DT.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Is the weather good to play outside?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The folder `datasets` contains two files:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "- `weather.numeric.csv`:\n",
    "\n",
    "```\n",
    "temperature,humidity,windy,play\n",
    "85,85,0,no\n",
    "80,90,1,no\n",
    "83,86,0,yes\n",
    "70,96,0,yes\n",
    "68,80,0,yes\n",
    "65,70,1,no\n",
    "64,65,1,yes\n",
    "72,95,0,no\n",
    "69,70,0,yes\n",
    "75,80,0,yes\n",
    "75,70,1,yes\n",
    "72,90,1,yes\n",
    "81,75,0,yes\n",
    "71,91,1,no\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `weather.nominal.csv`:\n",
    "\n",
    "```\n",
    "outlook,temperature,humidity,windy,play\n",
    "sunny,hot,high,FALSE,no\n",
    "sunny,hot,high,TRUE,no\n",
    "overcast,hot,high,FALSE,yes\n",
    "rainy,mild,high,FALSE,yes\n",
    "rainy,cool,normal,FALSE,yes\n",
    "rainy,cool,normal,TRUE,no\n",
    "overcast,cool,normal,TRUE,yes\n",
    "sunny,mild,high,FALSE,no\n",
    "sunny,cool,normal,FALSE,yes\n",
    "rainy,mild,normal,FALSE,yes\n",
    "sunny,mild,normal,TRUE,yes\n",
    "overcast,mild,high,TRUE,yes\n",
    "overcast,hot,normal,FALSE,yes\n",
    "rainy,mild,high,TRUE,no\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use Decision Tress from the `scikit-learn` library to create accurate models, first for the numerical dataset, then for the nominal dataset.\n",
    "\n",
    "Explain your reasoning, and justify any choices of the hyperparameters (and/or run experiments to find the optimal ones).\n",
    "\n",
    "Use the provided datasets for training, and create testing datasets based on your experience.\n",
    "\n",
    "Evaluate your models, and use visualisation to show the trees and any relevant plots."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Marking scheme\n",
    "\n",
    "|Item|Mark|\n",
    "|:----|---:|\n",
    "|**Numerical dataset**:||\n",
    "|Explanation, Justification|/4|\n",
    "|DT model|/3|\n",
    "|Evaluation|/3|\n",
    "|**Nominal dataset**:||\n",
    "|Explanation, Justification|/4|\n",
    "|DT model|/3|\n",
    "|Evaluation|/3|\n",
    "|||\n",
    "|**Total**:     |/20|\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-10-25T11:37:52.790405Z",
     "start_time": "2022-10-25T11:37:50.972952Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-10-25T11:37:52.837819Z",
     "start_time": "2022-10-25T11:37:52.790405Z"
    }
   },
   "outputs": [],
   "source": [
    "data1 = pd.read_csv('datasets/weather.numeric.csv')\n",
    "data2 = pd.read_csv('datasets/weather.nominal.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-10-25T11:37:52.890705Z",
     "start_time": "2022-10-25T11:37:52.843562Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": "    temperature  humidity  windy play\n0            85        85      0   no\n1            80        90      1   no\n2            83        86      0  yes\n3            70        96      0  yes\n4            68        80      0  yes\n5            65        70      1   no\n6            64        65      1  yes\n7            72        95      0   no\n8            69        70      0  yes\n9            75        80      0  yes\n10           75        70      1  yes\n11           72        90      1  yes\n12           81        75      0  yes\n13           71        91      1   no",
      "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>temperature</th>\n      <th>humidity</th>\n      <th>windy</th>\n      <th>play</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>85</td>\n      <td>85</td>\n      <td>0</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>80</td>\n      <td>90</td>\n      <td>1</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>83</td>\n      <td>86</td>\n      <td>0</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>70</td>\n      <td>96</td>\n      <td>0</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>68</td>\n      <td>80</td>\n      <td>0</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>65</td>\n      <td>70</td>\n      <td>1</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>64</td>\n      <td>65</td>\n      <td>1</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>72</td>\n      <td>95</td>\n      <td>0</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>69</td>\n      <td>70</td>\n      <td>0</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>75</td>\n      <td>80</td>\n      <td>0</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>75</td>\n      <td>70</td>\n      <td>1</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>72</td>\n      <td>90</td>\n      <td>1</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>81</td>\n      <td>75</td>\n      <td>0</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>71</td>\n      <td>91</td>\n      <td>1</td>\n      <td>no</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2022-10-25T11:37:52.922669Z",
     "start_time": "2022-10-25T11:37:52.895684Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": "     outlook temperature humidity  windy play\n0      sunny         hot     high  False   no\n1      sunny         hot     high   True   no\n2   overcast         hot     high  False  yes\n3      rainy        mild     high  False  yes\n4      rainy        cool   normal  False  yes\n5      rainy        cool   normal   True   no\n6   overcast        cool   normal   True  yes\n7      sunny        mild     high  False   no\n8      sunny        cool   normal  False  yes\n9      rainy        mild   normal  False  yes\n10     sunny        mild   normal   True  yes\n11  overcast        mild     high   True  yes\n12  overcast         hot   normal  False  yes\n13     rainy        mild     high   True   no",
      "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>outlook</th>\n      <th>temperature</th>\n      <th>humidity</th>\n      <th>windy</th>\n      <th>play</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>sunny</td>\n      <td>hot</td>\n      <td>high</td>\n      <td>False</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>sunny</td>\n      <td>hot</td>\n      <td>high</td>\n      <td>True</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>overcast</td>\n      <td>hot</td>\n      <td>high</td>\n      <td>False</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>rainy</td>\n      <td>mild</td>\n      <td>high</td>\n      <td>False</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>rainy</td>\n      <td>cool</td>\n      <td>normal</td>\n      <td>False</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>rainy</td>\n      <td>cool</td>\n      <td>normal</td>\n      <td>True</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>overcast</td>\n      <td>cool</td>\n      <td>normal</td>\n      <td>True</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>sunny</td>\n      <td>mild</td>\n      <td>high</td>\n      <td>False</td>\n      <td>no</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>sunny</td>\n      <td>cool</td>\n      <td>normal</td>\n      <td>False</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>rainy</td>\n      <td>mild</td>\n      <td>normal</td>\n      <td>False</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>sunny</td>\n      <td>mild</td>\n      <td>normal</td>\n      <td>True</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>overcast</td>\n      <td>mild</td>\n      <td>high</td>\n      <td>True</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>overcast</td>\n      <td>hot</td>\n      <td>normal</td>\n      <td>False</td>\n      <td>yes</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>rainy</td>\n      <td>mild</td>\n      <td>high</td>\n      <td>True</td>\n      <td>no</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data2"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Numeric Dataset"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Explanation/Justification"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "The numeric dataset was a relatively straightforward implementation using Scikit-learn, as the inputs from the csv set are numeric integers and require no preprocessing. So in the below code, I simply set my column names and imported the data using panda (skipping the first row to allow for title row), and then seperated out the features into my X DataFrame and my \"play\" labels into my Y DataFrame. I then used Scikit-learn's \"train_test_split\" function to separate this data into train and test data (70% train, 30% test) and shuffled the data once. I opted for a 70/30 split as with such a small pool of data, we require a large proportion of training data while still allowing for tests (I tried with lower percentage of test data but this doesn't provide enough inputs to perform the test). I shuffle the data only once as the numeric data is already unordered in the csv file. I then simply create my classifier and fit the train data to it. Once training is completed, I use the test data to evaluate the success of my model and use Scikit-learn's metrics library to print out the model accuracy.\n",
    "\n",
    "Once all is done, I plot my tree using basic parameters (as we are working on a very simple problem) and use the graphviz package to save the output as a pdf (to have data available for external use). These pdfs are already included in the project file, however if you would like to save a new pdf to test with your own parameters, please download graphviz at the following link and add it to PATH (for Windows users): https://graphviz.org/download/"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### DT Model"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.4\n",
      "Tree PDF saved in project file\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.tree import export_graphviz\n",
    "from sklearn.tree import plot_tree\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import metrics\n",
    "from sklearn import preprocessing\n",
    "import graphviz\n",
    "\n",
    "columns = ['temperature', 'humidity', 'windy', 'play']\n",
    "numeric_data = pd.read_csv('datasets/weather.numeric.csv', names=columns, skiprows=1)\n",
    "\n",
    "features = ['temperature', 'humidity', 'windy']\n",
    "X = numeric_data[features] # Features\n",
    "Y = numeric_data.play\n",
    "\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3, random_state=1)\n",
    "\n",
    "clf = DecisionTreeClassifier()\n",
    "clf.fit(X_train,Y_train)\n",
    "\n",
    "Y_pred = clf.predict(X_test)\n",
    "\n",
    "print(\"Accuracy:\",metrics.accuracy_score(Y_test, Y_pred))\n",
    "\n",
    "plot_tree(clf)\n",
    "\n",
    "dot_data = export_graphviz(clf, out_file=None)\n",
    "graph = graphviz.Source(dot_data)\n",
    "graph.render(\"numeric_data\")\n",
    "print(\"Tree PDF saved in project file\")"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Nominal Dataset"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Explanation/Justification"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "My implementation for the nominal dataset is very similar to my numeric implementation as much of the same functions and steps are used in both cases when building decision trees with Scikit-Learn. The main differences we have is an extra value in the training data and the need to preprocess the input into a format that works with classifier decision trees.\n",
    "\n",
    "When making a decision tree, training data must be integers, as our input data here is nominal we need to find a solution to convert our strings into ints in a useful way (not hashmap as this can cause memory management issues). So, we are using an encoder, 3 common encoding solutions exist, Ordinal Encoding, One-Hot Encoding and Dummy Variable Encoding. The Ordinal Encoder encodes our string inputs into ordered integers (if 3 options, then values are 0,1,2), a One-Hot Encoder turns strings into arrays of bits that each represent a possible value of that category (if 3 options, then values are [1,0,0], [0,1,0], [0,0,1]) and Dummy Variable Encoder acts similarly to One-Hot but removes redundant values by dropping unnecessary bits (if 3 options, then values are [1,0], [0,1], [0,0]). For our use case, we will use One-Hot for encoding as we do not have naturally ordered data and Dummy is only useful for much more complicated problems.\n",
    "\n",
    "To do encoding, we simply import Scikit-learn's preprocessing library, create our encoder (handle_unknown is set to ignore so that any input that doesn't fit is skipped instead of raising an error) and use it's fit_transform function to convert our data into an encoded form. We do this for both the X training data and X testing data.\n"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### DT Model"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.4\n",
      "Tree PDF saved in project file\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "columns = ['outlook','temperature', 'humidity', 'windy', 'play']\n",
    "\n",
    "nominal_data = pd.read_csv('datasets/weather.nominal.csv', names=columns, skiprows=1)\n",
    "\n",
    "features = ['outlook','temperature', 'humidity', 'windy']\n",
    "X = nominal_data[features] # Features\n",
    "Y = nominal_data.play\n",
    "\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3, random_state=1)\n",
    "\n",
    "clf = DecisionTreeClassifier()\n",
    "\n",
    "enc = preprocessing.OneHotEncoder(handle_unknown='ignore')\n",
    "X_train = enc.fit_transform(X_train)\n",
    "\n",
    "clf.fit(X_train,Y_train)\n",
    "\n",
    "X_test = enc.fit_transform(X_test)\n",
    "Y_pred = clf.predict(X_test)\n",
    "\n",
    "print(\"Accuracy:\",metrics.accuracy_score(Y_test, Y_pred))\n",
    "\n",
    "plot_tree(clf)\n",
    "\n",
    "dot_data = export_graphviz(clf, out_file=None)\n",
    "graph = graphviz.Source(dot_data)\n",
    "graph.render(\"nominal_data\")\n",
    "print(\"Tree PDF saved in project file\")"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "For both of our decision trees, we have a fluctuating accuracy between 0.2 and 0.4, this means that when performing tests, the tree predicts the output correctly only between 20% and 40% of the time. As a rule of thumb, in machine learning, a result that can be considered a \"good accuracy\" level is at least 60% to 70% as this will lead to the model predicting correctly a majority of the time. Here, our models aren't performing accurately enough as there is a significant lack of data both for training and testing. To correct this, a new model can be created with much more data, below is a copy of our numeric implementation but with 50 rows of data instead of 14. New rows were added and some data was duplicated. As our quantity of data is much higher, I have reduced the percentage of test data to 20% and have shuffled the data an extra time."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 1.0\n",
      "Tree PDF saved in project file\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "columns = ['temperature', 'humidity', 'windy', 'play']\n",
    "numeric_data = pd.read_csv('datasets/weather.numericBetter.csv', names=columns, skiprows=1)\n",
    "\n",
    "features = ['temperature', 'humidity', 'windy']\n",
    "X = numeric_data[features] # Features\n",
    "Y = numeric_data.play\n",
    "\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=2)\n",
    "\n",
    "clf = DecisionTreeClassifier()\n",
    "clf.fit(X_train,Y_train)\n",
    "\n",
    "Y_pred = clf.predict(X_test)\n",
    "\n",
    "print(\"Accuracy:\",metrics.accuracy_score(Y_test, Y_pred))\n",
    "\n",
    "plot_tree(clf)\n",
    "\n",
    "dot_data = export_graphviz(clf, out_file=None)\n",
    "graph = graphviz.Source(dot_data)\n",
    "graph.render(\"numeric_data\")\n",
    "print(\"Tree PDF saved in project file\")"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "As we can see, increasing the amount of data has made the accuracy jump up to 1.0, so the model should be trained to predict an outcome properly 100% of the time."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# List of references\n",
    "\n",
    "- Jason Brownlee. (2020). Ordinal and One-Hot Encodings for Categorical Data. https://machinelearningmastery.com/one-hot-encoding-for-categorical-data/\n",
    "- Scikit Learn. (2022). 6.3.4.Encoding categorical features. https://scikit-learn.org/stable/modules/preprocessing.html\n",
    "- Avinash Navlani. (2018). Decision Tree Classification in Python Tutorial. https://www.datacamp.com/tutorial/decision-tree-classification-python"
   ]
  }
 ],
 "metadata": {
  "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": "3.8.3"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "vscode": {
   "interpreter": {
    "hash": "6d1e45cadc3597bb8b6600530fbdf8c3eefe919a24ef54d9d32b318795b772e0"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}