ECS708_Lab0.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "ECS708_Lab0.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "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": "3.8.5"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/anbuatgithub/ml-labs/blob/main/ECS708_Lab0.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oDwYmglELzV7"
      },
      "source": [
        "# <font color='gray'> Lab 0: Basic Data Exploration & Visualisation: Getting to Know Your Data </font>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FYZv8R_TImsZ"
      },
      "source": [
        "## 0. Introduction \n",
        "\n",
        "The purpose of this introductory lab is to familiarise you with  **basic exploratory data analysis** of a sample dataset. We will first be using  the Python's **pandas** package in this **notebook** environment. \n",
        "\n",
        "*This session <ins>is a warm-up that will not count toward your final grade</ins>.*\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qx5GxUx1DS5z"
      },
      "source": [
        "**Before we start:** this environment that allows us to enter both text and run codes interactively, is called ***[notebook](https://jupyter.org/)***).\n",
        "\n",
        "There are two types of cells: *Text* and *Code*. You can add your own cells. You can also edit the texts by double-clicking on them. It follows the [markup rules](https://github.com/adam-p/markdown-here/wiki/Markdown-Cheatsheet). \n",
        "\n",
        "In order to execute (run) a cell, you can use one of the following ways:\n",
        "\n",
        "0. `Shift + Enter` : executes a cell and goes to the next one.\n",
        "1. `Ctrl + Enter` : executes a block but stays at the same block. This is equivalent to clicking on the *run* butten to the left of the cell, which appears when you hover the mouse over the `[ ]` icon. \n",
        "2. Use the `Runtime` tab (at the top of the page), which gives you more options as well.\n",
        "\n",
        "Now, test your knowledge(!) by running the next cell (the output should appear right after):"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v0T8VcP9DSWe",
        "outputId": "4753603d-d6a5-4c25-c7a7-1f643d12c595"
      },
      "source": [
        "for i in range(5):\n",
        "  print('{:2d} - Welcome to the Machine Learning lab!'.format(3**i))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " 1 - Welcome to the Machine Learning lab!\n",
            " 3 - Welcome to the Machine Learning lab!\n",
            " 9 - Welcome to the Machine Learning lab!\n",
            "27 - Welcome to the Machine Learning lab!\n",
            "81 - Welcome to the Machine Learning lab!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T_hB_CXQNTho"
      },
      "source": [
        "As you might have noticed, we will be using Python code. Should we have a word on why `Python`? \n",
        "\n",
        "*Well, for one, it is the most preferred language among data scientist, according to [this poll](https://www.kdnuggets.com/2018/05/poll-tools-analytics-data-science-machine-learning-results.html).\n",
        "Also it helps that it is a general purpose programming language with a simple syntax. It also has a lot of great open-source libraries -- which, in Python, we call them `packages` -- developed by an actve community. If interested, [read more](https://www.cbtnuggets.com/blog/technology/data/why-data-scientists-love-python) on why Python is particularly so popular among data scientist.*"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A0dzRa5JLXwm"
      },
      "source": [
        "#### Some more `notebook` cool tricks:\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yqt3m-dISdQH"
      },
      "source": [
        "- You can use the **`tab`** button for auto-completing. You can also use `tab` after a **`dot`** to be shown a drop-down list of the available attributes and methods on an object or a class. For instance, let's create a string object called `mystring` (by executing the first cell), and use the `tab` key after a `dot` to see a list of available attributes and methods for a string object in Python (in the second cell: don't run it, just use the tab key after dot!)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5lWMrlANLVgS"
      },
      "source": [
        "mystring = '  Hello, World!'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 129
        },
        "id": "AaQ7X-h6OVOm",
        "outputId": "0595cd81-532a-4ee9-8d45-839e3a2621e3"
      },
      "source": [
        "mystring."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "SyntaxError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;36m  File \u001b[0;32m\"<ipython-input-3-41df6beba87c>\"\u001b[0;36m, line \u001b[0;32m1\u001b[0m\n\u001b[0;31m    mystring.\u001b[0m\n\u001b[0m             ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m invalid syntax\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p7v8mCdQOpgZ"
      },
      "source": [
        "> Note that in Python, every variables is an objects. \n",
        ">Also by now, you shoul have noticed that we don't need to explicitly declare variables and their types in Python before using them!\n",
        "\n",
        "> Another thing to keep in mind is that objects and functions (runtime variables) persist between different cells in the same notebook session. If you want to clear the memroy, you can choose \"`Reset all runtimes...`\" under the \"`Runtime`\" tab from the top-left menue. \n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9FjvhE_VR2BC"
      },
      "source": [
        "- Another cool point about the `notebook` environment is that if you want to get help on anything (a method, a function, an object, etc), you can just put a question mark in front of it and execute that line. For instance, let's get help on what the `strip` method does on a string oject, by executing the folloing cell:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RJTJ4wLFSXTz"
      },
      "source": [
        "mystring.strip?\n",
        "# This opens the help on the method \"strip\" at the bottom of this\n",
        "# page. You can close the help page after reading it.\n",
        "# By the way, this is how we designate a comment line in Python!"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RQrmMEyJU9wa"
      },
      "source": [
        "From now on, we may give a hint about answering a question by putting a question mark in front of a command to open the help page for it for you. Your task is to complete the command. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WOhjRZRZVIVC",
        "outputId": "e6a17ee5-0260-45d2-861b-4d75aa45b52f"
      },
      "source": [
        "print(mystring)\n",
        "print(mystring.strip())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "  Hello, World!\n",
            "Hello, World!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kGfrmq6NnBFG"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eetQ-iqXnBy6"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KKlNqAg6MzRP"
      },
      "source": [
        "## 1.   Explore the Dataset -- using Python's `Pandas`"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jMy4U6wpdclq"
      },
      "source": [
        "Here, we are going to perform some basic *exploratory data analysis* on some sample data. The data is about the \"cars for sale\", which was collected in Summer 2014 from the website _Autotrader.com_,  by one of our MSc student who worked on a data mining project. The data is saved in a CSV file, called **`LondonCars2014.csv`**."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MGVOTd6ml6NT"
      },
      "source": [
        "---\n",
        "> **Q0:** Why are such files called `CSV`? (Find out what CSV stands for. It may also help to try and open the fie using a basic text editor).\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dNKGLzR0OTSj"
      },
      "source": [
        "> **A0:** CSV stands for comma-separated values . A comma-separated values (CSV) file is a delimited text file that uses a comma to separate values. Each line of the file is a data record. Each record consists of one or more fields, separated by commas\n",
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mD3zkB8aF2NX"
      },
      "source": [
        "#### 0.   Upload the CSV file to this machine.\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZvDYSO7gPnfU"
      },
      "source": [
        ">-   On the left side of this page, click on the arrow to \"Open the left pane\". \n",
        "-   From the \"Files\" tab, click on `UPLOAD`\n",
        "- Choose the CSV file that you downloaded from QM+ to your local machine, to be uploaded to *this machine*.\n",
        "\n",
        ">_Note:_ *By \"this machine\", we mean the virtual machine that is allocated to your account and running this notebook, hosted by Google Comptute Engine. Note that next time you log in, the file may be gone and needs to be re-uplaoded, as Google recycles Virtual machines when they are idle for a while.*"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5m8tYlpfNJ93"
      },
      "source": [
        "\n",
        "\n",
        "#### 1.   Load the dataset from the CSV file:\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k0gxX165m7JI"
      },
      "source": [
        "We will use a package in Python called [pandas](https://pandas.pydata.org/): it has many useful features to work with structured data, and is popualr for its ease of use (and as is the case for any Python package, it is [open-source](https://github.com/pandas-dev/pandas)).\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gFA_eR7m516a"
      },
      "source": [
        "So, let's start by reading the csv file into a `pandas' DataFrame`:\n",
        "\n",
        "(Note that we use **`import`** to (guess what) import a package in Python! We can also assign a different name (alias) to it, usually a shorthand name for convenience. For example, we typically `import pandas as pd`, because we lazy!)\n",
        "\n",
        "Execute the following code block. \n",
        "\n",
        "\n",
        "\n",
        " \n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "cellView": "both",
        "id": "KjvGRNGgNnzF"
      },
      "source": [
        "import pandas as pd\n",
        "df = pd.read_csv('./LondonCars2014.csv')\n",
        "#df = pd.read_csv(\"./LondonCars2014.csv\", index_col = None)\n",
        "#df.tail()"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p_m8B7bVT98P"
      },
      "source": [
        "Let's get more information about the `read_csv` method that we used by executing the folloing cell:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zFLpA98_QmZa"
      },
      "source": [
        "pd.read_csv?"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "exMOTcLuU20P"
      },
      "source": [
        "---\n",
        "> **Q1:** You should notice that the command has many options. When these options are not specified, their default values are taken. For example, find out what these default values were for our `read_csv`. Also find out, in particular, what the options `sep`, `header`, `index_col`, `usecols`, `dtype`, `na_values`, and `encoding` do."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rt0KmDoOWURV"
      },
      "source": [
        "> **A1:** \n",
        "\n",
        "**sep** str, default ‘,’\n",
        "Delimiter to use. If sep is None, the C engine cannot automatically detect the separator, but the Python parsing engine can, meaning the latter will be used and automatically detect the separator by Python’s builtin sniffer tool, csv.Sniffer. In addition, separators longer than 1 character and different from '\\s+' will be interpreted as regular expressions and will also force the use of the Python parsing engine. Note that regex delimiters are prone to ignoring quoted data. Regex example: '\\r\\t'.\n",
        "\n",
        "**header** int, list of int, default ‘infer’\n",
        "Row number(s) to use as the column names, and the start of the data. Default behavior is to infer the column names: if no names are passed the behavior is identical to header=0 and column names are inferred from the first line of the file, if column names are passed explicitly then the behavior is identical to header=None. Explicitly pass header=0 to be able to replace existing names. The header can be a list of integers that specify row locations for a multi-index on the columns e.g. [0,1,3]. Intervening rows that are not specified will be skipped (e.g. 2 in this example is skipped). Note that this parameter ignores commented lines and empty lines if skip_blank_lines=True, so header=0 denotes the first line of data rather than the first line of the file.\n",
        "\n",
        "**index_col** int, str, sequence of int / str, or False, default None\n",
        "Column(s) to use as the row labels of the DataFrame, either given as string name or column index. If a sequence of int / str is given, a MultiIndex is used.\n",
        "\n",
        "Note: index_col=False can be used to force pandas to not use the first column as the index, e.g. when you have a malformed file with delimiters at the end of each line.\n",
        "\n",
        "**usecols** list-like or callable, optional\n",
        "Return a subset of the columns. If list-like, all elements must either be positional (i.e. integer indices into the document columns) or strings that correspond to column names provided either by the user in names or inferred from the document header row(s). For example, a valid list-like usecols parameter would be [0, 1, 2] or ['foo', 'bar', 'baz']. Element order is ignored, so usecols=[0, 1] is the same as [1, 0]. To instantiate a DataFrame from data with element order preserved use pd.read_csv(data, usecols=['foo', 'bar'])[['foo', 'bar']] for columns in ['foo', 'bar'] order or pd.read_csv(data, usecols=['foo', 'bar'])[['bar', 'foo']] for ['bar', 'foo'] order.\n",
        "\n",
        "If callable, the callable function will be evaluated against the column names, returning names where the callable function evaluates to True. An example of a valid callable argument would be lambda x: x.upper() in ['AAA', 'BBB', 'DDD']. Using this parameter results in much faster parsing time and lower memory usage.\n",
        "\n",
        "**dtype** Type name or dict of column -> type, optional\n",
        "Data type for data or columns. E.g. {‘a’: np.float64, ‘b’: np.int32, ‘c’: ‘Int64’} Use str or object together with suitable na_values settings to preserve and not interpret dtype. If converters are specified, they will be applied INSTEAD of dtype conversion.\n",
        "\n",
        "**na_values** scalar, str, list-like, or dict, optional\n",
        "Additional strings to recognize as NA/NaN. If dict passed, specific per-column NA values. By default the following values are interpreted as NaN: ‘’, ‘#N/A’, ‘#N/A N/A’, ‘#NA’, ‘-1.#IND’, ‘-1.#QNAN’, ‘-NaN’, ‘-nan’, ‘1.#IND’, ‘1.#QNAN’, ‘<NA>’, ‘N/A’, ‘NA’, ‘NULL’, ‘NaN’, ‘n/a’, ‘nan’, ‘null’.\n",
        "\n",
        "**encoding** str, optional\n",
        "Encoding to use for UTF when reading/writing (ex. ‘utf-8’). List of Python standard encodings .\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9izm3YejNhhM"
      },
      "source": [
        "\n",
        "\n",
        "#### 2.   Get general information about the data:\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_OM7Gcip50oE"
      },
      "source": [
        "So far, we have read the csv file into a variable we called **`df`**: it is a pandas `DataFrame` object that contains the information in the csv file, along with many useful attributes and methods. For isntance, let's print the first few entries along with the column names, to get a quick feeling about the data:\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 360
        },
        "id": "OyU38RdluTuA",
        "outputId": "f28731ba-2417-4d23-b7dc-e749081c1b4f"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 25,
      "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>Make</th>\n",
              "      <th>Model</th>\n",
              "      <th>Year</th>\n",
              "      <th>Mileage</th>\n",
              "      <th>Price</th>\n",
              "      <th>Body Style</th>\n",
              "      <th>Ex Color</th>\n",
              "      <th>In Color</th>\n",
              "      <th>Engine</th>\n",
              "      <th>Transmission</th>\n",
              "      <th>Doors</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Toyota</td>\n",
              "      <td>Avalon XLE</td>\n",
              "      <td>2014</td>\n",
              "      <td>4725</td>\n",
              "      <td>28995</td>\n",
              "      <td>Sedan</td>\n",
              "      <td>Gray</td>\n",
              "      <td>Black</td>\n",
              "      <td>6 Cyl</td>\n",
              "      <td>Automatic</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Ford</td>\n",
              "      <td>Escape SE</td>\n",
              "      <td>2014</td>\n",
              "      <td>17201</td>\n",
              "      <td>24994</td>\n",
              "      <td>SUV</td>\n",
              "      <td>Silver</td>\n",
              "      <td>Black</td>\n",
              "      <td>4 Cyl</td>\n",
              "      <td>Automatic</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Hyundai</td>\n",
              "      <td>Santa Fe Sport</td>\n",
              "      <td>2014</td>\n",
              "      <td>6279</td>\n",
              "      <td>25998</td>\n",
              "      <td>SUV</td>\n",
              "      <td>Silver</td>\n",
              "      <td>Gray</td>\n",
              "      <td>4 Cyl</td>\n",
              "      <td>Automatic</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Ford</td>\n",
              "      <td>Escape SE</td>\n",
              "      <td>2014</td>\n",
              "      <td>16262</td>\n",
              "      <td>26684</td>\n",
              "      <td>SUV</td>\n",
              "      <td>Black</td>\n",
              "      <td>Black</td>\n",
              "      <td>4 Cyl</td>\n",
              "      <td>Automatic</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Jeep</td>\n",
              "      <td>Grand Cherokee Limited Edition</td>\n",
              "      <td>2014</td>\n",
              "      <td>35572</td>\n",
              "      <td>33440</td>\n",
              "      <td>SUV</td>\n",
              "      <td>Black</td>\n",
              "      <td>Black</td>\n",
              "      <td>6 Cyl</td>\n",
              "      <td>Automatic</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      Make                           Model  Year  ...  Engine  Transmission Doors\n",
              "0   Toyota                      Avalon XLE  2014  ...   6 Cyl     Automatic     4\n",
              "1     Ford                       Escape SE  2014  ...   4 Cyl     Automatic     4\n",
              "2  Hyundai                  Santa Fe Sport  2014  ...   4 Cyl     Automatic     4\n",
              "3     Ford                       Escape SE  2014  ...   4 Cyl     Automatic     4\n",
              "4     Jeep  Grand Cherokee Limited Edition  2014  ...   6 Cyl     Automatic     4\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "14A7tQYoV8PH"
      },
      "source": [
        "We still don't know if the data-types are read correctly, we didn't specify them. The `info` method gives us a summary information: "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cOKXPWk3N6sm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6b3a13b7-d534-438e-d88e-04f7722be01f"
      },
      "source": [
        "df.info()"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 9080 entries, 0 to 9079\n",
            "Data columns (total 11 columns):\n",
            " #   Column        Non-Null Count  Dtype \n",
            "---  ------        --------------  ----- \n",
            " 0   Make          9080 non-null   object\n",
            " 1   Model         9080 non-null   object\n",
            " 2   Year          9080 non-null   int64 \n",
            " 3   Mileage       9080 non-null   int64 \n",
            " 4   Price         9080 non-null   int64 \n",
            " 5   Body Style    9080 non-null   object\n",
            " 6   Ex Color      9080 non-null   object\n",
            " 7   In Color      9080 non-null   object\n",
            " 8   Engine        9080 non-null   object\n",
            " 9   Transmission  9080 non-null   object\n",
            " 10  Doors         9080 non-null   int64 \n",
            "dtypes: int64(4), object(7)\n",
            "memory usage: 780.4+ KB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C5LylI4xWT9N"
      },
      "source": [
        "You should notice that the data-types are not exactly correct. For instance, the type for the `Doors` attribute (column) is infered as numeric (64-bit integer), but this is wrong (why?). \n",
        "\n",
        "So let's fix them! "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vZjVv-6mXrAe"
      },
      "source": [
        "df = df.astype({'Make':'category', 'Model':'category', 'Year':'category', \n",
        "                'Mileage':'int32', 'Price':'int64', 'Body Style':'category', \n",
        "                'Ex Color':'category' , 'In Color':'category', \n",
        "                'Engine':'category', 'Transmission':'category', 'Doors':'category'})"
      ],
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tRqHtd0DFZKz"
      },
      "source": [
        "Now, use the `info` method again to check if it it had the desied effect:\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CivnwTo2ZtME",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "612d0592-14d2-4f26-9d2d-6e7e6555fa0d"
      },
      "source": [
        "df.info()"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 9080 entries, 0 to 9079\n",
            "Data columns (total 11 columns):\n",
            " #   Column        Non-Null Count  Dtype   \n",
            "---  ------        --------------  -----   \n",
            " 0   Make          9080 non-null   category\n",
            " 1   Model         9080 non-null   category\n",
            " 2   Year          9080 non-null   category\n",
            " 3   Mileage       9080 non-null   int32   \n",
            " 4   Price         9080 non-null   int64   \n",
            " 5   Body Style    9080 non-null   category\n",
            " 6   Ex Color      9080 non-null   category\n",
            " 7   In Color      9080 non-null   category\n",
            " 8   Engine        9080 non-null   category\n",
            " 9   Transmission  9080 non-null   category\n",
            " 10  Doors         9080 non-null   category\n",
            "dtypes: category(9), int32(1), int64(1)\n",
            "memory usage: 248.0 KB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jeNSjWhdcFu6"
      },
      "source": [
        "---\n",
        "> **Q2:** You should also notice a change between the reported `memory ussage` (the last line reported by the `info` method). What is the change? How do you explain the change?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "id8BF8wvb9g8"
      },
      "source": [
        "> **A2:** Memory has been descresed from 780.4 KB to 248 KB due to columns dtype which we chnaged. \n",
        "---\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HG2CamjdPfFs"
      },
      "source": [
        "\n",
        "\n",
        "### 3. Basic exploratory questions using Python\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8yo2wMMFdINB"
      },
      "source": [
        "---\n",
        "> **Q3:** Use the provided hints to answer the questions about exploring your dataset."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LIDYDNj7TJdd"
      },
      "source": [
        "> a. How many instances does the dataset have?  **`9080`**\r\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PHyGIQ2OTwfl",
        "outputId": "d6ad5bf9-4af9-496f-d774-25f646940722"
      },
      "source": [
        "# hint: interpret the output of the following code:\n",
        "print(df.shape)\n"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(9080, 11)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IGbvaybkUIUR"
      },
      "source": [
        "> b. How many attributes? \r\n",
        "**`11`**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Lq38Z5FVRQ5i",
        "outputId": "966b4bca-1596-4d79-b4af-46dd09556ce9"
      },
      "source": [
        "# again, you can answer the question by interpretting the output of the same code:\n",
        "print(df.shape)"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(9080, 11)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8SPh2orQW8KG"
      },
      "source": [
        "> c. What are the attributes? \r\n",
        "```\r\n",
        " 'Make', 'Model', 'Year', 'Mileage', 'Price', 'Body Style', 'Ex Color',\r\n",
        "       'In Color', 'Engine', 'Transmission', 'Doors'\r\n",
        "```\r\n",
        "\r\n",
        "\r\n",
        "\r\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BzDyp-DvXDiw",
        "outputId": "2143802f-04df-444e-c0d0-f894316b82c7"
      },
      "source": [
        "# a pandas dataframe has an an attribute called columns:\n",
        "print(df.columns)"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Index(['Make', 'Model', 'Year', 'Mileage', 'Price', 'Body Style', 'Ex Color',\n",
            "       'In Color', 'Engine', 'Transmission', 'Doors'],\n",
            "      dtype='object')\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5CvTYxKMXQ_5"
      },
      "source": [
        "> d. What are the possible values for **Body Style** & **External Color**?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-bIdyCeWXhFG",
        "outputId": "665a9655-22e7-436d-faa7-a3d64231f996"
      },
      "source": [
        "# the \"unique\" method helps: \n",
        "print('Possible body styles:')\n",
        "print(df['Body Style'].unique())\n",
        "\n",
        "# now your turn, for external colour:\n",
        "print('\\nPossible external colours:')\n",
        "print(df['Ex Color'].unique())"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Possible body styles:\n",
            "['Sedan', 'SUV', 'Hatchback', 'Wagon', 'Coupe', 'Minivan', 'Van', 'Pickup', 'Convertible', 'Chassis']\n",
            "Categories (10, object): ['Sedan', 'SUV', 'Hatchback', 'Wagon', ..., 'Van', 'Pickup', 'Convertible',\n",
            "                          'Chassis']\n",
            "\n",
            "Possible external colours:\n",
            "['Gray', 'Silver', 'Black', 'White', 'Red', ..., 'Orange', 'Beige', 'Yellow', 'Bronze', 'Pink']\n",
            "Length: 15\n",
            "Categories (15, object): ['Gray', 'Silver', 'Black', 'White', ..., 'Beige', 'Yellow', 'Bronze',\n",
            "                          'Pink']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nNED-99zZf6w"
      },
      "source": [
        "> e. What is the *minimum*, *maximum*, *average* and *median* price?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NRvSaH7IZppP",
        "outputId": "ab119414-0f84-432b-b4bc-86216e8229a1"
      },
      "source": [
        "# these are easy, to get the minimum:\n",
        "print('min = {}'.format(df['Price'].min()))\n",
        "# now, you do the rest:\n",
        "print('max = {}'.format(df['Price'].max()))\n",
        "print('mean = {}'.format(df['Price'].mean()))\n",
        "print('median = {}'.format(df['Price'].median()))"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "min = 1490\n",
            "max = 499898\n",
            "mean = 24410.92665198238\n",
            "median = 21500.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vm4fAH2RbHuX"
      },
      "source": [
        "> f. Why might the median price be different than the average price?  \r\n",
        "\r\n",
        "```\r\n",
        "The median is the middle point of a number set . The average  is the sum of all the numbers in the set divided by the amount of numbers```\r\n",
        "\r\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yk9PQPvIbQpl"
      },
      "source": [
        "> g. What is the most common year of car? `2011`\r\n",
        "\r\n",
        "\r\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "b2mcEe-hbrpc",
        "outputId": "889acf9d-846c-41f3-b3c8-fbc577cb0e34"
      },
      "source": [
        "# hint: you can either use the 'mode()' method, \n",
        "# or the value_counts() along with idxmax()\n",
        "print(df['Year'].mode())\n",
        "print(df['Year'].value_counts().idxmax())\n"
      ],
      "execution_count": 51,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0    2011\n",
            "Name: Year, dtype: category\n",
            "Categories (24, int64): [1990, 1993, 1994, 1995, ..., 2012, 2013, 2014, 2015]\n",
            "2011\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2XLXYcq7cd9F"
      },
      "source": [
        "> f. What is the ratio of 2-door to 4-door cars? "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fZrDVln8bAVc",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fda7e1c3-9f33-4d2c-c299-2f15f0a64015"
      },
      "source": [
        "# Hint: you can use the output of the value_counts()...\n",
        "print(df['Doors'].value_counts())\n",
        "print('\\nThe ratio of 2-door to 4-door cars is: {} : {} '.format('766','8221'))"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "4    8221\n",
            "2     766\n",
            "3      93\n",
            "Name: Doors, dtype: int64\n",
            "\n",
            "The ratio of 2-door to 4-door cars is: 766 : 8221 \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BbxDhMO0dj6h"
      },
      "source": [
        "> g. What is the average price of a Honda car versus a Mercedes-Benz car? **Hint:** Try $\\texttt{df.loc}$."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T6yes3WGdvJ2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "993dc7c8-ee5c-48dd-8fe1-49b5a2037a30"
      },
      "source": [
        "# For Honda:\n",
        "print('Average price of a Honda car = {:.2f}'.format(df['Price'].mean()))\n",
        "# Now you do for Mercedes-Benz:\n",
        "print('Average price of a Mercedes-Benz car = {:.2f}'.format(df.loc[df['Make'] == 'Mercedes-Benz']['Price'].mean()))\n"
      ],
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Average price of a Honda car = 24410.93\n",
            "Average price of a Mercedes-Benz car = 37620.97\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FWoLpa4kld5u"
      },
      "source": [
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZViWoH9Pk2g3"
      },
      "source": [
        "### 4. Basic data visualisation using Python\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Qn_a3nqYXH8N"
      },
      "source": [
        "---\n",
        "> **Q4:** Execute the following command and interpret what the plot is showing. In particular, describe the general trend between `milage` and `price`, as well as `milage` and `year`."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xztMxUgTl6Vx"
      },
      "source": [
        "\n",
        "\n",
        "The pairs plot builds on two basic figures, the histogram and the scatter plot. The histogram on the diagonal allows us to see the distribution of a single variable while the scatter plots on the upper and lower triangles show the relationship (or lack thereof) between two variables.\n",
        "\n",
        "from the scatter plots , we can able to see the less mileage will be having high price and vice-versa\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        },
        "id": "h9MF9IA2mBAE",
        "outputId": "8f746b22-872e-445e-bcf5-52d839efa298"
      },
      "source": [
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "sns.pairplot(df)"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<seaborn.axisgrid.PairGrid at 0x7fee6f431b00>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 6 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}