dtypes\n",
        "    columns = df.columns\n",
        "    for i in range(df.shape[1]):\n",
        "      plt.figure(i)\n",
        "      sns.set(style=\"darkgrid\")\n",
        "      if data_types[i] == 'object':\n",
        "        if df.iloc[:,i].nunique() <= 8:\n",
        "          ax = sns.countplot(x=pd.Series(df.iloc[:,i], name = columns[i]),data=df)\n",
        "      else:\n",
        "        if df.iloc[:,i].nunique() <= 8:\n",
        "          ax = sns.countplot(x=pd.Series(df.iloc[:,i], name = columns[i]),data=df)\n",
        "        else:\n",
        "          ax = sns.distplot(pd.Series(df.iloc[:,i], name = columns[i]))\n",
        "    return\n",
        "visualize_data(url1)   # change the input to url1, url2, url3, url4 to test the data for visualisation"
      ],
      "execution_count": 252,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEMCAYAAAA1VZrrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de1xU953/8dfcuA4gMw7DcBEVFScixluMazXVaiAVgzU/i2vMI/mZmE1NY2PaJHTb1ZjG7Wp/zW7ianrJZddtt3FNGo1ojNpcvORiTEw0IooIojCAMnKX28z5/UGclaDMgAMzA5/n4+FDmPOd73zOEXnP+Z45369KURQFIYQQogtqXxcghBDC/0lYCCGEcEvCQgghhFsSFkIIIdySsBBCCOGWhIUQQgi3JCyEEEK4pfV1Ad5w+XIDTmfXt4sYjXqqqur7qCLvCtTape6+JXX3vUCt3WjUd/s5/SIsnE7FbVhcbReoArV2qbtvSd19L5Br7w4ZhhJCCOGWhIUQQgi3JCyEEEK4JWEhhBDCLQkLIYQQbklYCCGEcEvCQgghhFv94j6L/q6usYWG5jav9Res06KVtwlCiG6QsAgAV5ra+Oxkhdf6m2w1ow2Wf3ohhOfk/aUQQgi3JCyEEEK4JWEhhBDCLQkLIYQQbklYCCGEcEvCQgghhFsSFkIIIdzyKCyKiorIzs4mPT2d7OxsiouLO7XZuHEjc+fOZd68eSxYsIADBw64tl25coXHH3+cOXPmkJGRwfvvv+/RNiGEEP7BozuzVq9ezeLFi8nKymL79u2sWrWKzZs3d2iTlpbG0qVLCQ0NJT8/nyVLlnDw4EFCQkJ45ZVX0Ov17N27l+LiYu6991727NlDeHh4l9uEEEL4B7dnFlVVVeTl5ZGZmQlAZmYmeXl52O32Du2mT59OaGgoACkpKSiKQnV1NQDvvPMO2dnZAAwdOpTU1FT279/vdpsQQgj/4DYsbDYbZrMZjUYDgEajISYmBpvNdsPnbNu2jSFDhhAbGwtAWVkZ8fHxru0Wi4Xy8nK324QQQvgHr08QdPjwYV544QVeffVVb3d9Q0aj3qN2JlNEL1fSOyrtjUToQ7zWX1hYMCZDmNf660qgHnOpu28Fat0Q2LV3h9uwsFgsVFRU4HA40Gg0OBwOKisrsVgsndoePXqUJ598kk2bNjF8+HDX43FxcZSWlmIwGID2s5UpU6a43eapqqp6nE6lyzYmUwQXL9Z1q1+/odFQV9/kte4aG5u56HB4rb8bCdRjLnX3rUCtGwK39p4EnNthKKPRiNVqJTc3F4Dc3FysVqvrl/tVx44dY+XKlbz44ouMGTOmw7aMjAy2bNkCQHFxMcePH2f69OlutwkhhPAPKkVRun5LDhQWFpKTk0NtbS2RkZGsW7eO4cOHs2zZMlasWMHYsWO55557KC0txWw2u563fv16UlJSaGxsJCcnh5MnT6JWq3nyySeZPXs2QJfbPNXfzywUjYYPPy/xWn+TrWbC+2CK8kA95lJ33wrUuiFwa+/JmYVHYeHvJCy6R8Kia1J33wrUuiFwa++VYSghhBBCwkIIIYRbEhZCCCHckrAQQgjhloSFEEIItyQshBBCuCVhIYQQwi0JCyGEEG5JWAghhHBLwkIIIYRbEhZCCCHckrAQQgjhloSFEEIItyQshBBCuCVhIYQQwi2PwqKoqIjs7GzS09PJzs6muLi4U5uDBw+yYMECUlNTWbduXYdtTz31FFlZWa4/o0eP5m9/+xsAGzZsYOrUqa5ta9asufm9EkII4VUerYCzevVqFi9eTFZWFtu3b2fVqlVs3ry5Q5vExETWrl3L7t27aWlp6bBt/fr1rq/z8/O5//77OyydOn/+fJ5++umb2Q8hhBC9yO2ZRVVVFXl5eWRmZgKQmZlJXl4edru9Q7ukpCSsVitabdf588YbbzBv3jyCgoJuomwhhBB9yW1Y2Gw2zGYzGo0GAI1GQ0xMDDabrdsv1tLSwo4dO7jnnns6PL5z507mzZvH0qVLOXr0aLf7FUII0bt6fyHma+zbt4+4uDisVqvrsUWLFvHII4+g0+k4dOgQy5cvZ9euXURHR3vcr9Go96hdT9ad9QeV9kYi9CFe6y8sLBiTIcxr/XUlUI+51N23ArVuCOzau8NtWFgsFioqKnA4HGg0GhwOB5WVlVgslm6/2JtvvtnprMJkMrm+njZtGhaLhYKCAm677TaP+62qqsfpVLpsE6gLqwOg0VBX3+S17hobm7nocHitvxsJ1GMudfetQK0bArf2ngSc22Eoo9GI1WolNzcXgNzcXKxWKwaDoVsvVF5ezueff868efM6PF5RUeH6+uTJk5SWljJs2LBu9S2EEKJ3eTQM9cwzz5CTk8OmTZuIjIx0fTR22bJlrFixgrFjx3LkyBGeeOIJ6uvrURSFnTt3snbtWtennt566y1mzpxJVFRUh76ff/55Tpw4gVqtRqfTsX79+g5nG0IIIXxPpShK1+M3AaC/D0MpGg0ffl7itf4mW82EB/f+5apAPeZSd98K1LohcGvvlWEoIYQQQsJCCCGEWxIWQggh3JKwEEII4ZaEhRBCCLckLIQQQrglYSGEEMItCQshhBBuSVgIIYRwS8JCCCGEWxIWQggh3JKwEEII4ZaEhRBCCLckLIQQQrglYSGEEMItCQshhBBueRQWRUVFZGdnk56eTnZ2NsXFxZ3aHDx4kAULFpCamupaSe+qDRs2MHXqVLKyssjKymLNmjWubVeuXOHxxx9nzpw5ZGRk8P7779/cHgkhhPA6j5ZLW716NYsXLyYrK4vt27ezatUqNm/e3KFNYmIia9euZffu3bS0tHTqY/78+Tz99NOdHn/llVfQ6/Xs3buX4uJi7r33Xvbs2UN4eHgPd0kIIYS3uT2zqKqqIi8vj8zMTAAyMzPJy8vDbrd3aJeUlITVakWr7d5yne+88w7Z2dkADB06lNTUVPbv39+tPoQQQvQut7/ZbTYbZrMZjUYDgEajISYmBpvNhsFg8PiFdu7cycGDBzGZTDz22GOMHz8egLKyMuLj413tLBYL5eXl3doJo1HvUbuerDvrDyrtjUToQ7zWX1hYMCZDmNf660qgHnOpu28Fat0Q2LV3R/dOA3po0aJFPPLII+h0Og4dOsTy5cvZtWsX0dHRXum/qqoep1Ppsk2gLqwOgEZDXX2T17prbGzmosPhtf5uJFCPudTdtwK1bgjc2nsScG6HoSwWCxUVFTi++eXicDiorKzEYrF0ozATOp0OgGnTpmGxWCgoKAAgLi6O0tJSV1ubzUZsbGy3dkIIIUTvchsWRqMRq9VKbm4uALm5uVit1m4NQVVUVLi+PnnyJKWlpQwbNgyAjIwMtmzZAkBxcTHHjx9n+vTp3doJIYQQvcujYahnnnmGnJwcNm3aRGRkpOujscuWLWPFihWMHTuWI0eO8MQTT1BfX4+iKOzcuZO1a9cyffp0nn/+eU6cOIFarUan07F+/XpMJhMADz74IDk5OcyZMwe1Ws2zzz6LXu/ZNQghhBB9Q6UoSteD/QGgv1+zUDQaPvy8xGv9TbaaCQ/u/ctVgXrMpe6+Fah1Q+DW3ivXLIQQQggJCyGEEG5JWAghhHBLwkIIIYRbEhZCCCHckrAQQgjhloSFEEIItyQshBBCuCVhIYQQwi0JCyGEEG5JWAghhHBLwkIIIYRbEhZCCCHckrAQQgjhloSFEEIItzwKi6KiIrKzs0lPTyc7O5vi4uJObQ4ePMiCBQtITU11LY501caNG5k7dy7z5s1jwYIFHDhwwLUtJyeHGTNmkJWVRVZWFi+99NLN7ZEQQgiv82gFnNWrV7N48WKysrLYvn07q1atYvPmzR3aJCYmsnbtWnbv3k1LS0uHbWlpaSxdupTQ0FDy8/NZsmQJBw8eJCQkBICHH36YJUuWeGmXhBBCeJvbM4uqqiry8vLIzMwEIDMzk7y8POx2e4d2SUlJWK1WtNrO+TN9+nRCQ0MBSElJQVEUqqurvVG/EEKIPuD2zMJms2E2m9FoNABoNBpiYmKw2WwYDIZuv+C2bdsYMmQIsbGxrsdee+01tmzZQmJiIj/96U9JTk7udr+iXXOrg9Ml1ZyvrCc4SEN4iJak2AgsxnBflyaECGC9vxDzNQ4fPswLL7zAq6++6nps5cqVmEwm1Go127Zt46GHHmLfvn2ucPKE0aj3qF1P1p31B5X2RiL0IW7bHT1dyeETFbQ5nJgNYTS3OrhYfYXT52tITohiWlocEWFBhIUFYzKE9UHlgXvMpe6+Fah1Q2DX3h1uw8JisVBRUYHD4UCj0eBwOKisrMRisXTrhY4ePcqTTz7Jpk2bGD58uOtxs9ns+nr+/Pn8+te/pry8nPj4eI/7rqqqx+lUumwTqAurA6DRUFff1GWT/HOXOXyykgRTOONHDSY6oj1cHA4nJ4rsHD9rp6S8jtmTEhidGMVFh6PXyw7UYy51961ArRsCt/aeBJzbaxZGoxGr1Upubi4Aubm5WK3Wbg1BHTt2jJUrV/Liiy8yZsyYDtsqKipcXx84cAC1Wt0hQIR7xbZaDp+sJDFGz3fHx7uCAkCjUZM2YjB3f2coIUEa9n12gcLSGh9WK4QIRCpFUbp+Sw4UFhaSk5NDbW0tkZGRrFu3juHDh7Ns2TJWrFjB2LFjOXLkCE888QT19fUoikJERARr165l+vTp3HPPPZSWlnYIgfXr15OSksIDDzxAVVUVKpUKvV7PU089xa233tqtnejvZxaKRsOHn5dcd1t1fTO5h84xeFAIsycloNXcOP8bm1rZ89kFmlra+Nmi8YyIj+qtkoHAPeZSd98K1LohcGvvyZmFR2Hh7wZyWPzt8wtUXr7CD2YMIyTI/SWoK81tvP9FKa1tTlY9MJnoiGBvl+sSqMdc6u5bgVo3BG7tvTIMJfxX2aUGSi82kJZs9CgoAEKDtTx89xiaWh38+1+P0drW+9cuhBCBT8IiQDmdCkfyK9GH6hidNKhbz7UMDmdZ5i0U2er4057TvVShEKI/kbAIUIVlNVTXtzAxxYRG3f1/xgmjTMydmsSBYzY+P1XZCxUKIfoTCYsApCgKJ4svY4gMZojZs3tMrifrO8NIio3gP3eforq+2YsVCiH6GwmLAFRZfYXq+hZSEgehUql63I9Wo+bhebfQ0urgtV359IPPOggheomERQA6XVKNTqtmqCXypvuyGMNZOHMEx89W8cHRUi9UJ4TojyQsAkxTSxvnyutJjotEp/XOP9+sCfGkDjOw5b0z2KoavNKnEKJ/kbAIMGcu1OBUFEYN6d4noLqiUqn4v9+3otOqeTk3jzaH02t9CyH6BwmLAKIoCqfP12A2hDJI792b6aIjgrk/YzRFtjpyPyr2at9CiMAnYRFAqmqbqb/SSnJc70zTMWl0DFPHmMn96BzF5bW98hpCiMAkYRFAzpXXoVJBYkzPPy7rzuI5o4jSB/Fy7km5u1sI4SJhESAURaGkog6LMYzgIM/X+uiu8BAd//eu0ZRdauCt/UW99jpCiMAiYREg7HXN1DW2khTb+wutpA438t3x8bx7uITT52X5WyGEhEXAKOmDIahr/XBmMoMHhfDKzjyaWtr65DWFEP5LwiIAKIpCcXkdsYYwj2eXvVkhQVoenHsLl6qb+J/3C/vkNYUQ/sujsCgqKiI7O5v09HSys7MpLi7u1ObgwYMsWLCA1NRU1q1b12Gbw+FgzZo1zJ49mzlz5rB161aPtol2pRfr24egzH271u+oxEGk3zaED46WcvxsVZ++thDCv3gUFqtXr2bx4sW8++67LF68mFWrVnVqk5iYyNq1a3nwwQc7bduxYwclJSXs2bOHLVu2sGHDBi5cuOB2m2j3VcElABJvYtLAnvrBjGEkmMJ5JTePmoaWPn99IYR/cBsWVVVV5OXlkZmZCUBmZiZ5eXnY7fYO7ZKSkrBarWi1nYdJdu3axcKFC1Gr1RgMBmbPns3u3bvdbhPtvj57CUNkMKHBfTMEdS2dVsM/3D2GKy0OXtmZh1MmGxRiQHIbFjabDbPZjEbT/nFNjUZDTEwMNpvN4xex2WzExcW5vrdYLJSXl7vdJqCxqY3CCzXEDQ73WQ3xJj2LZo3g67N29n123md1CCF8p+/fqvYCo9Gz4ZmerDvrax8fL8PhVBiZGE2EPsQrfYaFBWMyhHXrOQvvHE1BWS1vfFjI7ePiSU7wbG6qQDzmIHX3tUCtGwK79u5wGxYWi4WKigocDgcajQaHw0FlZSUWi8XjF7FYLJSVlZGWlgZ0PJvoapunqqrqcTq7Hh4J1IXVD31ZSkiQhvBgDXX1TV7ps7GxmYuO7t+dvfh7Izl17jL/8p+fsfqByW5vDgzUYy51961ArRsCt/aeBJzbYSij0YjVaiU3NxeA3NxcrFYrBoPB4xfJyMhg69atOJ1O7HY7+/btIz093e22gU5RFL4+W8Utw4yo1T1f5Mhb9KE6Hsq8hQp7I/+9T9buFmIg8ejTUM888wx/+tOfSE9P509/+hNr1qwBYNmyZRw/fhyAI0eOMGPGDF577TVef/11ZsyYwYEDBwDIysoiISGBO++8kx/+8Ic8+uijJCYmut020JVVNVJV20zqcKOvS3GxJkXz/W/W7v4sX9buFmKgUCn9YC3N/joM9e7hEra8d4Z1P/4OJ85c9Fq/k61mwm/ik1VtDif/8ucvsFU1smbpZAZHhV63XSAec5C6+1qg1g2BW3uvDEMJ3zlRZMdiDLvhL2Nf0WrUPHz3GBRF4Q878nA4ZbEkIfo7CQs/5XA6KbhQw+ikaF+Xcl0xg0K5Lz2FMxdqyP3onK/LEUL0MgkLP1VSUU9zq4OURO8tn+ptU8fEMnWMmR2HiiksrfF1OUKIXiRh4adOlbRPDT7Kj8MC4N45KURHBPOHHSe40iyz0wrRX0lY+KnT56sxR3t/rW1vCwvRsmzeLVyqaeIv+wp8XY4QopdIWPghp6Jw+ny1359VXDUqcRDfvz2Jg8dtHJGP0wrRL0lY+KHSiw00NreRMiQwwgIg6zvDGBobwX/uzudyXbOvyxFCeJmEhR86VXIZ8P/rFde6+nHaVodTZqcVoh+SsPBDp89XY4wM9rv7K9yJNYSx6HsjySu+zHufy5okQvQnEhZ+RnFdr/DP+yvcuWNcHKnDDLz54Vkq7I2+LkcI4SUSFn6m3N5IbWNrQF2vuJZKpeL+jNGggn//ny/pB7PJCCGQsPA7p84Hxv0VXTFGhfDD7ybzZcFFDh7zfJEsIYT/krDwM6fPVxMVHoQ5OrCuV3zbHePjGTPcyP+8f4ZaWbtbiIDXL1bK6y8UReFUSfv9FSpV761foVKraPDi3dY6rZbWts79LZo9itUvf8Jf/lbAkvQUj/sL1mnRytsYIfyKhIUfuVTTxOW65l4fgmpudfDVae9NeT5ulOm6/UXoQ7glKZpP8yoYpA/C7OFSrpOtZrQ3MYW6EML75P2bHzn9zfUKf548sLvSRhgJD9HyaV6F2zVHhBD+y6O3b0VFReTk5FBdXc2gQYNYt24dQ4cO7dDG4XDw3HPPceDAAVQqFQ8//DALFy4E4KmnnuLUqVOutqdOnWLjxo1873vfY8OGDfz3f/83MTExAEyYMIHVq1d7afcCy6nz1YSHaIkzhfu6FK/RatTcdouZ978o5VRJNdahgfmRYCEGOo/CYvXq1SxevJisrCy2b9/OqlWr2Lx5c4c2O3bsoKSkhD179lBdXc38+fOZOnUqCQkJrF+/3tUuPz+f+++/n+nTp7semz9/Pk8//bSXdilwnf7meoW6F69X+EKCKRyLMYyvCi8xLC6SkCCNr0sSQnST22Goqqoq8vLyyMzMBCAzM5O8vDzsdnuHdrt27WLhwoWo1WoMBgOzZ89m9+7dnfp74403mDdvHkFBQV7ahf7hcl0zldVXAvojszeiUqmYPDqG1jYnX5255OtyhBA94PbMwmazYTab0Wja3w1qNBpiYmKw2WwYDIYO7eLi4lzfWywWysvLO/TV0tLCjh07+I//+I8Oj+/cuZODBw9iMpl47LHHGD9+fLd2wmjUe9SuJ+vO9pWTF9oXD7o9Lb5TnZX2RiL0IV57LZ1O22f9XX08Qh9C6nAjX5+tYnyKGWPUjV8/LCwYk4cXw3uLP/+sdEXq7nuBXHt39OlHTvbt20dcXBxWq9X12KJFi3jkkUfQ6XQcOnSI5cuXs2vXLqKjPR/brqqqd3vx1N8XVj9yopyQIA36IFXnOjUa6uqbvPZara1tfdJfhD6kw+PWpGhOnbvMwS8vMGtiwg37a2xs5qLD4bX6usvff1ZuROrue4Fae08Czu0wlMVioaKiAsc3/3kdDgeVlZVYLJZO7crKylzf22w2YmNjO7R58803ueeee75VtAmdTgfAtGnTsFgsFBQMvEV0Tp2vZkRCFBp1//2AWkiQhtThBi5cbKDisswbJUQgcfubyWg0YrVayc3NBSA3Nxer1dphCAogIyODrVu34nQ6sdvt7Nu3j/T0dNf28vJyPv/8c+bNm9fheRUVFa6vT548SWlpKcOGDbupnQo0tY0tlF1q6Fcfmb2R0UnRhAZr+OLUJZk3SogA4tEw1DPPPENOTg6bNm0iMjKSdevWAbBs2TJWrFjB2LFjycrK4quvvuLOO+8E4NFHHyUxMdHVx1tvvcXMmTOJiorq0Pfzzz/PiRMnUKvV6HQ61q9fj8lk8tb+BYSC8+3XK1ICdKbZ7tBq1IxLHswneRWUXmwgIcaz601CCN/yKCySk5PZunVrp8f/+Mc/ur7WaDSsWbPmhn386Ec/uu7jV4NnIDt1/jJBWjVDLQPjQtmIhChOFNv54vRF4kzh/e6jwkL0R/13gDyAnD5fTXJ8FFrNwPjnUKtV3DpyMNX1LRTban1djhDCAwPjt5Mfa2xq43xFfb+8v6IrQ2MjMEQG82VBFQ6ZBkQIvydh4WNnSqtRCOz1K3pCpVIxfqSJ+iutFHwzJ5YQwn9JWPjYqZJqNGoVyXGRvi6lz8UNDsNsCOVYYRWtbU5flyOE6IKEhY+dPl/NsLhIgnQDb74klUrFhFEmmloc5J+77OtyhBBdkLDwoeYWB8XldQPi/oobMQ0KJd4UzoliOy2tvrtrWwjRNQkLHzpTVoPDqQy46xXfduuIwbS0OskrlrMLIfyVhIUPnS6pRqWCEfFR7hv3Y8aoEIaY9Zw8d5nmFjm7EMIfSVj40Onz1SSZIwiVJUQZN2IwrW1OThTb3TcWQvQ5CQsfaW1zUlhWO+CHoK6KjghmaGwE+ecuU9fY4utyhBDfImHhI0W2WtocTlKGSFhcNW6EEYdD4W9HLvi6FCHEt0hY+Mipb25EG5kgYXFVlD6YYXGR7P+qjJr6Zl+XI4S4hoSFj5w+X02CKRx9qM7XpfiVtGQjDoeTnR+f83UpQohrSFj4QJvDyZkLNQNiSvLuigwP4rZbzHzwZSn2Wu+t5ieEuDkSFj5QUlFPc6uDUXK94roypiShKJArZxdC+A2PwqKoqIjs7GzS09PJzs6muLi4UxuHw8GaNWuYPXs2c+bM6bD+xYYNG5g6dSpZWVlkZWV1WPfiypUrPP7448yZM4eMjAzef//9m98rP5df0n7zmXwS6vqMUSFMHxfHga/KuFR9xdflCCHwcPGj1atXs3jxYrKysti+fTurVq1i8+bNHdrs2LGDkpIS9uzZQ3V1NfPnz2fq1KkkJCQAMH/+fJ5++ulOfb/yyivo9Xr27t1LcXEx9957L3v27CE8PNwLu+efTp67TPzgcKLCg3xdit/KnJrEwWM23v6omKXft/q6HCEGPLdnFlVVVeTl5ZGZmQlAZmYmeXl52O0db57atWsXCxcuRK1WYzAYmD17Nrt373ZbwDvvvEN2djYAQ4cOJTU1lf379/dkXwJCm8NJwYVqRifJ9YquGCJD+O6tcXx0vJyKy42+LkeIAc9tWNhsNsxmMxpN+6yoGo2GmJgYbDZbp3ZxcXGu7y0WC+Xl5a7vd+7cybx581i6dClHjx51PV5WVkZ8fPwNn9ffnC2rpaXViVXCwq25U5PQalS8fbDI16UIMeD1yTwTixYt4pFHHkGn03Ho0CGWL1/Orl27iI72zi9Mo1HvUTuTyfdrXO87WoZKBdPGJ6AP82wYqtLeSIQ+xGs16HTaPuuvJ68TFhaMyRCGyRTB3O8MZ9uHZ7hv7hgSzX337+cPPys9IXX3vUCuvTvchoXFYqGiogKHw4FGo8HhcFBZWYnFYunUrqysjLS0NKDjmYbJZHK1mzZtGhaLhYKCAm677Tbi4uIoLS3FYDC4njdlypRu7URVVT1ON0tzmkwRXLxY161+e8PneeUMiYngSkMzVxo8vPFMo6Gu3nsfI21tbeuT/iL0IT16ncbGZi462icUvCMtll2HiviPHV/zSFbqTdfqCX/5WekuqbvvBWrtPQk4t8NQRqMRq9VKbm4uALm5uVitVtcv96syMjLYunUrTqcTu93Ovn37SE9PB6CiosLV7uTJk5SWljJs2DDX87Zs2QJAcXExx48fZ/r06d3ekUDQ3OqgsKxGhqC6ITIsiNmTEjh8spILlfW+LkeIAcujYahnnnmGnJwcNm3aRGRkJOvWrQNg2bJlrFixgrFjx5KVlcVXX33FnXfeCcCjjz5KYmIiAM8//zwnTpxArVaj0+lYv36962zjwQcfJCcnhzlz5qBWq3n22WfR6z0bVgo0Z0praHMocnG7m9JvG8J7X1zgrQNneeyeNF+XI8SA5FFYJCcnd7hv4qo//vGPrq81Gk2H+yeudTVcricsLIwXX3zRkzICXv65y2jUKkYmDOz1K7pLH6oj/bYhbDtQxJkLNYyQ4ydEn5M7uPtQXrGdYZZIWb+iB9InDyFKH8SW9wpQlK6vTwkhvE/Coo/UNrZQbKsjdbjBfWPRSXCQhgXTh1NYVstn+ZW+LkeIAUfCoo/kFdlRgLHDjb4uJWBNG2shwaTnjQ8KaW1z+rocIQYUCYs+cvxsFfpQHUmxA+Mz2b1BrVaRPWsEl2qa2PNZia/LEWJAkbDoA05F4esiO6nDDKhVKl+XE9DGDDMwfuRgdnxULFOYC9GHJCz6QElFHd2ESm0AABqQSURBVHWNrTIE5SV//72RKAq8/t4ZX5cixIAhYdEHjp9tn3RxzDC5uO0NgweFMndqEkfyKzlRbHf/BCHETZOw6ANfn60iKTaCSJmS3GvumjKEmEGh/Ne7p2hudfi6HCH6PQmLXtbQ1EphaS1j5SOzXqXTarj/rtFUXr7CW/vP+rocIfo9CYte9tWZSzgVhXEjBvu6lH7HmhTNd8fHs/ez85wprfF1OUL0axIWvezzUxeJjghmmCXS16X0Swu/m4whMphXd56U4SghepGERS9qbnHwdZGdCSNN8pHZblCpVTQ0t3n0xwn8/ZxRVNgb2fzuqeu2kfv3hLh5MklRLzp+torWNicTUkzuGwuX5lYHX52+2K3npA438PHX5WjUKobHdTyLm2w1o5X5uIS4KXJm0Yu+OH0RfaiOUYkyS2pvGzdiMDHRoXxyopya+hZflyNEvyNh0Uta25x8VXiJW0cORqOWw9zb1GoV08dZ0KjVvPfFBZpa2nxdkhD9ike/xYqKisjOziY9PZ3s7GyKi4s7tXE4HKxZs4bZs2czZ86cDutfbNy4kblz5zJv3jwWLFjAgQMHXNtycnKYMWMGWVlZZGVl8dJLL938XvmBk+cuc6XZwcRRMgTVV8JDdMycEE9DUxvvf1GGwyEXK4TwFo8GclevXs3ixYvJyspi+/btrFq1is2bN3dos2PHDkpKStizZw/V1dXMnz+fqVOnkpCQQFpaGkuXLiU0NJT8/HyWLFnCwYMHCQkJAeDhhx9myZIl3t87Hzp8soLQYA23DJVV8fpSTHQo30mzsP/LMg4eL2d6msX9k4QQbrk9s6iqqiIvL4/MzEwAMjMzycvLw27vOM3Crl27WLhwIWq1GoPBwOzZs9m9ezcA06dPJzQ0FICUlBQURaG6utrb++I3rjS3ceRUJbdZzei0Gl+XM+AMjY1gYoqJc+V1HPhKzjCE8Aa3YWGz2TCbzWg07b/0NBoNMTEx2Gy2Tu3i4uJc31ssFsrLyzv1t23bNoYMGUJsbKzrsddee4158+axfPlyCgsLe7wz/uJIfiUtrU6mjZV3tb4yZpiBSSkmzlXU80ruSVrkHgwhbkqffp7w8OHDvPDCC7z66quux1auXInJZEKtVrNt2zYeeugh9u3b5wonTxiNeo/amUx9s5bE4VMXiTeFc/u4eFReuL+i0t5IhD7EC5W10+m0fdZfT17HW/VNGRtHWFgQ+4+W8v+2fMnP77+NGEOYR8/tq58Vb5O6+14g194dbsPCYrFQUVGBw+FAo9HgcDiorKzEYrF0aldWVkZaWhrQ+Uzj6NGjPPnkk2zatInhw4e7Hjebza6v58+fz69//WvKy8uJj4/3eCeqqupxOrtel9lkiuDixTqP++ypisuNnDhbxT13DOfSpXrvdKrRUFfvvbUbWlvb+qS/CH1Ij17Hm/UNNeuxZo1h8+58fvL8BzyUeQtpyV1PFd9XPyveJnX3vUCtvScB53YYymg0YrVayc3NBSA3Nxer1YrB0HFivIyMDLZu3YrT6cRut7Nv3z7S09MBOHbsGCtXruTFF19kzJgxHZ5XUVHh+vrAgQOo1eoOARJoPjpejkoFf5cqQ1D+InW4kVX3TyZKH8S/bf2K3799gpoGuRdDiO7waBjqmWeeIScnh02bNhEZGcm6desAWLZsGStWrGDs2LFkZWXx1VdfceeddwLw6KOPkpiYCMCaNWtoampi1apVrj7Xr19PSkoKTz/9NFVVVahUKvR6PS+99BJabWDebdvmcHLoaxu3DDUQHRHs63LENcyGMFbdP5ldn5xj58fFHCu8xOyJicyZnIg+VOfr8oTweypFUboevwkA/jIM9UleOX94O48V96Rx60jvzTKraDR8+Ln31pweN8rU7ek0etJfT4ehvF3fZKuZ8Gum+7BVNfDXD8/y+emLBAdpmJYay/S0OIaY9ahUqoAeWpC6+1ag1t6TYajAfAvvhxRF4d1PzxNrCCNthCyf6s8sxnAeXTCWCxfreeeTc+z/ysZ7X5QSNziccSOM3DFxCMZwrdx5L8Q1JCy8JP/cZc5V1HF/RorMMBsgEkx6ls0bw+I5rRzOq+DIqYvsOXyedz4pITxEy9hkI2nJRlKHGWWoSgx4EhZesvvweSLDdPxdaqz7xsKvtE8TksDMCQk0NrVx3t7IgS8ucKywik9OVKBSQXJcVHt4DDe6hquEGEgkLLzgfGU9x89W8YPpw+SO7QAXFqLlO+PiSYmLxOlUKLLVcqywiuNnq3hr/1ne2n+WwVEhTEwxMXF0DBZjuFdeN1inRSujXsKPSVjcJEVR+J/3CggL1jJzQoKvyxFepFarSI6PIjk+ih/MGE5NQwvHzlzi47wK9hw+z7uHz3+zCmIEw+OiCAvp+X8nWXND+Dv56bxJxwqrOFF8mUXfGynj2v1cVHgQ08fFMWF0DPu/LKW4vI6islq+OH2JowWXSDDpGZkQRZwpXK5biX5HwuImtDmcbHnvDOboUGZN8PyOcxH4QoO1WJOisSZFU9vQQsGFGgpLazhfWU9YsJYRCVGMiI9CHyZvIET/IGFxE97/opRyeyMr7klDq5EB54EqMjyIiSkmbh05mAuV9RRcqOFYYRXHCquwGMMYmTiIxBg9GrWcbYjAJWHRQ6WXGnjzw0JShxkYJ/dVCECjVpEUG0FSbAT1V1opLK2h4EIN+78sI1inITk+kpEJUUTp5e5+EXgkLHqgpdXB77d/TXCQhqVzrfIxStGJPlTHuBGDGZtsxHapkYIL1Zw8d5m84suYBoWSFKsnMUZPRFiQr0sVwiMSFj2w5b0zXLjYwOMLxzFI3iWKLqhVKuJN4cSbwrnS3EZhaQ1ny2o5kn+RI/kX0YfqiIkOpbnVwZihBuIGy8Vx4Z8kLLpp96clvH+0lPTbEt1OdS38g0qtoqG5zeP2ir2Rxi7au5mG7IZCg7WkDjeSOtxIXWMLFyobqLjcSNmlBrb87YyrjcUYRkx0KOboMGIMocQMCmNQRBAhQV3/d722bp1WS2ub5/vsjr/fB9LmhObWgbO/viBh0Q17j5znf94/w23WGP7Pd5N9XY7wUHOro1sTE7qbAHHcKNNN1xQRFoR1aBDWodEoikKCOYL9R0u5WH2FmvoWvrzUQFNLx9X9tBoVYcFaQkO0hAZp0WnVBOk0BGnV6LRqIvUhOBwOgrRqRg8zUHShBo1GhUatRqtRodGoUKtUPRo29ff7QJpb2/jsZIX7hh7y9/31BTkaHnAqCjs/KuatA0VMGGXiocxbZJI54TUqlYrBg0JdNwBe1dLqoKahhbrGVhqb27jS1Nb+d3MbVbVNtLY5aWlzXnfG5XcPn7/h62m/CRCNRoVWrUKjUaNRq9Bq1N+Ei4ognQZ9qI6IUB36MB11jS2EBWn6/fU5p1PBqSg0tbShU6vQadX9fp89JWHhRm1DC3/ccYITxZe5/RYzS+da5WOyok8E6TSYBoViGhTaZTuHoz00goJ0XK5tpKXVyZDYCM6cr8bhVGhzKDicThwOhTangsPhvO5jDqdCa7ODNqdCc4ujw5nN1ckVh5gjSIzRk2SOINGsx2IMC6g3Ti1tDmrqWqhtbA/husb/DeOWVgdtjvbg/e+9BQCoVBASpMUQGYwxMgRjZAiGyGBiosNIjNFj8HBJ5/5AwuIGWlodvPdFKTs/Lqalzcn9GSnMGBcn7zKE39Fo1IRq1ETog9Go2n/ZjRlupK3NeVP9trY5abjSSt2VVqIjgrlU3cT5yjreP1pK6zd9B2nVJJr1DDVHMtTS/rHhOGM4aj+4p6SppQ17bTP22iaqvvm7rrG1Q5vwEC0RYUFYjGEEfzOkp1arGGKOQKNW0dzq5EpTG/a6JqpqmygsraGh6X+vjQQHaYgfHE5ijJ4hMXqGmCNIiNETrOt/c8R5FBZFRUXk5ORQXV3NoEGDWLduHUOHDu3QxuFw8Nxzz3HgwAFUKhUPP/wwCxcuvKltvmCrauDjExUcPFZGdX0LqcMM/HDWCBJMA+cdhBAAOq2aQRHBDIoI7rCAlMPppLyqkZKKes5V1FFsq+Xg1zb+9sUFAIJ0aobERDD0m3tOkmIjMEeH9tokm20OJ2WXGjhbVkt1fTPVdc3Y65ppvOaXuj5UhyEymOT4KKIjgokMC0IfduM1S769YNa1mlraqLBfoaSyjqq6Fk6fs3Mkv5IPvywD2s9GYg1hJMVGMCQmgiSznoQYPfpQXUC/2fQoLFavXs3ixYvJyspi+/btrFq1is2bN3dos2PHDkpKStizZw/V1dXMnz+fqVOnkpCQ0ONtva3+SivlVY3YqhooLKvh9Pkayu2NqFQwZqiBf7g7iZQh0b1ehxCBRKNWE2/SE2/SM/WbKfmdToVyeyPnyusoKq/lXHkd+4+V0fJ5+xmICoiODCZmUCgx0WEYo0KICNOhD9GRUNuMo7mVYF37NRGVqv06jloFzW3t7+yvfHOtpqGpDXttE5dqm6iqaeJSzRXstc04vrluo1K1z+Fljg7F8M2QkSEihOAg7wVVSJDWFYJXV8pTFAV7bTPnKuooqaijpKKeUyXVfHLify+6h4doMRvCMEeHEWtor2+QPpio8CCi9EF+HyZuw6Kqqoq8vDxee+01ADIzM/nVr36F3W7HYDC42u3atYuFCxeiVqsxGAzMnj2b3bt389BDD/V4m6c8PeW9tl3B+Wpezs3j6qXB0CAto4YMYu7fDWXcCCORfnSzlKJWERbivTmGtBp1n/QXGqzF0db91+mr+m7EXd3ers9bfV5bd28cw67+n6nVKhJi2t9BT0uzAO0BcrGmibJL9d/8Ym8fyjl/sZ78kss3VU9EeBCGiGASYvQYI0OIjgymur6ZiNAgrwyBudvfa7W3U2GKDsUUHcqk0TGubfVXWim71ECFvZGLNU1cqr7CpZomCstqOvWjUasIDtISrFUTHKwhWKtBp1Xz7TKcCrR8Mwy48LvJmA1hPd7P7nAbFjabDbPZjEbTnswajYaYmBhsNluHsLDZbMTFxbm+t1gslJeX39Q2T0VHe7amgPGai1FGo57bbw2cKcXnTvfuR3WHJ3j3jEn6878+e6PG7jKZIrhlxM1/1NifGbu4yG0Ekvzg38EbAudjDEIIIXzGbVhYLBYqKipwONo/RudwOKisrMRisXRqV1ZW5vreZrMRGxt7U9uEEEL4B7dhYTQasVqt5ObmApCbm4vVau0wBAWQkZHB1q1bcTqd2O129u3bR3p6+k1tE0II4R9UiqK4nemmsLCQnJwcamtriYyMZN26dQwfPpxly5axYsUKxo4di8Ph4Nlnn+XQoUMALFu2jOzsbIAebxNCCOEfPAoLIYQQA5tc4BZCCOGWhIUQQgi3JCyEEEK4JWEhhBDCrQE362xOTg4fffQR0dHtd1VmZGTwox/9yMdVXZ8nEzj6o1mzZhEUFERwcPuSsz/72c+YPn26j6vqbN26dbz77ruUlpayY8cORo0aBfj/cb9R3f5+3C9fvsxTTz1FSUkJQUFBJCUl8eyzz2IwGPjyyy9ZtWoVzc3NxMfH85vf/Aaj0T9Wouyq7pSUFEaNGoX6mwkJ169fT0pKio8r7mj58uVcuHABtVpNWFgY//RP/4TVau3+z7kywDz99NPKf/3Xf/m6DI/cd999yrZt2xRFUZRt27Yp9913n48r8szMmTOVU6dO+boMtz777DOlrKysU73+ftxvVLe/H/fLly8rn3zyiev7f/mXf1F+/vOfKw6HQ5k9e7by2WefKYqiKBs3blRycnJ8VWYnN6pbURRl1KhRSn19va9K80htba3r67179yrz589XFKX7P+cyDOWnrk7gmJmZCbRP4JiXl4fdbvdxZf3HpEmTOs1EEAjH/Xp1B4JBgwYxZcoU1/e33norZWVlfP311wQHBzNp0iQAFi1axO7du31VZic3qjtQREREuL6ur69HpVL16Od8wA1DAbz22mts2bKFxMREfvrTn5Kc7H/raXs6gaO/+tnPfoaiKEycOJEnnniCyMhIX5fkETnufcPpdPKXv/yFWbNmdZpM1GAw4HQ6XcMj/uTauq+67777cDgczJgxg8cee4ygIP+ZsfqqX/ziFxw6dAhFUXj55Zd79HPe784sfvCDHzBlypTr/nE4HKxcuZK9e/eyY8cO7rzzTh566CHXvFfCO/785z/z9ttv8+abb6IoCs8++6yvSxoQAum4/+pXvyIsLIwlS5b4upRu+XbdH3zwAX/961/585//zJkzZ9i4caOPK7y+tWvX8sEHH7By5UrWr1/foz76XVi89dZbfPrpp9f9o9FoMJvNrotR8+fPp7GxsdtTovcFTydw9EdXawwKCmLx4sV88cUXPq7Ic3Lce9+6des4d+4c//Zv/4Zare40majdbketVvvdWcW364b/PeZ6vZ6FCxf67TG/av78+Xz66afExsZ2++e834WFOxUV/7ty1YEDB1Cr1ZjNZh9WdH2eTuDobxobG6mrqwNAURR27dqF1Wr1cVWek+Peu55//nm+/vprNm7c6BquSU1NpampiSNHjgDw+uuvk5GR4csyO7le3TU1NTQ1NQHQ1tbGu+++63fHvKGhAZvN5vr+vffeIyoqqkc/5wNubqgHHniAqqoqVCoVer2ep556iltvvdXXZV3XjSZw9Gfnz5/nsccew+Fw4HQ6SU5O5pe//CUxMTHun9zHnnvuOfbs2cOlS5eIjo5m0KBB7Ny50++P+/Xq/t3vfuf3x72goIDMzEyGDh1KSEgIAAkJCWzcuJEvvviC1atXd/jo7ODBg31ccbsb1f3QQw+xatUqVCoVbW1tjB8/nn/8x38kPNyzxdj6wqVLl1i+fDlXrlxBrVYTFRXF008/zZgxY7r9cz7gwkIIIUT3DbhhKCGEEN0nYSGEEMItCQshhBBuSVgIIYRwS8JCCCGEWxIWQggh3JKwEH5t7ty5fPrppz1+/ttvv83SpUtvuP2+++5j69atPe5fiIFCwkL4tZ07d3aY8bO77r77bl599VUvVnR9f/3rX/n7v//7m+7n448/JiMjg3HjxnHfffdRWlrq2tbS0sLPf/5zJkyYwLRp03jttdfc9tfS0sKKFSuYNWsWKSkpNwzelpYW7rrrLmbMmHHT+yD6JwkLIfyE3W7nxz/+MT/5yU84fPgwqamprFy50rV9w4YNnDt3jvfff5/Nmzfz8ssvs3//frf9TpgwgfXr12MymW7Y5pVXXvH7KU2Eb0lYCL82a9YsPvroIzZs2MBPfvITnnrqKcaPH8/cuXM5fvy4q53NZuPHP/4xt99+O1OmTHHNuPrtd/yHDh0iIyODiRMn8uyzz/LtCQzeeOMN7rrrLiZPnsyDDz7Y4Z19SkoKf/nLX7jzzjuZNGkSa9asQVEUCgsLWb16NV9++SXjx493rcvw4Ycf8v3vf5/x48czffp0XnnllS73de/evYwcOZK77rqL4OBgHnvsMfLz8yksLATaJ8lcvnw5UVFRJCcns3DhQt56660u+wwKCuKBBx5g0qRJrsnvvu38+fO8/fbbPPzww132JQY2CQsRMN577z3mzp3LkSNHmDVrFr/61a+A9hkz/+Ef/oG4uDjee+899u/fz/e///1Oz7/6zv3xxx/nk08+YciQIR1mCd23bx+///3v+fd//3c+/vhjJk6cyE9/+tMOfXzwwQe88cYbvP3227zzzjscOHCA5ORk1qxZw6233srRo0ddE+L94he/4Nlnn+Xo0aPk5uZy++23d7l/BQUFHZbkDAsLY8iQIZw5c4aamhouXrzI6NGjXdtHjx7NmTNnun8gv+W5557jiSeecM17JMT1SFiIgDFx4kTuuOMONBoNWVlZ5OfnA3Ds2DEqKyt56qmnCAsL67Dq2rX279/PyJEjycjIQKfTcf/993eYrO7111/n4YcfJjk5Ga1WyyOPPMLJkyc7nF0sW7aMyMhI4uLimDJliquG69FqtZw5c4b6+nqioqIYM2ZMl/vX2NjYYVUzaJ/6uqGhgcbGRqDjqmcRERE0NDR02ac7e/fuxeFwMGfOnJvqR/R/EhYiYFz7iz0kJITm5mba2tpcK61ptV0v/FhZWUlsbKzre5VK1WH+/rKyMv75n/+ZSZMmMWnSJG677TYURekwrf214/6hoaFd/rJ+8cUX+fDDD5k5cyZLlizh6NGjXdYXFhZGfX19h8caGhoIDw8nLCwMoMP2+vr6m5rhtLGxkd/85jf88pe/7HEfYuAYkMuqiv7FYrFgs9loa2vrMjBMJlOHha4URekw17/FYuGRRx7h7rvv7nYNKpWq02NpaWm89NJLtLa28uc//5nHH3+cDz/88IZ9jBw5ssM1iMbGRkpKShgxYgRRUVGYTCby8/OZNm0aAPn5+YwYMaLbtV517tw5SktLuffeewFobW2lrq6OadOmsWXLFhISEnrct+h/5MxCBLy0tDRMJhO//e1vaWxspLm5mc8//7xTuzvuuIOCggL27NlDW1sbmzdv5tKlS67tixYt4g9/+AMFBQUA1NXV8c4773hUg9FopKKigpaWFqD9o6hvv/02dXV16HQ6wsPDb3iB+ao5c+ZQUFDAu+++S3NzMxs3biQlJcW1Rvz8+fN56aWXqKmpobCwkK1bt/KDH/zAbW0tLS00NzcD7YHQ3NyMoiiMHDmSDz74gG3btrFt2zaee+45jEYj27ZtC4iVAUXfkrAQAU+j0fC73/2Oc+fOMXPmTGbMmHHdX/IGg4EXXniB3/72t0yZMoVz584xYcIE1/Y5c+bw0EMP8cQTTzBhwgQyMzM9+mgqwO23386IESP4zne+47ovZPv27cyaNYsJEybw+uuv85vf/KbLPgwGAxs2bOBf//VfmTx5MseOHeP55593bV+xYgWJiYnMnDmT++67jwcffNCj+yIyMjJIS0ujoqKCBx98kLS0NEpLS9FqtZhMJtefqKgo1Go1JpMJjUbj0X6LgUMWPxJCCOGWnFkIIYRwSy5wC9GHfve73/H73/++0+MTJ07k5Zdf9ps+hfg2GYYSQgjhlgxDCSGEcEvCQgghhFsSFkIIIdySsBBCCOGWhIUQQgi3/j8LgiJbJI3ZgwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ecF3C1OIcG5X",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}