Predicting Bike Rentals.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting Bike Rentals "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Many American cities have communal bike sharing stations where people can rent bicycles by the hour or day. Washington, D.C. is one such city. The District collects detailed data on the number of bicycles people rent by the hour and day.\n",
    "\n",
    "Hadi Fanaee-T at the University of Porto compiled this data into a CSV file, we will use in this project. The file contains 17380 rows, with each row representing the number of bike rentals for a single hour of a single day. The data used is publicly available to download from UC Irvine's website here: http://archive.ics.uci.edu/ml/datasets/Bike+Sharing+Dataset\n",
    "\n",
    "Here are the descriptions for the relevant columns:\n",
    "\n",
    "- instant - A unique sequential ID number for each row\n",
    "- dteday - The date of the rentals\n",
    "- season - The season in which the rentals occurred\n",
    "- yr - The year the rentals occurred\n",
    "- mnth - The month the rentals occurred\n",
    "- hr - The hour the rentals occurred\n",
    "- holiday - Whether or not the day was a holiday\n",
    "- weekday - The day of the week (as a number, 0 to 7)\n",
    "- workingday - Whether or not the day was a working day\n",
    "- weathersit - The weather (as a categorical variable)\n",
    "- temp - The temperature, on a 0-1 scale\n",
    "- atemp - The adjusted temperature\n",
    "- hum - The humidity, on a 0-1 scale\n",
    "- windspeed - The wind speed, on a 0-1 scale\n",
    "- casual - The number of casual riders (people who hadn't previously signed up with the bike sharing program)\n",
    "- registered - The number of registered riders (people who had already signed up)\n",
    "- cnt - The total number of bike rentals (casual + registered)\n",
    "\n",
    "In this project, we will try to predict the total number of bikes people rented (cnt column) in a given hour using all the other columns, except for casual and registered. To find the best prediction, we will create a few different machine learning models and evaluate their performance. We will then compare the performance of each model to find the optimal machine learning method.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import numpy as np\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import r2_score\n",
    "import time\n",
    "import itertools"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>instant</th>\n",
       "      <th>dteday</th>\n",
       "      <th>season</th>\n",
       "      <th>yr</th>\n",
       "      <th>mnth</th>\n",
       "      <th>hr</th>\n",
       "      <th>holiday</th>\n",
       "      <th>weekday</th>\n",
       "      <th>workingday</th>\n",
       "      <th>weathersit</th>\n",
       "      <th>temp</th>\n",
       "      <th>atemp</th>\n",
       "      <th>hum</th>\n",
       "      <th>windspeed</th>\n",
       "      <th>casual</th>\n",
       "      <th>registered</th>\n",
       "      <th>cnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>01/01/2011</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.2879</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>13</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>01/01/2011</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.2727</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>01/01/2011</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.2727</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "      <td>27</td>\n",
       "      <td>32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>01/01/2011</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.2879</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>01/01/2011</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.2879</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   instant      dteday  season  yr  mnth  hr  holiday  weekday  workingday  \\\n",
       "0        1  01/01/2011       1   0     1   0        0        6           0   \n",
       "1        2  01/01/2011       1   0     1   1        0        6           0   \n",
       "2        3  01/01/2011       1   0     1   2        0        6           0   \n",
       "3        4  01/01/2011       1   0     1   3        0        6           0   \n",
       "4        5  01/01/2011       1   0     1   4        0        6           0   \n",
       "\n",
       "   weathersit  temp   atemp   hum  windspeed  casual  registered  cnt  \n",
       "0           1  0.24  0.2879  0.81        0.0       3          13   16  \n",
       "1           1  0.22  0.2727  0.80        0.0       8          32   40  \n",
       "2           1  0.22  0.2727  0.80        0.0       5          27   32  \n",
       "3           1  0.24  0.2879  0.75        0.0       3          10   13  \n",
       "4           1  0.24  0.2879  0.75        0.0       0           1    1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bike_rentals = pd.read_csv('bike_rental_hour.csv')\n",
    "bike_rentals.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(17379, 17)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bike_rentals.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Distribution of cnt')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram of the cnt column of bike_rentals to observe the distribution of total rentals.\n",
    "\n",
    "bike_rentals['cnt'].hist()\n",
    "plt.title('Distribution of cnt')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The histogram shows the cnt column is positively skewed (skewed to the right). About 40% of rentals numbers are less than 100. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "instant       0.278379\n",
       "season        0.178056\n",
       "yr            0.250495\n",
       "mnth          0.120638\n",
       "hr            0.394071\n",
       "holiday      -0.030927\n",
       "weekday       0.026900\n",
       "workingday    0.030284\n",
       "weathersit   -0.142426\n",
       "temp          0.404772\n",
       "atemp         0.400929\n",
       "hum          -0.322911\n",
       "windspeed     0.093234\n",
       "casual        0.694564\n",
       "registered    0.972151\n",
       "cnt           1.000000\n",
       "Name: cnt, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Explore how each column is correlated with cnt.\n",
    "\n",
    "bike_rentals.corr()[\"cnt\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x14c0471eaf0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# These correlations can be showed graphically:\n",
    "\n",
    "plt.figure(figsize=(15, 10))\n",
    "sns.heatmap(bike_rentals.corr(), annot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "cnt           1.000000\n",
       "registered    0.972151\n",
       "casual        0.694564\n",
       "temp          0.404772\n",
       "atemp         0.400929\n",
       "hr            0.394071\n",
       "hum           0.322911\n",
       "Name: cnt, dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's find out which variables have an absolute correlation with cnt (total number of bike rentals) greater than 0.3:\n",
    "\n",
    "corr_coefficients = abs(bike_rentals.corr()['cnt'])\n",
    "corr_coefficients[corr_coefficients>0.3].sort_values(ascending = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Excluding registered and casual columns since cnt is the sum of them, the following columns have high correlations with the cnt column:\n",
    "\n",
    "- temperature\n",
    "- hour\n",
    "- humidity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Engineering "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It can often be helpful to calculate features before applying machine learning models. Features can enhance the accuracy of models by introducing new information or distilling existing information.\n",
    "\n",
    "For example, the hr column in bike_rentals contains the hours during which bikes are rented, from 1 to 24. A machine treats each hour differently, without understanding that certain hours are related. We can introduce some order into the process by creating a new column with labels for morning, afternoon, evening, and night. This bundles similar times together, enabling the model to make better decisions.\n",
    "\n",
    "We will create a function assign_label() that takes in a numeric value for an hour and returns:\n",
    "\n",
    "- 1 if the hour is from 6 to 12\n",
    "- 2 if the hour is from 12 to 18\n",
    "- 3 if the hour is from 18 to 24\n",
    "- 4 if the hour is from 0 to 6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def assign_label(hour):\n",
    "    if hour >=0 and hour < 6:\n",
    "        return 4\n",
    "    elif hour >=6 and hour < 12:\n",
    "        return 1\n",
    "    elif hour >= 12 and hour < 18:\n",
    "        return 2\n",
    "    elif hour >= 18 and hour <=24:\n",
    "        return 3\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    4375\n",
       "3    4368\n",
       "1    4360\n",
       "4    4276\n",
       "Name: time_label, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bike_rentals[\"time_label\"] = bike_rentals[\"hr\"].apply(assign_label)\n",
    "bike_rentals['time_label'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train and Test Data Sets "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before applying machine learning methods, we need to split the data into training and testing sets. This will enable us to train an algorithm using the training set, and evaluate its accuracy on the testing set. As if you train an algorithm on the training data, then evaluate its performance on the same data, you can get an unrealistically low error value, due to overfitting.\n",
    "\n",
    "We will select 80% of the rows in bike_rentals to be part of the training set with the rest of the rows part of the testing set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train shape:  (13903, 18)\n",
      "test shape:  (3476, 18)\n"
     ]
    }
   ],
   "source": [
    "train = bike_rentals.sample(frac = .8, random_state = 1)\n",
    "test = bike_rentals.loc[~bike_rentals.index.isin(train.index)]\n",
    "print(\"train shape: \", train.shape)\n",
    "print(\"test shape: \", test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following metrics are used to evaluate the performance of the machine learning algorithms:\n",
    "\n",
    "- Root of the mean square (RMSE): This gives us an absolute number on how much our predicted results deviate from the actual number. Each model is applied to both the training and test data set and then the RMSE os calculated for both (called training error and validation error).\n",
    "\n",
    "- R-Squared (R2): This indicates how much of the variance in observations the model accounts for. It is a good measure to determine how well the model fits the dependent variables. It is a value between 0 and 1 and generally a higher value is better but unrealistically too high values can be a sign of overfitting.\n",
    "\n",
    "- Difference between training and validation error: To avoid overfitting, we will minimize the difference between these two errors while also trying to minimize the RMSEs and looking for a high enough R-Squared value. \n",
    "\n",
    "To compare results between models and parameters, a combined error score will be calculated as follows:\n",
    "\n",
    "((1/2)(rmse1(0.4) + rmse2(0.6))) + ((-1/4)(r2_1(0.5) + r2_2(0.5))) + (1/4)(abs(rmse2 - rmse1))\n",
    "\n",
    "where:\n",
    "\n",
    "- rmse1: RMSE from training dataset\n",
    "- rmse2: RMSE from test dataset\n",
    "- r2_1: R2 from training dataset\n",
    "- r2_2: R2 from test dataset\n",
    "\n",
    "So, we give greater weight towards the RMSEs for training and test datasets but more weight is given to the RMSE for the test dataset as decision tree models are prone to overfitting and we could end up with a model that has a very low average RMSE being best even though it is highly overfitted and a poor predictor of data outside the training dataset. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Selection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to create a list of predictor columns to use in training and testing machine learning models. We ignore the 'casual' and 'registered' columns because cnt is derived from them. \n",
    "\n",
    "'temp' and 'atemp' columns are highly correlated so one of them is sufficient for prediction. Thus, we exclude 'atemp' column. \n",
    "\n",
    "We exclude 'dteday' since it is not numeric and its data is mentioned in other columns. We also remove 'instant' as it is only an id. The target variable 'cnt' is removed as it cannot be its own predictor. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['season', 'yr', 'mnth', 'hr', 'holiday', 'weekday', 'workingday',\n",
       "       'weathersit', 'temp', 'atemp', 'hum', 'windspeed', 'time_label'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "excluded_columns = ['cnt', 'casual', 'registered','dteday', 'instant']\n",
    "features = bike_rentals.columns.drop(excluded_columns)\n",
    "features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Machine Learning Methods to Predict Bike Rentals "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1: Linear Regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Linear regression should work fairly well on this data, given that many of the columns are highly correlated with cnt, as linear regression works best when predictors are linearly correlated to the target and also independent - in other words, they don't change meaning when we combine them with each other. The good thing about linear regression is that it's fairly resistant to overfitting because it's straightforward. It also can be prone to underfitting the data, however, and not building a powerful enough model. This means that linear regression usually isn't the most accurate option."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Linear Regression\n",
      "*************************\n",
      "train error (rmse, r2): (132.6124243916765, 0.4702152740503345)\n",
      "validation error (rmse, r2): (130.5946379586095, 0.46211423723840306)\n",
      "Difference:  -2.017786433067016\n"
     ]
    }
   ],
   "source": [
    "def linear_regression(train , test, features, target):\n",
    "    lr = LinearRegression()\n",
    "    lr.fit(train[features], train[target])\n",
    "    predictions = lr.predict(test[features])\n",
    "    mse = mean_squared_error(test[target], predictions)\n",
    "    rmse = np.sqrt(mse)\n",
    "    r2 = r2_score(test[target], predictions)\n",
    "    return rmse, r2\n",
    "\n",
    "train_error = linear_regression(train, train, features, 'cnt')\n",
    "test_error = linear_regression(train, test, features, 'cnt')\n",
    "print(\"Linear Regression\")\n",
    "print(\"*************************\")\n",
    "print(\"train error (rmse, r2):\", train_error)\n",
    "print(\"validation error (rmse, r2):\", test_error)\n",
    "print(\"Difference: \", test_error[0] - train_error[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The errors should be explained in comparison with the errors of other models but they look high at first glance. This may bb because the data has a few extremely high rental counts but otherwise mostly low counts and large errors are penalized more with RMSE or MSE measurements which leads to a higher total error.  \n",
    "\n",
    "Also, both training and validation errors are high and the training error is higher than the validation error. So the model seems too poor for prediction and there is a probability of underfitting."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2: Decision Tree Algorithm "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now apply a decision tree algorithm and will compare the error performance of this model to that of the linear regression model to pick the right algorithm for this dataset.\n",
    "\n",
    "Decision trees tend to predict outcomes much more reliably than linear regression models, because a decision tree is a fairly complex model, it also tends to overfit, particularly when we don't tweak parameters like maximum depth and minimum number of samples per leaf. Decision trees are also prone to instability - small changes in the input data can result in a very different output model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree Regression with default settings\n",
      "***********************************************\n",
      "train error (rsme, r2): (0.5693317794626366, 0.9999902352336465)\n",
      "validation error (rsme, r2): (57.9713978521288, 0.8940095093156276)\n",
      "Difference:  57.402066072666166\n"
     ]
    }
   ],
   "source": [
    "def DecisionTreeRegressor_model(train, test, feature, target, min_samples_leaf = 1, max_depth = None):\n",
    "    dtr = DecisionTreeRegressor(random_state=1, min_samples_leaf=min_samples_leaf, max_depth = max_depth)\n",
    "    dtr.fit(train[features], train[target])\n",
    "    predictions = dtr.predict(test[features])\n",
    "    mse = mean_squared_error(test[target], predictions)\n",
    "    rmse = np.sqrt(mse)\n",
    "    r2 = r2_score(test[target], predictions)\n",
    "    return rmse, r2\n",
    "\n",
    "train_error = DecisionTreeRegressor_model(train, train, features, 'cnt')\n",
    "test_error = DecisionTreeRegressor_model(train, test, features, 'cnt')\n",
    "print(\"Decision Tree Regression with default settings\")\n",
    "print(\"***********************************************\")\n",
    "print(\"train error (rsme, r2):\", train_error)\n",
    "print(\"validation error (rsme, r2):\", test_error)\n",
    "print(\"Difference: \", test_error[0] - train_error[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The error is significantly less than the error of the linear regression algorithm. This would be expected as the decision tree regressor can take nonlinear predictors into account to form more accurate predictions. However, with the current (default) parameters used, there is signs of overfitting. \n",
    "\n",
    "A low training error while significantly larger validation error shows overfitting. Too high r2 value for training (0.99) can also be another sign of overfitting. Let's see if we can tune the model by changing the min_samples_leaf to optimize the model's performance. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision Tree Parameter Optimization "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will find the decision tree algorithm with the optimal minimum sample leaves and maximum depth."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def decision_tree_model_optimization(feature, target, min_samples_leaf = 1, max_depth = None):\n",
    "    feature = features \n",
    "    \n",
    "    test1 = train\n",
    "    dtr1 = DecisionTreeRegressor(random_state=1, min_samples_leaf=min_samples_leaf, max_depth = max_depth)\n",
    "    dtr1.fit(train[features], train[target])\n",
    "    predictions1 = dtr1.predict(test1[features])\n",
    "    mse1 = mean_squared_error(test1[target], predictions1)\n",
    "    rmse1 = np.sqrt(mse1)\n",
    "    r2_1 = r2_score(test1[target], predictions1)\n",
    "    \n",
    "    \n",
    "    test2 = test\n",
    "    dtr2 = DecisionTreeRegressor(random_state=1, min_samples_leaf=min_samples_leaf, max_depth = max_depth)\n",
    "    dtr2.fit(train[features], train[target])\n",
    "    predictions2 = dtr2.predict(test2[features])\n",
    "    mse2 = mean_squared_error(test2[target], predictions2)\n",
    "    rmse2 = np.sqrt(mse2)\n",
    "    r2_2 = r2_score(test2[target], predictions2)\n",
    "    \n",
    "    combined_performance = ((1/2)*(rmse1*0.4 + rmse2*0.6)) + ((-1/4)*(r2_1*0.5 + r2_2*0.5)) + (1/4)*abs(rmse2 - rmse1)\n",
    "    \n",
    "    return combined_performance\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "380\n"
     ]
    }
   ],
   "source": [
    "# Find possible combinations of minimum sample leaves and maximum depth:\n",
    "\n",
    "# Range of values to consider:\n",
    "range_values = range(1,20,1)\n",
    "\n",
    "# List of possible combinations for minimum sample leaves and maximum depth:\n",
    "combinations = []\n",
    "\n",
    "for subset in itertools.combinations(range_values, 2):\n",
    "    # Convert values stored as tuples into lists\n",
    "    subset_values = [item for item in subset]\n",
    "    combinations.append(subset_values)\n",
    "\n",
    "# Add combinations in reverse order e.g. [3, 2] instead of only [2, 3]:\n",
    "combo_copy = combinations.copy()\n",
    "for combo in combo_copy:\n",
    "    combinations.append([combo[1], combo[0]])\n",
    "    \n",
    "# Add double values like e.g. [2,2], [3,3] ... \n",
    "for subset in itertools.combinations(range_values, 1):\n",
    "    # Convert values stored as tuples into lists\n",
    "    subset_values = [item for item in subset]\n",
    "    subset_values.append([item for item in subset][0])\n",
    "    combinations.append(subset_values)\n",
    "\n",
    "# Add max_depth = None and min_sample_leaves combinations:\n",
    "nones = [ [i, None] for i in range(1,20,1)]\n",
    "for n in nones:\n",
    "    combinations.append(n)\n",
    "    \n",
    "print(len(combinations)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dictionary to store combined error score:\n",
    "\n",
    "# Key is in format: min_samples_leaf, max_depth:\n",
    "optimal_dict_decision_tree = {}\n",
    "\n",
    "for combo in combinations:\n",
    "    optimal_dict_decision_tree[combo[0],combo[1]] = decision_tree_model_optimization(features, 'cnt', min_samples_leaf = combo[0], max_depth = combo[1])\n",
    "\n",
    "# optimal_dict_decision_tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Find optimal combination\n",
    "\n",
    "def optimal_parameters(optimal_dictionary):\n",
    "    base_optimal_combo = combinations[0]\n",
    "    base_optimal_score = optimal_dictionary[combinations[0][0], combinations[0][1]]\n",
    "    \n",
    "    for k in optimal_dictionary:\n",
    "        if optimal_dictionary[k] < base_optimal_score:\n",
    "            base_optimal_combo = k\n",
    "            base_optimal_score = optimal_dictionary[k]\n",
    "        else:\n",
    "            pass\n",
    "        \n",
    "    return base_optimal_combo, base_optimal_score "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For the decision tree model, the optimal minimum number of leaves is: 8 and the optimal maximum depth is: None\n"
     ]
    }
   ],
   "source": [
    "print(\"For the decision tree model, the optimal minimum number of leaves is:\",str(optimal_parameters(optimal_dict_decision_tree)[0][0]),\"and the optimal maximum depth is:\",str(optimal_parameters(optimal_dict_decision_tree)[0][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3: Random Forests "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now apply the random forest algorithm, which improves on the decision tree algorithm. Random forests tend to be much more accurate than simple models like linear regression. Due to the way random forests are constructed, they tend to overfit much less than decision trees. However, random forests are prone to overfitting, so it's important we tune parameters like maximum depth and minimum samples per leaf to avoid this and obtain the optimal model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "random_forest_model with default setting\n",
      "***********************************************\n",
      "train error: (15.895826268783399, 0.992388027359284)\n",
      "validation error: (44.310634484224785, 0.9380765285401584)\n",
      "Difference:  28.414808215441386\n"
     ]
    }
   ],
   "source": [
    "def random_forest_model(train, test, feature, target, n_estimators = 100, min_samples_leaf = 1, max_depth = None):\n",
    "    rf = RandomForestRegressor(n_estimators = n_estimators, random_state=1, min_samples_leaf = min_samples_leaf, max_depth = max_depth)\n",
    "    rf.fit(train[feature], train[target])\n",
    "    predictions = rf.predict(test[feature])\n",
    "    mse = mean_squared_error(test[target], predictions)\n",
    "    rmse = np.sqrt(mse)\n",
    "    r2 = r2_score(test[target], predictions)\n",
    "    return rmse, r2\n",
    "\n",
    "train_error = random_forest_model(train, train, features, 'cnt')\n",
    "test_error = random_forest_model(train, test, features, 'cnt')\n",
    "print(\"random_forest_model with default setting\")\n",
    "print(\"***********************************************\")\n",
    "print(\"train error:\", train_error)\n",
    "print(\"validation error:\", test_error)\n",
    "print(\"Difference: \", test_error[0] - train_error[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forests Parameter Optimization "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will find the random forest algorithm with the optimal minimum sample leaves and maximum depth. In doing so we use the default number of trees (estimators) which is 100. This is for time purposes. In theory we could also search for the optimal number of trees (or just choose a larger number of trees as more trees provides greater accuracy) however this would take considerably more time and the impact of increasing the number of trees on accuracy is subject to diminishing returns. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def random_forest_model_optimization(feature, target, n_estimators = 100, min_samples_leaf = 1, max_depth = None):\n",
    "    feature = features \n",
    "    \n",
    "    test1 = train\n",
    "    rf1 = RandomForestRegressor(n_estimators = n_estimators, random_state=1, min_samples_leaf = min_samples_leaf, max_depth = max_depth)\n",
    "    rf1.fit(train[feature], train[target])\n",
    "    predictions1 = rf1.predict(test1[feature])\n",
    "    mse1 = mean_squared_error(test1[target], predictions1)\n",
    "    rmse1 = np.sqrt(mse1)\n",
    "    r2_1 = r2_score(test1[target], predictions1)\n",
    "     \n",
    "    test2 = test\n",
    "    rf2 = RandomForestRegressor(n_estimators = n_estimators, random_state=1, min_samples_leaf = min_samples_leaf, max_depth = max_depth)\n",
    "    rf2.fit(train[feature], train[target])\n",
    "    predictions2 = rf2.predict(test2[feature])\n",
    "    mse2 = mean_squared_error(test2[target], predictions2)\n",
    "    rmse2 = np.sqrt(mse2)\n",
    "    r2_2 = r2_score(test2[target], predictions2)\n",
    "    \n",
    "    combined_performance = ((1/2)*(rmse1*0.4 + rmse2*0.6)) + ((-1/4)*(r2_1*0.5 + r2_2*0.5)) + (1/4)*abs(rmse2 - rmse1)\n",
    "    \n",
    "    return combined_performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "22.922476883101982\n",
      "It took 8.55 seconds to compute\n"
     ]
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "print(random_forest_model_optimization(features, 'cnt', n_estimators = 100, min_samples_leaf = 2, max_depth = None))\n",
    "\n",
    "end_time = time.time()\n",
    "print(f\"It took {end_time-start_time:.2f} seconds to compute\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dictionary to store combined error score:\n",
    "# Key is in format: min_samples_leaf, max_depth:\n",
    "\n",
    "optimal_dict_random_forest = {}\n",
    "\n",
    "for combo in combinations:\n",
    "    optimal_dict_random_forest[combo[0],combo[1]] = random_forest_model_optimization(features, 'cnt', n_estimators = 100, min_samples_leaf = combo[0], max_depth = combo[1])\n",
    "\n",
    "# optimal_dict_random_forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Find optimal combination\n",
    "\n",
    "def optimal_parameters(optimal_dictionary):\n",
    "    base_optimal_combo = combinations[0]\n",
    "    base_optimal_score = optimal_dictionary[combinations[0][0], combinations[0][1]]\n",
    "    \n",
    "    for k in optimal_dictionary:\n",
    "        if optimal_dictionary[k] < base_optimal_score:\n",
    "            base_optimal_combo = k\n",
    "            base_optimal_score = optimal_dictionary[k]\n",
    "        else:\n",
    "            pass\n",
    "        \n",
    "    return base_optimal_combo, base_optimal_score \n",
    "            \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For the random forest model, the optimal minimum number of leaves is: 2 and the optimal maximum depth is: None\n"
     ]
    }
   ],
   "source": [
    "print(\"For the random forest model, the optimal minimum number of leaves is:\",str(optimal_parameters(optimal_dict_random_forest)[0][0]),\"and the optimal maximum depth is:\",str(optimal_parameters(optimal_dict_random_forest)[0][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Therefore, 2 minimum sample leaves and no maximum depth (while all other parameters are using their default values) is the optimal random forest model. \n",
    "\n",
    "We could find the optimal set of parameters for all parameters (i.e. using no default values and comparing all possible combinations) however this would take too long to run. In addition, there is unlikely to be a significant improvement over the optimal combination we propose here as minimum sample leaves and maximum depth are the most common parameters to optimize over."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this project, we have used various machine learning techniques to predict bike rentals. We will now compare the results of each model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>train_rsme</th>\n",
       "      <th>train_r2</th>\n",
       "      <th>test_rsme</th>\n",
       "      <th>test_r2</th>\n",
       "      <th>diff_rsme</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Linear Regression</th>\n",
       "      <td>132.612</td>\n",
       "      <td>0.470</td>\n",
       "      <td>130.595</td>\n",
       "      <td>0.462</td>\n",
       "      <td>-2.018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Decision Tree Default</th>\n",
       "      <td>0.569</td>\n",
       "      <td>1.000</td>\n",
       "      <td>57.971</td>\n",
       "      <td>1.000</td>\n",
       "      <td>57.402</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Decision Tree with Optimal Parameters</th>\n",
       "      <td>40.560</td>\n",
       "      <td>0.950</td>\n",
       "      <td>53.786</td>\n",
       "      <td>0.909</td>\n",
       "      <td>13.226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Random Forest Default</th>\n",
       "      <td>15.896</td>\n",
       "      <td>0.992</td>\n",
       "      <td>44.311</td>\n",
       "      <td>0.938</td>\n",
       "      <td>21.941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Random Forest with Optimal Parameters</th>\n",
       "      <td>22.370</td>\n",
       "      <td>0.985</td>\n",
       "      <td>44.148</td>\n",
       "      <td>0.939</td>\n",
       "      <td>21.778</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       train_rsme  train_r2  test_rsme  \\\n",
       "Linear Regression                         132.612     0.470    130.595   \n",
       "Decision Tree Default                       0.569     1.000     57.971   \n",
       "Decision Tree with Optimal Parameters      40.560     0.950     53.786   \n",
       "Random Forest Default                      15.896     0.992     44.311   \n",
       "Random Forest with Optimal Parameters      22.370     0.985     44.148   \n",
       "\n",
       "                                       test_r2  diff_rsme  \n",
       "Linear Regression                        0.462     -2.018  \n",
       "Decision Tree Default                    1.000     57.402  \n",
       "Decision Tree with Optimal Parameters    0.909     13.226  \n",
       "Random Forest Default                    0.938     21.941  \n",
       "Random Forest with Optimal Parameters    0.939     21.778  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "algorithms_df = pd.DataFrame(index = ['Linear Regression', 'Decision Tree Default', 'Decision Tree with Optimal Parameters', 'Random Forest Default', 'Random Forest with Optimal Parameters'],\n",
    "            columns = ['train_rsme', 'train_r2', 'test_rsme', 'test_r2', 'diff_rsme'], \n",
    "            data = np.array([[linear_regression(train, train, features, 'cnt')[0], linear_regression(train, train, features, 'cnt')[1], linear_regression(train, test, features, 'cnt')[0], linear_regression(train, test, features, 'cnt')[1], linear_regression(train, test, features, 'cnt')[0] - linear_regression(train, train, features, 'cnt')[0]],\n",
    "                             [DecisionTreeRegressor_model(train, train, features, 'cnt')[0], DecisionTreeRegressor_model(train, train, features, 'cnt')[1], DecisionTreeRegressor_model(train, test, features, 'cnt')[0], DecisionTreeRegressor_model(train, train, features, 'cnt')[1], DecisionTreeRegressor_model(train, test, features, 'cnt')[0] - DecisionTreeRegressor_model(train, train, features, 'cnt')[0]],\n",
    "                             [DecisionTreeRegressor_model(train, train, features, 'cnt', min_samples_leaf = optimal_parameters(optimal_dict_decision_tree)[0][0], max_depth = optimal_parameters(optimal_dict_decision_tree)[0][1])[0], DecisionTreeRegressor_model(train, train, features, 'cnt', min_samples_leaf = optimal_parameters(optimal_dict_decision_tree)[0][0], max_depth = optimal_parameters(optimal_dict_decision_tree)[0][1])[1], DecisionTreeRegressor_model(train, test, features, 'cnt', min_samples_leaf = optimal_parameters(optimal_dict_decision_tree)[0][0], max_depth = optimal_parameters(optimal_dict_decision_tree)[0][1])[0], DecisionTreeRegressor_model(train, test, features, 'cnt', min_samples_leaf = optimal_parameters(optimal_dict_decision_tree)[0][0], max_depth = optimal_parameters(optimal_dict_decision_tree)[0][1])[1], DecisionTreeRegressor_model(train, test, features, 'cnt', min_samples_leaf = optimal_parameters(optimal_dict_decision_tree)[0][0], max_depth = optimal_parameters(optimal_dict_decision_tree)[0][1])[0] - DecisionTreeRegressor_model(train, train, features, 'cnt', min_samples_leaf = optimal_parameters(optimal_dict_decision_tree)[0][0], max_depth = optimal_parameters(optimal_dict_decision_tree)[0][1])[0]],\n",
    "                             [random_forest_model(train, train, features, 'cnt', n_estimators = 100, min_samples_leaf = 1, max_depth = None)[0], random_forest_model(train, train, features, 'cnt', n_estimators = 100, min_samples_leaf = 1, max_depth = None)[1], random_forest_model(train, test, features, 'cnt', n_estimators = 100, min_samples_leaf = 1, max_depth = None)[0], random_forest_model(train, test, features, 'cnt', n_estimators = 100, min_samples_leaf = 1, max_depth = None)[1], random_forest_model(train, test, features, 'cnt', n_estimators = 100, min_samples_leaf = 1, max_depth = None)[0] - random_forest_model(train, train, features, 'cnt', n_estimators = 100, min_samples_leaf = 2, max_depth = None)[0]],\n",
    "                             [random_forest_model(train, train, features, 'cnt', n_estimators = 100, min_samples_leaf = optimal_parameters(optimal_dict_random_forest)[0][0], max_depth = optimal_parameters(optimal_dict_random_forest)[0][1])[0], random_forest_model(train, train, features, 'cnt', n_estimators = 100, min_samples_leaf = optimal_parameters(optimal_dict_random_forest)[0][0], max_depth = optimal_parameters(optimal_dict_random_forest)[0][1])[1], random_forest_model(train, test, features, 'cnt', n_estimators = 100, min_samples_leaf = optimal_parameters(optimal_dict_random_forest)[0][0], max_depth = optimal_parameters(optimal_dict_random_forest)[0][1])[0], random_forest_model(train, test, features, 'cnt', n_estimators = 100, min_samples_leaf = optimal_parameters(optimal_dict_random_forest)[0][0], max_depth = optimal_parameters(optimal_dict_random_forest)[0][1])[1], random_forest_model(train, test, features, 'cnt', n_estimators = 100, min_samples_leaf = optimal_parameters(optimal_dict_random_forest)[0][0], max_depth = optimal_parameters(optimal_dict_random_forest)[0][1])[0] - random_forest_model(train, train, features, 'cnt', n_estimators = 100, min_samples_leaf = optimal_parameters(optimal_dict_random_forest)[0][0], max_depth = optimal_parameters(optimal_dict_random_forest)[0][1])[0]]]))\n",
    "\n",
    "algorithms_rounded_df = algorithms_df.round(decimals=3)\n",
    "algorithms_rounded_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thus using our error metric as described earlier, which combines RMSE, R-Squared and differential RMSE values for test and train datasets but weighs most heavily on RMSE test and train dataset values, the random forest method with optimal parameters (2 minimum sample leaves and no maximum depth) is the optimal model across the 3 machine learning methods covered in this project.   "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}