Peppi-Lotta_Saari_DAKD_2022_exercise_1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-block alert-warning\">\n",
    "    <h1><center> DAKD 2022 EXERCISE 1: DATA UNDERSTANDING  </center></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This exercise relates to the _data understanding_ and  _data preparation_ stages of the Crisp Data Mining (CRISP-DM) model presented on the course. The questions at this stage of a data-analysis project are for example: \n",
    "\n",
    "- Is the data quality sufficient?\n",
    "- How can we check the data for problems?\n",
    "- How do we have to clean the data?\n",
    "- How is the data best transformed for modeling?\n",
    "\n",
    "It may be tempting to just run a model on data without checking it. However, not doing basic checks can ruin your whole analysis and make your results invalid as well as mislead you in further analyses. There is no excuse for not plotting and checking that the data is as we expect and clean. In this exercise we do just that, check the validity of data and familiarize ourselves with a dataset, also discussing preprocessing and multi-dimensional plotting. \n",
    " \n",
    "------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <font color = red> *** FILL YOUR INFORMATION BELOW *** </font>\n",
    "Peppi-Lotta Saari <br>\n",
    "517334 <br>\n",
    "plsaar@utu.fi  <br>\n",
    "14.11.2022  <br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "W3jIT0S_3BPb"
   },
   "source": [
    "\n",
    "#### General guidance for exercises \n",
    "\n",
    "-  You can add more code and markup cells, as long as the flow of the notebook stays readable and logical.\n",
    "- Answer **all** questions (except the bonus if you do not want to attempt it), even if you can't get your script to fully work\n",
    "- Write clear and easily readable code, include explanations of what your code does\n",
    "- Make informative illustrations: include labels for x and y axes, legends and captions for your plots\n",
    "- Before saving the ipynb file (and possible printing) run: \"Restart & Run all\", to make sure you return a file that works as expected.\n",
    "- Grading: *Fail*/*Pass*/*Pass with honors* (+1)\n",
    "- If you encounter problems, Google first. If you can't find an answer to the problem, don't hesitate to ask in the Moodle discussion or directly via moodle chat.\n",
    "- Don't leave it to the last moment! No feedback service during weekends."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2I2WLapM3BPc"
   },
   "source": [
    "### <font color = red> Packages needed for this exercise: </font>\n",
    "- The exercise can be done without importing any extra packages, but you can import new ones but bear in mind that if you are importing many new packages, you may be complicating your answer. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "id": "kNsRA1WF3BPc"
   },
   "outputs": [],
   "source": [
    "# --- Libraries with a short description ---\n",
    "import pandas as pd # for data manipulation\n",
    "import matplotlib.pyplot as plt # for plotting\n",
    "import numpy as np #for numeric calculations and making simulated data.\n",
    "import seaborn as sns # for plotting, an extension on matplotlib \n",
    "\n",
    "# - sklearn has many data analysis utility functions like scaling as well as a large variety of modeling tools.\n",
    "from sklearn.decomposition import PCA \n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import minmax_scale\n",
    "from sklearn.preprocessing import scale\n",
    "from sklearn.manifold import TSNE\n",
    "\n",
    "# This forces plots to be shown inline in the notebook\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "<div class=\"alert alert-block alert-warning\">\n",
    "    <h1><center> PLOTTING TUTORIAL </center></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This small explanation of the matplotlib package aims to avoid confusion and help you avoid common mistakes and frustration. Matplotlib is an object-oriented plotting package with the benefit of giving the user a lot of control. The downside is that it can be confusing to new users. **If you are having problems with the plotting exercises, return to this tutorial as it explains the needed concepts to do the exercises!**\n",
    "\n",
    "-----------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Figure and axes \n",
    "\n",
    "\n",
    "All plots in matplotlib are structured with the **<font color = dimgrey> figure </font>** and **<font color = blue> axes </font>** objects. \n",
    "\n",
    "- The **<font color = dimgrey> figure </font>** object is a container for all plotting elements (in other words, everything we see).  \n",
    "- A figure can have many **<font color = blue> axes </font>**. They are the objects you plot on to. The axes can be anywhere inside the figure and can even overlap. Position of axes is defined relative to the figure. \n",
    "\n",
    "The **<font color = blue> axes </font>** objects have the methods you will use to define most of your plots. For example axes.hist() is used to draw a histogram and axes.set_title() to give one axes a title. The name of the object can be a bit confusing as it does not refer to the axes in the way \"x-axis\" does but to the container of a single plot.\n",
    "\n",
    "\n",
    "--------------\n",
    "     \n",
    "- Below is an example that illustrates how **<font color = dimgrey> figures </font>**and **<font color = blue> axes </font>** work together in matplotlib. The comments explain what is done in every row of code. <font color = green> You are encouraged to play around with it, but its not required in terms of the exercise </font>. Below, we will create all figures and axes separately, but later on we will use a quicker way to do so. \n",
    "\n",
    " This is not yet a part of the exercises themselves and you do not need to change anything ! "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "#  --- Lets make some example data. ---\n",
    "x_example_data = np.linspace(0,5,10)\n",
    "y_example_data = x_example_data**2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(3, 5, 'This is a text object relative to the inner axes')"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Create a figure ###\n",
    "example_figure = plt.figure(figsize =(5,5)) #you give the size of the figure as a tuple of inches\n",
    "\n",
    "### Create an axes separately and add it to the figure ###\n",
    "example_axes_outer = example_figure.add_axes([0.1, 0.1, 0.9, 0.9]) #the argument gives the relative location of the axes in percentage from the corners of the figure. The order is left, bottom, right, top. \n",
    "\n",
    "### Set labels and titles to the axes ###\n",
    "example_axes_outer.set_xlabel(\"This is how you set an x-axis label to an axes\")\n",
    "example_axes_outer.set_ylabel(\"The y-label of an axes is set like this\")\n",
    "example_axes_outer.set_title(\"We learned how to give an axes a title!\")\n",
    "example_axes_inner = example_figure.add_axes([0.45, 0.45, 0.4, 0.3])\n",
    "example_axes_inner.set_title(\"This axes has a title too\")\n",
    "\n",
    "### Add something to the axes ###\n",
    "example_axes_inner.scatter(x_example_data, y_example_data)\n",
    "\n",
    "# Multiple things, like lines can be plotted on same axis. \n",
    "example_axes_outer.plot(x_example_data**4, y_example_data**2)\n",
    "example_axes_outer.plot(x_example_data**7, y_example_data**2)\n",
    "\n",
    "# If you want to add other objects, you add them to axes too, like text\n",
    "# Now you specify the location relative to the parent axes\n",
    "example_axes_inner.text(3, 5, \"This is a text object relative to the inner axes\")\n",
    "\n",
    "#Many more things can be added to axes in a similar way, not just text. \n",
    "#For more information there are many good tutorials available for example in youtube. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Subplots: creating multiple axes and placing them in a grid on the figure \n",
    "An established convention of using matplotlib is to start plotting by calling the **<font color = blue> plt.subplots </font>** function, which automatically creates a figure and a determined number of axes in a grid inside it and automatically links the axes to the figure. Even when creating just one axes this is a often used way to start making a plot.\n",
    "\n",
    "The most important arguments to **<font color = blue> plt.subplots </font>** are **nrows**, **ncols**, **figsize**, **sharex** and **sharey**\n",
    "- **nrows** controls how many subplots there will be in the grid by row, **ncol** controls the number of columns\n",
    "- **figsize** is a tuple e.g (1,5) which controls the size of the **<font color = dimgrey> figure </font>**, first is width and then height.\n",
    "- sharex (True, False) tells matplotlib whether all axes in the grid should have same x-axis scale and ticks, sharey does the same for all y.\n",
    "\n",
    "--------\n",
    "Below an example on creating subplots is presented. There is also a template-like example on how to fill the subplots in a loop using the  **enumerate** function of python for indexing into the subplots. The function **enumerate()** will give you an additional int indexer over the object you are looping over. This indexer can be used to loop over the different subplot elements like the axes for each of the subplots.\n",
    "\n",
    "**<font color = dimgrey> plt.tight_layout() </font>** is also a good command to know with subplots. It attempts to automatically arrange the different axes in a pretty way. It should be called after the plot is finished. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ----- Create some random data for the example, 3 continuous numeric features and 3 binary -----\n",
    "#dont worry about understanding the function, it creates lists and is shorthand for a for loop called list comprehension.\n",
    "numeric_datas = [np.random.rand(10,2) for _ in range(0,3)] #this creates list of lists of linear data, using list comprehension\n",
    "binary_datas = [(np.unique(np.random.randint(0, 2, size= 10), return_counts = True)[1]) for _ in range(0,3)] # create list of lists of samples of 0,1 like (co"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x500 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create figure with six axes in a 2*3 grid and set up titles --------------------------------------------------------\n",
    "fig, axes = plt.subplots(2,3, figsize = (10,5)) # now axes have indexes like axes[i, j]\n",
    "numeric_plot_titles = ['scatter_plot_1', 'a second plot', 'yet a third plot' ]#some titles for the different axes\n",
    "binary_plot_titles = ['coin_tosses1', 'tossing again', 'still tossing' ]#some titles for the different axes\n",
    "\n",
    "\n",
    "# Enumerate the index into the axes, fill the first 3 columns of first row with scatterplots of numeric_datas --------\n",
    "i = 0 # for indexing to the row of the axes [**i**, j]\n",
    "for j, numeric_data in enumerate(numeric_datas): # j = [0,1, ... n_datasets] for filling the columns, i stays constant as its the row\n",
    "    axes[i, j].scatter(x = numeric_data[:, 0], y = numeric_data[:, 1]) #plots are called on the axes\n",
    "    axes[i, j].set_title(numeric_plot_titles[j]) #set a title for each axes\n",
    "plt.tight_layout()\n",
    "\n",
    "\n",
    "# Plot the binary data -----------------------------------------------------------------------------------------------\n",
    "i = 1 # second row\n",
    "for j, binary_data in enumerate(binary_datas): # j = [0,1, ... n_datasets] for filling the columns, i stays constant as its the row\n",
    "    axes[i, j].bar(x = [\"0\",\"1\"], height = binary_data) #make a barplot\n",
    "    axes[i, j].set_title(binary_plot_titles[j]) #set a title for each axes\n",
    "    axes[i, j].set_ylim((0,10)) # set the yaxis limits, set_xlim works the same way.  \n",
    "\n",
    "fig.suptitle(\"fig.suptitle gives the figure a title and axes.set_title the axes\")\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "####  <font color = maroon> Seaborn and matplotlib </font>\n",
    "- Finally, it is good to know, that the popular Seaborn plotting library is based on matplotlib, and was designed to be an extension of it and to be more user-friendly and faster to use.  \n",
    "\n",
    "- One tip in particular that might help new users with seaborn is that two kinnds of plotting functions: for figure-level and axes-level plots. Axes level plots can be put into subplots like matplotlib plots as you saw in the example above whereas figure-level plots are done completely with seaborn. (For more information on this see https://seaborn.pydata.org/tutorial/function_overview.html)\n",
    "\n",
    "- For axes-level plots, the matplotlib-axes object is usually given to the seaborn plotting function as an argument. There is an example below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Count'>"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(2)\n",
    "\n",
    "# make some data \n",
    "random_data_a = np.random.rand(30)\n",
    "random_data_b = np.random.rand(100)\n",
    "\n",
    "# print the data we are plotting\n",
    "sns.histplot(data = random_data_a, ax = axes[0]) # we make a seaborn plot and put it into one of the axes we created\n",
    "sns.histplot(data =  random_data_b, ax = axes[1]) # we make a seaborn plot and put it into one of the axes we created"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "<div class=\"alert alert-block alert-warning\">\n",
    "    <h1><center> START OF EXERCISES </center></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8AZs1Rdr3BPd"
   },
   "source": [
    "##  <font color = dimgrey> 1. Introduction to the dataset </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2Y4dstDz3BPe"
   },
   "source": [
    "The exercise data depicts data gathered on hospital patients with and without cardiovascular disease, depicted by the target variable **_cardio_**, a common disease associated with a buildup of fatty deposits inside the arteries (blood vessels) of the heart. \n",
    " \n",
    " -------\n",
    "As is often the case with data analysis projects, the features/variables have been retrieved from different sources: \n",
    "- doctors notes (texts)\n",
    "- examination variables that have come from a database containing lab results or taken during a doctors examination\n",
    "- self reported variables\n",
    "\n",
    "--------------\n",
    "The exercise data has the following columns/attributes:\n",
    "\n",
    "| Feature | Type | Explanation |\n",
    "| :- | :- | :-\n",
    "| age | numeric | The age of the patient in days \n",
    "| sex | binary | Male/Female\n",
    "| weight | numeric | Measured weight of the patient (cm)\n",
    "| height | numeric | Measured weight of the patient (kg)\n",
    "| ap_hi | numeric | Measured Systolic blood pressure\n",
    "| ap_lo | numeric | Measured Diastolic blood pressure\n",
    "| smoke | binary | A subjective feature based on asking the patient whether or not he/she smokes\n",
    "| active | binary |  A subjective feature based on asking the patient whether or not he/she exercises regularly\n",
    "| cholesterol | categorical | Cholesterol associated risk information evaluated by a doctor\n",
    "| cardio | binary | Whether or not the patient has been diagnosed with cardiac disease."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue>weight and height have a typo/mistake here <br>\n",
    "Should be something like this</font><br>\n",
    "| Feature | Type | Explanation |\n",
    "| :- | :- | :-\n",
    "| weight | numeric | Measured weight of the patient (kg)\n",
    "| height | numeric | Measured height of the patient (cm)\n",
    "\n",
    "<font color = skyblue>male/female has 1/0 values. It is unclear which is which. </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jMatXBHE3BPh"
   },
   "source": [
    "-----------\n",
    "#### ***Reading data*** \n",
    "\n",
    "It is good practice to read the features in using their correct types instead of fixing them later. Below, there is ready-maed code for you to read in the data, using the data types and column names listed in the above table. Don't change the name of the variable, _data_. It is important in later exercises (for example in ex. 5e) that this is the name of the variable. <font color = red> If you have the dataset in the same folder as this notebook, the path already given to you should work. </font>\n",
    "\n",
    "---------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "# --- READ IN DATA (no need to change) --------\n",
    "data_path = \"cardiovascular_disease_data_ex_1.csv\" #if you just give the name of the file it will look for the data in the same folder as your script\n",
    "data = pd.read_csv(data_path, dtype = {'age': 'int', 'height': 'int', 'weight':'int', 'ap_lo':'int', 'ap_hi':'int', 'sex': 'boolean', 'smoke': 'boolean',\n",
    "       'active':'boolean', 'cardio':'boolean', 'cholesterol':'category'}) #the main data you use in this exercise should have this variable name, so that code given for you further on will run. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---------\n",
    "***Exercise 1 a)***\n",
    "1. First, print out the first five rows of the data.\n",
    "\n",
    "2. Then, save the feature names to lists by their types: make three lists named **numeric_features**, **binary_features** and **categorical_features**, containing the **names** of the features of each corresponding type (*you can think in terms of this exercise that binary variables can also be called booleans*).\n",
    "\n",
    "_Saving the feature names to a list or lists and using them to subset prevents typing errors and avoids repetition!_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>active</th>\n",
       "      <th>cardio</th>\n",
       "      <th>cholesterol</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>19797</td>\n",
       "      <td>False</td>\n",
       "      <td>161</td>\n",
       "      <td>55</td>\n",
       "      <td>102</td>\n",
       "      <td>68</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>elevated</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>22571</td>\n",
       "      <td>True</td>\n",
       "      <td>178</td>\n",
       "      <td>68</td>\n",
       "      <td>120</td>\n",
       "      <td>70</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>16621</td>\n",
       "      <td>True</td>\n",
       "      <td>169</td>\n",
       "      <td>69</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16688</td>\n",
       "      <td>False</td>\n",
       "      <td>156</td>\n",
       "      <td>77</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19498</td>\n",
       "      <td>True</td>\n",
       "      <td>170</td>\n",
       "      <td>98</td>\n",
       "      <td>130</td>\n",
       "      <td>80</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>elevated</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     age    sex  height  weight  ap_hi  ap_lo  smoke  active  cardio  \\\n",
       "0  19797  False     161      55    102     68  False    True   False   \n",
       "1  22571   True     178      68    120     70  False   False   False   \n",
       "2  16621   True     169      69    120     80  False    True   False   \n",
       "3  16688  False     156      77    120     80  False    True   False   \n",
       "4  19498   True     170      98    130     80   True    True    True   \n",
       "\n",
       "  cholesterol  \n",
       "0    elevated  \n",
       "1      normal  \n",
       "2      normal  \n",
       "3      normal  \n",
       "4    elevated  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['age', 'height', 'weight', 'ap_hi', 'ap_lo'] ['sex', 'smoke', 'active', 'cardio'] ['cholesterol']\n"
     ]
    }
   ],
   "source": [
    "#1\n",
    "display(data.head(5))\n",
    "\n",
    "# 2\n",
    "# save names in list \n",
    "numeric_features = ['age', 'height', 'weight', 'ap_hi', 'ap_lo' ]\n",
    "binary_features = ['sex', 'smoke', 'active', 'cardio']\n",
    "categorical_features = ['cholesterol']\n",
    "\n",
    "#print list to see if they are correct\n",
    "print(numeric_features, binary_features, categorical_features)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_________\n",
    "## <font color = dimgrey> 2. Checking data quality\n",
    "    \n",
    "Often in data analysis projects the data has not been gathered exclusively for the data analysis only but originally for other reasons. Because of this, the features are most often not nicely formatted and may have mistakes. It might be tempting to just use the data as is with a model, but it is very important to first check the data for possible mistakes as they can make all the conclusions you make based on your analysis misleading. One good routine for checking data quality is to first calculate statistical descriptives and then to plot the features to check if the values are realistic. \n",
    "\n",
    "  \n",
    "-----------    \n",
    "    \n",
    "Some descriptive statistics don't really make sense for certain kinds of features. In pandas, like in many other packages, some functions work differently depending on the data type of a column. In the following exercise we will look at the data descriptive statistics as well as how the behavior can change when the data types are different. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------\n",
    "***2 a)***  Print out the data types of your dataset below.\n",
    "\n",
    "_Perhaps the most common data types in pandas (see https://pandas.pydata.org/docs/user_guide/basics.html#basics-dtypes) are **float**, **int**, **bool** and **category**._"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age               int32\n",
       "sex             boolean\n",
       "height            int32\n",
       "weight            int32\n",
       "ap_hi             int32\n",
       "ap_lo             int32\n",
       "smoke           boolean\n",
       "active          boolean\n",
       "cardio          boolean\n",
       "cholesterol    category\n",
       "dtype: object"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#2a\n",
    "data.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "--------\n",
    "***2 b)*** Use the **DataFrame.describe() method** in the cell below on your data.   \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>19455.504762</td>\n",
       "      <td>164.180952</td>\n",
       "      <td>73.895238</td>\n",
       "      <td>127.857143</td>\n",
       "      <td>81.814286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2429.010199</td>\n",
       "      <td>7.534648</td>\n",
       "      <td>14.612326</td>\n",
       "      <td>17.508947</td>\n",
       "      <td>9.947652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>14367.000000</td>\n",
       "      <td>142.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>50.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>17635.750000</td>\n",
       "      <td>158.000000</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>80.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>19778.000000</td>\n",
       "      <td>164.000000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>80.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>21230.500000</td>\n",
       "      <td>170.000000</td>\n",
       "      <td>81.000000</td>\n",
       "      <td>140.000000</td>\n",
       "      <td>90.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>23565.000000</td>\n",
       "      <td>195.000000</td>\n",
       "      <td>125.000000</td>\n",
       "      <td>190.000000</td>\n",
       "      <td>120.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                age      height      weight       ap_hi       ap_lo\n",
       "count    210.000000  210.000000  210.000000  210.000000  210.000000\n",
       "mean   19455.504762  164.180952   73.895238  127.857143   81.814286\n",
       "std     2429.010199    7.534648   14.612326   17.508947    9.947652\n",
       "min    14367.000000  142.000000   45.000000   90.000000   50.000000\n",
       "25%    17635.750000  158.000000   64.000000  120.000000   80.000000\n",
       "50%    19778.000000  164.000000   70.000000  120.000000   80.000000\n",
       "75%    21230.500000  170.000000   81.000000  140.000000   90.000000\n",
       "max    23565.000000  195.000000  125.000000  190.000000  120.000000"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "--------\n",
    "***2 c)*** Did you get all of the features statistics or not? What do you think happened? "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> We don't get all features as describe-method only handles numeric values  </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------\n",
    "***2 d)*** Calculate descriptives for the binary (boolean) features and the categorical feature <br>\n",
    "\n",
    "_tip: in python, same type data structures can in many cases be concatenated using the + operator. If youre using the lists of names you created to subset, you can concatenate the two lists of feature names and use the resulting list to help you subset the dataframe_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_446bb\">\n",
       "  <caption>sex</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_446bb_level0_col0\" class=\"col_heading level0 col0\" >Frequency</th>\n",
       "      <th id=\"T_446bb_level0_col1\" class=\"col_heading level0 col1\" >Relative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_446bb_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_446bb_row0_col0\" class=\"data row0 col0\" >129</td>\n",
       "      <td id=\"T_446bb_row0_col1\" class=\"data row0 col1\" >0.614286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_446bb_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_446bb_row1_col0\" class=\"data row1 col0\" >81</td>\n",
       "      <td id=\"T_446bb_row1_col1\" class=\"data row1 col1\" >0.385714</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x233861b8a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_164bf\">\n",
       "  <caption>smoke</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_164bf_level0_col0\" class=\"col_heading level0 col0\" >Frequency</th>\n",
       "      <th id=\"T_164bf_level0_col1\" class=\"col_heading level0 col1\" >Relative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_164bf_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_164bf_row0_col0\" class=\"data row0 col0\" >186</td>\n",
       "      <td id=\"T_164bf_row0_col1\" class=\"data row0 col1\" >0.885714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_164bf_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_164bf_row1_col0\" class=\"data row1 col0\" >24</td>\n",
       "      <td id=\"T_164bf_row1_col1\" class=\"data row1 col1\" >0.114286</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x233861b8a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_c7d30\">\n",
       "  <caption>active</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_c7d30_level0_col0\" class=\"col_heading level0 col0\" >Frequency</th>\n",
       "      <th id=\"T_c7d30_level0_col1\" class=\"col_heading level0 col1\" >Relative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_c7d30_level0_row0\" class=\"row_heading level0 row0\" >True</th>\n",
       "      <td id=\"T_c7d30_row0_col0\" class=\"data row0 col0\" >162</td>\n",
       "      <td id=\"T_c7d30_row0_col1\" class=\"data row0 col1\" >0.771429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c7d30_level0_row1\" class=\"row_heading level0 row1\" >False</th>\n",
       "      <td id=\"T_c7d30_row1_col0\" class=\"data row1 col0\" >48</td>\n",
       "      <td id=\"T_c7d30_row1_col1\" class=\"data row1 col1\" >0.228571</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x233861b8a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_9d3c4\">\n",
       "  <caption>cardio</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_9d3c4_level0_col0\" class=\"col_heading level0 col0\" >Frequency</th>\n",
       "      <th id=\"T_9d3c4_level0_col1\" class=\"col_heading level0 col1\" >Relative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_9d3c4_level0_row0\" class=\"row_heading level0 row0\" >False</th>\n",
       "      <td id=\"T_9d3c4_row0_col0\" class=\"data row0 col0\" >105</td>\n",
       "      <td id=\"T_9d3c4_row0_col1\" class=\"data row0 col1\" >0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9d3c4_level0_row1\" class=\"row_heading level0 row1\" >True</th>\n",
       "      <td id=\"T_9d3c4_row1_col0\" class=\"data row1 col0\" >105</td>\n",
       "      <td id=\"T_9d3c4_row1_col1\" class=\"data row1 col1\" >0.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x233861b8a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_dd1eb\">\n",
       "  <caption>cholesterol</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_dd1eb_level0_col0\" class=\"col_heading level0 col0\" >Frequency</th>\n",
       "      <th id=\"T_dd1eb_level0_col1\" class=\"col_heading level0 col1\" >Relative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_dd1eb_level0_row0\" class=\"row_heading level0 row0\" >normal</th>\n",
       "      <td id=\"T_dd1eb_row0_col0\" class=\"data row0 col0\" >153</td>\n",
       "      <td id=\"T_dd1eb_row0_col1\" class=\"data row0 col1\" >0.728571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dd1eb_level0_row1\" class=\"row_heading level0 row1\" >at risk</th>\n",
       "      <td id=\"T_dd1eb_row1_col0\" class=\"data row1 col0\" >35</td>\n",
       "      <td id=\"T_dd1eb_row1_col1\" class=\"data row1 col1\" >0.166667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dd1eb_level0_row2\" class=\"row_heading level0 row2\" >elevated</th>\n",
       "      <td id=\"T_dd1eb_row2_col0\" class=\"data row2 col0\" >21</td>\n",
       "      <td id=\"T_dd1eb_row2_col1\" class=\"data row2 col1\" >0.100000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dd1eb_level0_row3\" class=\"row_heading level0 row3\" >elev ated</th>\n",
       "      <td id=\"T_dd1eb_row3_col0\" class=\"data row3 col0\" >1</td>\n",
       "      <td id=\"T_dd1eb_row3_col1\" class=\"data row3 col1\" >0.004762</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x233861b8a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#frequencies for categorial values \n",
    "categorial_data = data.select_dtypes(exclude='number')\n",
    "for c in categorial_data:\n",
    "        absolute = data[c].value_counts().to_frame()\n",
    "        relative = data[c].value_counts(normalize=True).to_frame()\n",
    "        result = pd.concat([absolute, relative], axis=1)\n",
    "        result.set_axis(['Frequency', 'Relative'], axis=1, inplace=True)\n",
    "        result = result.style.set_caption(c)\n",
    "        display(result)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------\n",
    "Now, we will see ***what would have happened if the data was read in using default settings*** and not giving information about the types of the features (dtypes), giving no arguments to pd.read_csv. \n",
    "\n",
    "Run the below cell (no need to modify the code) and look at the output of the cell with the wrongly read data. Compare it with the output of the cell where you used the correctly read data to get the descriptives. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>active</th>\n",
       "      <th>cardio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>210.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>19455.504762</td>\n",
       "      <td>0.385714</td>\n",
       "      <td>164.180952</td>\n",
       "      <td>73.895238</td>\n",
       "      <td>127.857143</td>\n",
       "      <td>81.814286</td>\n",
       "      <td>0.114286</td>\n",
       "      <td>0.771429</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2429.010199</td>\n",
       "      <td>0.487927</td>\n",
       "      <td>7.534648</td>\n",
       "      <td>14.612326</td>\n",
       "      <td>17.508947</td>\n",
       "      <td>9.947652</td>\n",
       "      <td>0.318918</td>\n",
       "      <td>0.420916</td>\n",
       "      <td>0.501195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>14367.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>142.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>17635.750000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>158.000000</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>19778.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>164.000000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>21230.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>170.000000</td>\n",
       "      <td>81.000000</td>\n",
       "      <td>140.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>23565.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>195.000000</td>\n",
       "      <td>125.000000</td>\n",
       "      <td>190.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                age         sex      height      weight       ap_hi  \\\n",
       "count    210.000000  210.000000  210.000000  210.000000  210.000000   \n",
       "mean   19455.504762    0.385714  164.180952   73.895238  127.857143   \n",
       "std     2429.010199    0.487927    7.534648   14.612326   17.508947   \n",
       "min    14367.000000    0.000000  142.000000   45.000000   90.000000   \n",
       "25%    17635.750000    0.000000  158.000000   64.000000  120.000000   \n",
       "50%    19778.000000    0.000000  164.000000   70.000000  120.000000   \n",
       "75%    21230.500000    1.000000  170.000000   81.000000  140.000000   \n",
       "max    23565.000000    1.000000  195.000000  125.000000  190.000000   \n",
       "\n",
       "            ap_lo       smoke      active      cardio  \n",
       "count  210.000000  210.000000  210.000000  210.000000  \n",
       "mean    81.814286    0.114286    0.771429    0.500000  \n",
       "std      9.947652    0.318918    0.420916    0.501195  \n",
       "min     50.000000    0.000000    0.000000    0.000000  \n",
       "25%     80.000000    0.000000    1.000000    0.000000  \n",
       "50%     80.000000    0.000000    1.000000    0.500000  \n",
       "75%     90.000000    0.000000    1.000000    1.000000  \n",
       "max    120.000000    1.000000    1.000000    1.000000  "
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read in the dataset with no arguments\n",
    "wrongly_read_data = pd.read_csv(data_path)\n",
    "\n",
    "# calculate descriptives for the data that was wrongly read in. \n",
    "wrongly_read_data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age             int64\n",
       "sex             int64\n",
       "height          int64\n",
       "weight          int64\n",
       "ap_hi           int64\n",
       "ap_lo           int64\n",
       "smoke           int64\n",
       "active          int64\n",
       "cardio          int64\n",
       "cholesterol    object\n",
       "dtype: object"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wrongly_read_data.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***2 e)*** \n",
    "- Looking at the above output, can you now say whats wrong with this presentation and why it was important to define the data types? "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> The binary and categorial data doesn't make any sense in this table. All binary data is typed as numeric and gategorial data is an object. This discription gives us info as they were all numeric and in case og boolean and categorial it does not make any sense.</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----------------------\n",
    "## 3. Plotting numeric features\n",
    "Descriptives don't really give a full or intuitive picture of the distribution of features. Next, we will make use of different plots to check the data quality.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------\n",
    "***3 a)*** \n",
    "\n",
    "Plot the numeric features as histograms (see tutorial if you need help).\n",
    "\n",
    "_tip: if you give only one grid-size argument for plt.subplots() like plt.subplots(3) the grid will be one-dimensional and you can index it with only one indexer._  \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "quantitative_data = data.select_dtypes(include='number')\n",
    "#histograms for quantitativa values\n",
    "for q in quantitative_data:\n",
    "    quantitative_data[q].plot(kind=\"hist\", color=\"pink\", legend=True)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_______\n",
    "## 4. Plotting binary and categorical features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***4 a)*** Plot **barplots** for each of the non-numeric features. **Use fractions, not the real frequencies of the levels of these features**.\n",
    "\n",
    "--------------\n",
    "\n",
    "_tip: For plotting, see documentation on axes.bar. To get the fractions, see the value_counts function and its optional argument normalize._\n",
    "\n",
    "_If you read in the dtypes to be pandas dtype.boolean, in some cases its easier to work with other packages, suchs as matplotlib when they are represented as numbers [0,1] and not True or False. If you get errors you can try to cast them momentarily to be int or float with astype. This does not mean that you've done the exercise incorrectly, just that you have to change them for the plotting package._ "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# bar plot for categorial variables\n",
    "for c in categorial_data:\n",
    "    categorial_data[c].value_counts(normalize=True).plot(kind='bar', color=\"pink\", legend=True)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**4 b)** \n",
    "\n",
    "Do you see something odd with one of the features? Fix it. \n",
    "\n",
    "_If you read the dtype of the categorical feature to be pandas dtype.categorical, **you have to also use the pandas function remove_categories to remove the category level from the feature**, even if you would have already removed the value. You can do this like: _data['feature_name'] = data['feature_name'].cat.remove_categories(\"category name to delete\")_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Cholesterol has values 'elevated' and 'elev ated' which are the same thing\n",
    "#This could also be seen in the frequency check in part 2d\n",
    "\n",
    "# choose cholesterol values 'elev ated' and set as 'elevated'\n",
    "data.loc[data['cholesterol'] == 'elev ated', 'cholesterol'] = 'elevated'\n",
    "\n",
    "#check if value 'elev ated' is still a category level \n",
    "# if it is we remove the level\n",
    "# (and if is not, we can still run this code over and over again without getting an error)\n",
    "if 'elev ated' in data['cholesterol'].cat.categories:\n",
    "    data['cholesterol'] = data['cholesterol'].cat.remove_categories(\"elev ated\")\n",
    "\n",
    "#bar blot to check\n",
    "data['cholesterol'].value_counts().plot(kind='bar', color=\"pink\", legend=True)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-------------\n",
    "\n",
    "## 5. Preprocessing numeric features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Scaling the data improves the performance of machine learning algorithms in many cases, or perhaps better put, can ruin performance if not done. For instance with distance based algorithms covered in the course such as PCA, T-SNE and KNN some features with large values can dominate the distance calculations. \n",
    "\n",
    "-----------\n",
    "We will look at two often used ways of bringing the values to the same scale: **min-max scaling to [0,1]** and **standardizing the features to 0 mean and unit variance**. We will see, that the decision has implications on how the data will look afterwards. Standardizing values is very common in statistics and min-max scaling is for example used in training neural networks, where we want the range to match the range of an activation function in the network. Its good to know both. \n",
    "\n",
    "Two functions, sklearn.minmax_scale and sklearn.scale have been imported for you and you can use them in the following exercises.\n",
    "__________________________\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5 a)** In this exercise we will use boxplots. Explain what a boxplot is and what information it gives you. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue> Box plot is distribution plot. It describes how the data is distributed. It shows the median, quntiles, minimum and maximun values and, outliers. </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5 b)** Min-max numeric attributes to [0,1] and **store the results in a new dataframe called data_min_maxed**. You might have to wrap the data to a dataframe again using pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.590346</td>\n",
       "      <td>0.358491</td>\n",
       "      <td>0.1250</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.257143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.891933</td>\n",
       "      <td>0.679245</td>\n",
       "      <td>0.2875</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.285714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.245053</td>\n",
       "      <td>0.509434</td>\n",
       "      <td>0.3000</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.428571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.252337</td>\n",
       "      <td>0.264151</td>\n",
       "      <td>0.4000</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.428571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.557839</td>\n",
       "      <td>0.528302</td>\n",
       "      <td>0.6625</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.428571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>205</th>\n",
       "      <td>0.246032</td>\n",
       "      <td>0.301887</td>\n",
       "      <td>0.1250</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.428571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>206</th>\n",
       "      <td>0.258208</td>\n",
       "      <td>0.528302</td>\n",
       "      <td>0.2875</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.285714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>207</th>\n",
       "      <td>0.951294</td>\n",
       "      <td>0.283019</td>\n",
       "      <td>0.4125</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.142857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>208</th>\n",
       "      <td>0.094477</td>\n",
       "      <td>0.207547</td>\n",
       "      <td>0.1250</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.428571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>209</th>\n",
       "      <td>0.399652</td>\n",
       "      <td>0.566038</td>\n",
       "      <td>0.4125</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.571429</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>210 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          age    height  weight  ap_hi     ap_lo\n",
       "0    0.590346  0.358491  0.1250   0.12  0.257143\n",
       "1    0.891933  0.679245  0.2875   0.30  0.285714\n",
       "2    0.245053  0.509434  0.3000   0.30  0.428571\n",
       "3    0.252337  0.264151  0.4000   0.30  0.428571\n",
       "4    0.557839  0.528302  0.6625   0.40  0.428571\n",
       "..        ...       ...     ...    ...       ...\n",
       "205  0.246032  0.301887  0.1250   0.30  0.428571\n",
       "206  0.258208  0.528302  0.2875   0.20  0.285714\n",
       "207  0.951294  0.283019  0.4125   0.10  0.142857\n",
       "208  0.094477  0.207547  0.1250   0.30  0.428571\n",
       "209  0.399652  0.566038  0.4125   0.50  0.571429\n",
       "\n",
       "[210 rows x 5 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#5b\n",
    "# https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.MinMaxScaler.html\n",
    "\n",
    "# minmax_scale\n",
    "data_min_maxed = minmax_scale(quantitative_data, feature_range=(0, 1))\n",
    "# make dataframe again\n",
    "data_min_maxed = pd.DataFrame(data_min_maxed)\n",
    "# insert column names back\n",
    "data_min_maxed.columns = numeric_features\n",
    "# display\n",
    "data_min_maxed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5 c)** Standardize numeric attributes to 0 mean and unit variance and **store the results in a new dataframe called data_standardized**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.140926</td>\n",
       "      <td>-0.423185</td>\n",
       "      <td>-1.296193</td>\n",
       "      <td>-1.480325</td>\n",
       "      <td>-1.392016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.285684</td>\n",
       "      <td>1.838449</td>\n",
       "      <td>-0.404407</td>\n",
       "      <td>-0.449822</td>\n",
       "      <td>-1.190484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.169727</td>\n",
       "      <td>0.641113</td>\n",
       "      <td>-0.335808</td>\n",
       "      <td>-0.449822</td>\n",
       "      <td>-0.182819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.142077</td>\n",
       "      <td>-1.088372</td>\n",
       "      <td>0.212983</td>\n",
       "      <td>-0.449822</td>\n",
       "      <td>-0.182819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.017537</td>\n",
       "      <td>0.774151</td>\n",
       "      <td>1.653560</td>\n",
       "      <td>0.122679</td>\n",
       "      <td>-0.182819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>205</th>\n",
       "      <td>-1.166013</td>\n",
       "      <td>-0.822297</td>\n",
       "      <td>-1.296193</td>\n",
       "      <td>-0.449822</td>\n",
       "      <td>-0.182819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>206</th>\n",
       "      <td>-1.119793</td>\n",
       "      <td>0.774151</td>\n",
       "      <td>-0.404407</td>\n",
       "      <td>-1.022324</td>\n",
       "      <td>-1.190484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>207</th>\n",
       "      <td>1.511004</td>\n",
       "      <td>-0.955335</td>\n",
       "      <td>0.281582</td>\n",
       "      <td>-1.594825</td>\n",
       "      <td>-2.198148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>208</th>\n",
       "      <td>-1.741280</td>\n",
       "      <td>-1.487484</td>\n",
       "      <td>-1.296193</td>\n",
       "      <td>-0.449822</td>\n",
       "      <td>-0.182819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>209</th>\n",
       "      <td>-0.582904</td>\n",
       "      <td>1.040225</td>\n",
       "      <td>0.281582</td>\n",
       "      <td>0.695180</td>\n",
       "      <td>0.824845</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>210 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          age    height    weight     ap_hi     ap_lo\n",
       "0    0.140926 -0.423185 -1.296193 -1.480325 -1.392016\n",
       "1    1.285684  1.838449 -0.404407 -0.449822 -1.190484\n",
       "2   -1.169727  0.641113 -0.335808 -0.449822 -0.182819\n",
       "3   -1.142077 -1.088372  0.212983 -0.449822 -0.182819\n",
       "4    0.017537  0.774151  1.653560  0.122679 -0.182819\n",
       "..        ...       ...       ...       ...       ...\n",
       "205 -1.166013 -0.822297 -1.296193 -0.449822 -0.182819\n",
       "206 -1.119793  0.774151 -0.404407 -1.022324 -1.190484\n",
       "207  1.511004 -0.955335  0.281582 -1.594825 -2.198148\n",
       "208 -1.741280 -1.487484 -1.296193 -0.449822 -0.182819\n",
       "209 -0.582904  1.040225  0.281582  0.695180  0.824845\n",
       "\n",
       "[210 rows x 5 columns]"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 5c\n",
    "# scale\n",
    "data_standardized = scale(quantitative_data)\n",
    "# make dataframe again\n",
    "data_standardized = pd.DataFrame(data_standardized)\n",
    "# insert column names back\n",
    "data_standardized.columns = numeric_features\n",
    "# display\n",
    "data_standardized"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**5 d)** Make two boxplots of the 'age' feature, one plot with the data_min_maxed and one with the data_standardized. Preferably put the plots side-by-side and give each titles. See the tutorial in the beginning for help. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'data_standardized'}>"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 5d\n",
    "\n",
    "fig, axes = plt.subplots(1,2)\n",
    "\n",
    "#set titles\n",
    "axes[0].title.set_text('data_min_maxed')\n",
    "axes[1].title.set_text('data_standardized')\n",
    "\n",
    "# plot\n",
    "sns.boxplot(data=data_min_maxed['age'], ax = axes[0], color='pink') # we make a seaborn plot and put it into one of the axes we created\n",
    "sns.boxplot(data=data_standardized['age'], ax = axes[1], color='pink') # we make a seaborn plot and put it into one of the axes we created"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---------\n",
    "\n",
    "Lets see the differences of these preprocessing methods through an example. We will add an \"outlier\" point (some point with a large value) to replace the \n",
    "last value in both data, then again minmax and standardize and plot. The code to add the value is given for you and you shouldn't change it. \n",
    "\n",
    "--------------------\n",
    "\n",
    "***5e) Do the following:***\n",
    "1. Take the data for the age feature (age_w_outlier) provided for you\n",
    "2. Make two variables, age_w_outlier_minmaxed, containing the min-maxed values of the age_w_outlier and\n",
    "3.  age_w_outlier_standardized containing the standardized values for the age_w_outlier\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Add an outlier, DONT CHANGE THIS CELL CODE, JUST RUN IT ###\n",
    "data_w_outlier = data.copy() #data should be the name of the variable where you have stored your data!\n",
    "data_w_outlier.loc[data.shape[0] -1 , 'age'] = 150 #change the last value of age to be 150\n",
    "age_w_outlier = data_w_outlier.age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 5e\n",
    "\n",
    "# create age_w_outlier_minmaxed, keep the names the same\n",
    "age_w_outlier_minmaxed = minmax_scale(age_w_outlier, feature_range=(0, 1))\n",
    "\n",
    "# create age_w_outlier_standardized\n",
    "age_w_outlier_standardized = scale(age_w_outlier)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Below there is pre-written code for you to plot the different cases. Run it. The code should run if you have named your features appropriately. Run the code.***"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Wrap in a dataframe that will have two features - the age feature without the outlier, and the age feature with it, min-maxed.\n",
    "minmaxed_datas = pd.DataFrame({\"minmaxed_age_no_outlier\" : data_min_maxed.age, \n",
    "              \"minmaxed_age_with_outlier\": age_w_outlier_minmaxed })\n",
    "\n",
    "# Wrap in a dataframe that will have two features - the age feature without the outlier, and the age feature with it, standardized. \n",
    "standardized_datas = pd.DataFrame({\"standardized_data_no_outlier\" : data_standardized.age, \n",
    "              \"standardized_data_w_outlier\": age_w_outlier_standardized })\n",
    "\n",
    "axes_minmaxed = minmaxed_datas[['minmaxed_age_no_outlier', 'minmaxed_age_with_outlier']].plot(kind='box', title='Minmax with and without outlier')\n",
    "axes_std = standardized_datas[['standardized_data_no_outlier', 'standardized_data_w_outlier']].plot(kind='box', title='Standardized with and without outlier')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------\n",
    "**5 f) Look at the output of the above cell and answer the following**: \n",
    "\n",
    "1. Can you notice a difference between the two cases (min-maxed and standardized)? \n",
    "2. Can you say something about the difference of the effect of min-maxing and standardization?\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"https://vitalflux.com/minmaxscaler-standardscaler-python-examples/#:~:text=The%20MinMaxscaler%20is%20a%20type,range%20from%20min%20to%20max.\">https://vitalflux.com/minmaxscaler-standardscaler-python-examples/#:~:text=The%20MinMaxscaler%20is%20a%20type,range%20from%20min%20to%20max</a><br>    \n",
    "\n",
    "<a href=\"https://www.simplypsychology.org/boxplots.html#:~:text=What%20is%20a%20box%20plot,(or%20percentiles)%20and%20averages.\">https://www.simplypsychology.org/boxplots.html#:~:text=What%20is%20a%20box%20plot,(or%20percentiles)%20and%20averages.</a>\n",
    "\n",
    "<font color = skyblue> \n",
    "    Min-max sets the minimun value as 0 and maximum value as 1 and scales rest of the values between 0 and 1 in relation to the minimum and maximum. We see in the minmaxed_age_with_outlier that most of the values are pushed closer to the value 1 when the outlier is taken to account and set as the minimum. The exsistance of outliers have a big impact on where the rest of the values are located. </font> <br><br>\n",
    "\n",
    "<font color = skyblue> \n",
    "    Standard scaling sets the min and max values as the top and bottom of the scale and place all the values in their place in that range. The standardized_data_no_outlier and standardized_data_w_outlier don't look as different from each other as minmaxed_age_no_outlier and minmaxed_age_with_outlier. Including outliers doesn't have as musch of and effect with standardization.  </font>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---------------\n",
    "## 6. Preprocessing categorical features\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can roughly divide categorical variables/features to two types:  ***nominal categorical***  and  ***ordinal categorical*** variables/features. Some cases are clear in terms of which of the two a feature falls into. For example nationality is not an ordered feature, but which grade in school someone is has a natural ordering. **One-hot encoding** was presented in the lectures and will be used in the following exercises with different learning methods.\n",
    "\n",
    "\n",
    "-----\n",
    "***Nominal categorical features need to be encoded***, because not encoding them implies that they have an order. For example, consider a dataset where you would have rows by different countries, encoded randomly with numbers, for ex. Finland = 1, Norway = 2 and so on. For some analyses and methods this would imply that Norway is somehow \"greater\" in value than Finland. For some algorithms, the implication would also be, that some of the countries would be \"closer\" to each other. \n",
    "\n",
    "------\n",
    "***Ordinal categorical features do not necessarily need to be encoded***, but there are cases where it can be wise. One case is that the categories are not even distance from each other, which is the case with the 'cholesterol' feature with the levels 'normal', 'elevated' and 'at risk'. Its not clear that these are equal in distance from each other. When unsure, it may also be better to one-hot encode, and a lot of packages do it for you behind the scenes. Here we decide to one-hot encode.  \n",
    "\n",
    "---------------------\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**6 a)** One-hot-encode the cholesterol-feature and add the one-hot features to the data. Give the new features meaningful names. Print the first rows of the resulting dataframe. \n",
    "\n",
    "_tip: pandas has a function for this, google!_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>height</th>\n",
       "      <th>weight</th>\n",
       "      <th>ap_hi</th>\n",
       "      <th>ap_lo</th>\n",
       "      <th>smoke</th>\n",
       "      <th>active</th>\n",
       "      <th>cardio</th>\n",
       "      <th>cholesterol_at risk</th>\n",
       "      <th>cholesterol_elev ated</th>\n",
       "      <th>cholesterol_elevated</th>\n",
       "      <th>cholesterol_normal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>19797</td>\n",
       "      <td>False</td>\n",
       "      <td>161</td>\n",
       "      <td>55</td>\n",
       "      <td>102</td>\n",
       "      <td>68</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>22571</td>\n",
       "      <td>True</td>\n",
       "      <td>178</td>\n",
       "      <td>68</td>\n",
       "      <td>120</td>\n",
       "      <td>70</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>16621</td>\n",
       "      <td>True</td>\n",
       "      <td>169</td>\n",
       "      <td>69</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16688</td>\n",
       "      <td>False</td>\n",
       "      <td>156</td>\n",
       "      <td>77</td>\n",
       "      <td>120</td>\n",
       "      <td>80</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19498</td>\n",
       "      <td>True</td>\n",
       "      <td>170</td>\n",
       "      <td>98</td>\n",
       "      <td>130</td>\n",
       "      <td>80</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     age    sex  height  weight  ap_hi  ap_lo  smoke  active  cardio  \\\n",
       "0  19797  False     161      55    102     68  False    True   False   \n",
       "1  22571   True     178      68    120     70  False   False   False   \n",
       "2  16621   True     169      69    120     80  False    True   False   \n",
       "3  16688  False     156      77    120     80  False    True   False   \n",
       "4  19498   True     170      98    130     80   True    True    True   \n",
       "\n",
       "   cholesterol_at risk  cholesterol_elev ated  cholesterol_elevated  \\\n",
       "0                    0                      0                     1   \n",
       "1                    0                      0                     0   \n",
       "2                    0                      0                     0   \n",
       "3                    0                      0                     0   \n",
       "4                    0                      0                     1   \n",
       "\n",
       "   cholesterol_normal  \n",
       "0                   0  \n",
       "1                   1  \n",
       "2                   1  \n",
       "3                   1  \n",
       "4                   0  "
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 6a\n",
    "# https://datagy.io/pandas-get-dummies/\n",
    "ohe = pd.get_dummies(\n",
    "    data=data,\n",
    "    columns=['cholesterol'])\n",
    "ohe.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-block alert-warning\">\n",
    "    <h1><center> BONUS EXERCISES </center></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Below are the bonus exercises. You can stop here, and get the \"pass\" grade. \n",
    "- By doing the bonus exercises below, you can get a \"pass with honors\", which means you will get one point bonus for the exam. \n",
    "\n",
    "The following exercises are more challenging and not as straight-forward and may require some research of your own. However, perfect written answers are not required, but answers that show that you have tried to understand the problems and explain them with your own words. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "bea2hNxp3BPv"
   },
   "source": [
    "____________\n",
    "##  <font color = dollargreen > 7. BONUS: Dimensionality reduction and plotting with PCA </font>\n",
    "In the lectures, PCA was introduced as a dimensionality reduction technique. Here we will use it to reduce the dimensionality of the numeric features of this dataset and use the resulting compressed view of the dataset to plot it. This means you have to, run PCA  and then project the data you used to fit the PCA to the new space, where the principal components are the axes.\n",
    "____________"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4mco9ZWp3BPv"
   },
   "source": [
    "**7 a)** Explain _in your own words_ what principal component analysis is.\n",
    "\n",
    "--------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = skyblue>  </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4mco9ZWp3BPv"
   },
   "source": [
    "-------------\n",
    "**7 b)** Do PCA with two components with and without z-score standardization **for the numeric features in the data**. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# --- Your for 7b) code here --- #"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-------------\n",
    "\n",
    "\n",
    "**7 c) Plot the data, projected on to the PCA space as a scatterplot, the x-axis being one component and y the other. **Add the total explained variance to your plot as an annotation**. See the documentation of the pca method on how to get the explained variance.\n",
    "\n",
    "- _Tip: It may be easier to try the seaborn scatterplot for this one. For help see documentation on how to do annotation see tutorial. The total explained variance is the sum of both the components explained variance_.\n",
    "\n",
    "- _Tip2_: Depending on how you approach annotating the plot, you might have to cast the feature name to be a string. One nice way to format values in python is the f - formatting string, which allows you to insert expressions inside strings (see example below):\n",
    "\n",
    "\n",
    "\n",
    "------\n",
    "name = Valtteri<br>\n",
    "print(f\"hello_{name}\")\n",
    "\n",
    "---------\n",
    "You can also set the number of wanted decimals for floats<br>\n",
    "For example f'{float_variable:.2f}' would result in 2 decimals making it to the string created\n",
    "\n",
    "----------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# --- Your code for 7c) --- you can make more cells if you like ---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " \n",
    "\n",
    "**7 d) Gather information for the next part of the exercise and print out the following things:**\n",
    "- First, the standard deviation of the original data features (not standardized, and with the numeric features only).\n",
    "- Second, the standard deviation of the standardized numeric features "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# --- Your code for 7d) here --- #\n",
    "\n",
    "print( \"+++\"*3, \"Standard deviations of numeric features, original data\", \"+++\" * 3)  \n",
    "#your print here\n",
    "\n",
    "print( \"+++\"*3, \"Standard deviations of numeric features, standardized data\", \"+++\" * 3)  \n",
    "#your print here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4mco9ZWp3BPv"
   },
   "source": [
    "----------\n",
    "**7 e) Look at the output above and the explained variance information you added as annotations to the plots. Try to think about the following questions and give a short answer of what you think has happened:** \n",
    "\n",
    "1. Where do you think the difference between the amounts of explained variance might come from? \n",
    "\n",
    "2. Can you say something about why it is important to scale the features for PCA by looking at the evidence youve gathered?\n",
    "\n",
    "__Answer in your own words, here it is not important to get the perfect answer but to try to think and figure out what has happened__\n",
    "\n",
    "------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = red> ***--- Answer 7e) here ---***  </font>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------------------\n",
    "\n",
    "## <font color = dollargreen > 8. Bonus: t-SNE and high dimensional data </font> "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another method that can be used to plot high-dimensional data introduced in the lectures was t-distributed Stochastic Neighbor Embedding (t-SNE)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***8 a)*** Shortly explain t-SNE **in your own words**. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = red> ***--- Answer 8 a) here ---***  </font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***8 b)*** Run t-SNE for both standardized and non standardized data (as you did with PCA)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# --- Your code for 8b) here --- #"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***8 c)*** Plot t-sne, similarly to PCA making the color of the points correspond to the levels of the cardio feature, but having only numerical features as a basis of the T-SNE.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# --- Code for 8 c) --- #"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***8 d)*** \n",
    "\n",
    "- What do you think might have happened between the two runs of t-SNE on unstandardized and standardized data? Why is it important to standardize before using the algorithm?\n",
    "\n",
    "_Here the aim is to think about this and learn, not come up with a perfect explanation. Googling is encouraged. Think about whether t-sne is a distance based algorithm or not?_ "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color = red> --- ***Answer 8d) here*** --- </font>"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "DADK_exercise_2_2021.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {
    "height": "180px",
    "width": "160px"
   },
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}