Diabetes Prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Diabetes Risk Prediction\n",
    "\n",
    "### 1. Diabetes mellitus\n",
    "\n",
    "[Diabetes mellitus](https://en.wikipedia.org/wiki/Diabetes) is a disease characterized by high blood sugar levels over long periods. If left untreated, it can result in serious medical complications, such as cardiovascular disease, chronic kidney disease, and even death. \n",
    "\n",
    "### 2. Prevalence of Diabetes, and Devising a Solution\n",
    "The prevalence of Diabetes in Singapore (10.5%) is not only higher than the world average (8.8%), but is also [ever increasing](https://www.healthhub.sg/a-z/diseases-and-conditions/626/diabetes). \n",
    "\n",
    "According to the World Health Organization (WHO), apart from prevention, [early diagnosis is one of the most effective ways](https://www.who.int/en/news-room/fact-sheets/detail/diabetes) to minimize the burden caused by diabetes. \n",
    "\n",
    "With this in mind, we are going to set out to construct a risk profile that measures an invidiual's susceptibility to the disease. The [dataset](https://www.kaggle.com/johndasilva/diabetes/downloads/diabetes.zip/1) we're going to work with was acquired from a hospital in Frankfurt, Germany. It contains a profile of health indicators such as BMI, Age, and Blood Glucose levels of individuals with and without Diabetes.\n",
    "\n",
    "### 3. Capitalizing on the effectiveness of Early Diagnosis\n",
    "Our objective here will be to train a predictive model using the available data to learn what the profile of a person afflicted with Diabetes looks like. This model will then receive the profile of an untested individual, and based on that person's health indicators, measure the similarity of their profile against that of a person with diabetes, and assign that person a value that represents their predicted risk of having Diabetes.\n",
    "\n",
    "Our end-goal is to better educate individuals on their diabetes risk factor, so the disease can be diagnosed and tackled earlier. The predictive model will serve as a means to impart (possibly life-saving) health literacy to its users.\n",
    "\n",
    "## Introduction\n",
    "\n",
    "We'll begin by reading and briefly exploring the `diabetes.csv` dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2000, 9)\n"
     ]
    },
    {
     "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>Pregnancies</th>\n",
       "      <th>Glucose</th>\n",
       "      <th>BloodPressure</th>\n",
       "      <th>SkinThickness</th>\n",
       "      <th>Insulin</th>\n",
       "      <th>BMI</th>\n",
       "      <th>DiabetesPedigreeFunction</th>\n",
       "      <th>Age</th>\n",
       "      <th>Outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>138</td>\n",
       "      <td>62</td>\n",
       "      <td>35</td>\n",
       "      <td>0</td>\n",
       "      <td>33.6</td>\n",
       "      <td>0.127</td>\n",
       "      <td>47</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>84</td>\n",
       "      <td>82</td>\n",
       "      <td>31</td>\n",
       "      <td>125</td>\n",
       "      <td>38.2</td>\n",
       "      <td>0.233</td>\n",
       "      <td>23</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>145</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>44.2</td>\n",
       "      <td>0.630</td>\n",
       "      <td>31</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>135</td>\n",
       "      <td>68</td>\n",
       "      <td>42</td>\n",
       "      <td>250</td>\n",
       "      <td>42.3</td>\n",
       "      <td>0.365</td>\n",
       "      <td>24</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>139</td>\n",
       "      <td>62</td>\n",
       "      <td>41</td>\n",
       "      <td>480</td>\n",
       "      <td>40.7</td>\n",
       "      <td>0.536</td>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pregnancies  Glucose  BloodPressure  SkinThickness  Insulin   BMI  \\\n",
       "0            2      138             62             35        0  33.6   \n",
       "1            0       84             82             31      125  38.2   \n",
       "2            0      145              0              0        0  44.2   \n",
       "3            0      135             68             42      250  42.3   \n",
       "4            1      139             62             41      480  40.7   \n",
       "\n",
       "   DiabetesPedigreeFunction  Age  Outcome  \n",
       "0                     0.127   47        1  \n",
       "1                     0.233   23        0  \n",
       "2                     0.630   31        1  \n",
       "3                     0.365   24        1  \n",
       "4                     0.536   21        0  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "diabetes = pd.read_csv('diabetes.csv')\n",
    "\n",
    "print(diabetes.shape)\n",
    "diabetes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 2000 entries, 0 to 1999\n",
      "Data columns (total 9 columns):\n",
      "Pregnancies                 2000 non-null int64\n",
      "Glucose                     2000 non-null int64\n",
      "BloodPressure               2000 non-null int64\n",
      "SkinThickness               2000 non-null int64\n",
      "Insulin                     2000 non-null int64\n",
      "BMI                         2000 non-null float64\n",
      "DiabetesPedigreeFunction    2000 non-null float64\n",
      "Age                         2000 non-null int64\n",
      "Outcome                     2000 non-null int64\n",
      "dtypes: float64(2), int64(7)\n",
      "memory usage: 140.7 KB\n"
     ]
    }
   ],
   "source": [
    "diabetes.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that the dataset has 2000 rows and 9 columns.\n",
    "\n",
    "Let's go through what the columns mean:\n",
    "\n",
    "- Pregnancies - Number of Pregnancies the person has had.\n",
    "- Glucose - Blood Sugar levels (mg/DL)\n",
    "- BloodPressure - Measured Blood Pressure (mmHg)\n",
    "- SkinThickness - Skin Thickness (mm).\n",
    "- Insulin - Insulin Level\n",
    "- BMI - A person's [Body Mass Index](https://en.wikipedia.org/wiki/Body_mass_index)\n",
    "- DiabetesPedigreeFunction - Measures genetic influence (hereditary risk) of Diabetes\n",
    "- Age - Age of the person\n",
    "- Outcome - \"1\" represents a person with diabetes. \"0\" represents a healthy individual.\n",
    "\n",
    "### Visualizing Distribution of the Data\n",
    "\n",
    "Before proceeding with Data Cleaning, let's get a picture of values in all the above 9 columns are distributed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x1008 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "# Making a list of col names:\n",
    "cols = diabetes.columns.tolist()\n",
    "\n",
    "fig = plt.figure(figsize=(14,14))\n",
    "for position, col in enumerate(cols):\n",
    "    ax = fig.add_subplot(3,3,position+1)\n",
    "    ax.hist(diabetes[col], bins=30)\n",
    "    plt.title(col)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some columns, like `BMI` and `BloodPressure` look normally distributed.\n",
    "\n",
    "One thing that stands out to us is that there are values in the `BMI`, `BloodPressure`, `Insulin`, and `SkinThickness` columns that are zero. \n",
    "\n",
    "Because realistically the values of these things could not possibly be 0 for a person, it likely means that the data in that row was either recorded incorrectly, or not recorded at all.\n",
    "\n",
    "Let's also visualize the distribution of the data based on whether a person had diabetes.\n",
    "\n",
    "### Visualizing distribution of Data based on Diabetes outcome"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Splitting dataframe\n",
    "\n",
    "with_diabetes = diabetes[diabetes['Outcome'] == 1]\n",
    "without_diabetes = diabetes[diabetes['Outcome'] == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x1008 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(14,14))\n",
    "for position, col in enumerate(cols):\n",
    "    ax = fig.add_subplot(3,3,position+1)\n",
    "    ax.hist(with_diabetes[col], histtype ='step', label='With Diabetes', bins = 20)\n",
    "    ax.hist(without_diabetes[col], histtype ='step', label = 'No Diabetes', bins=20)\n",
    "    ax.legend(loc=0)\n",
    "    plt.title(col)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We notice that people with diabetes on average tend to be older, unsurprisingly have higher blood glucose levels, and also have more pregnancies on average which might simply be because of the correlation with age, since an older person is more likely to have had more pregnancies than a younger person."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Processing\n",
    "\n",
    "We want to clean the data to make it suitable to pass into a Machine Learning model. In particular, we should look out for the following things:\n",
    "- Duplicate entries\n",
    "- Incomplete Data (denoted by the several 0 values we saw above)\n",
    "- Outlier Data, or data that is very obviously incorrect.\n",
    "\n",
    "### Dropping Duplicate Rows\n",
    "First, we'll drop any duplicate rows.\n",
    "\n",
    "Because all the data we're working with is from the same hospital, it's extremely unlikely that two rows would be identical across every column. A better explanation would be that the data was incorrectly duplicated. We'll thus drop the duplicate columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(744, 9)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes = diabetes.drop_duplicates(keep='first')\n",
    "\n",
    "diabetes.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that dropping the duplicate columns dramatically reduced the size of the `diabetes` dataframe, from 2000 rows down to only 744.\n",
    "\n",
    "### Handling incomplete Data\n",
    "\n",
    "In some rows, there were 0 values for columns like `BMI`, `BloodPressure`, `Insulin`, and `Glucose`. We'll drop rows that still contain values of 0 in these columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(381, 9)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes = diabetes.drop(diabetes[diabetes['BMI']==0].index) \n",
    "diabetes = diabetes.drop(diabetes[diabetes['BloodPressure']==0].index) \n",
    "diabetes = diabetes.drop(diabetes[diabetes['Insulin']==0].index)\n",
    "diabetes = diabetes.drop(diabetes[diabetes['Glucose']==0].index)\n",
    "diabetes = diabetes.drop(diabetes[diabetes['SkinThickness']==0].index)\n",
    "\n",
    "diabetes.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we see that we are further down to only 381 rows.\n",
    "\n",
    "The main thing left to do at this point is selecting the features intelligently.\n",
    "\n",
    "### Feature Selection\n",
    "\n",
    "The `Pregnancies` column contains information on number of pregnancies, but while there is a correlation between diabetes likelihood and number of pregnancies, our data-set does not actually distinguish between males and females. If, for instance, a male is more likely to have diabetes based on other indicators, we don't want the prediction for him having diabetes being marked down just because his number of `Pregnancies` is 0. To avoid the problem of the predictive model being confounded by this, we will get rid of the `Pregnancies` column. \n",
    "\n",
    "There is also the fact that we have very limited understanding on how the `DiabetesPedigreeFunction` column's values are determined. This column measures an individual's hereditary predisposition to diabetes. Realistically, it is not going to be possible for an individual to input this measurement into our predictive model. We'll thus drop this column as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'Age']\n"
     ]
    }
   ],
   "source": [
    "features = cols.copy()\n",
    "features.remove('Outcome')\n",
    "features.remove('DiabetesPedigreeFunction')\n",
    "features.remove('Pregnancies')\n",
    "\n",
    "\n",
    "print(features)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "At this junction, we are good to move on to training the model.\n",
    "\n",
    "## Predictive Modelling - Logistic Regression\n",
    "\n",
    "We'll first use Logistic Regression.\n",
    "\n",
    "The error metric we'll be focusing on for now is accuracy (total correct predictions/total predictions).\n",
    "\n",
    "We will assess its accuracy with KFold Cross Validation, to use our data with maximum efficiency."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.75       0.6875     0.75       0.87301587 0.79365079 0.79365079]\n",
      "The mean accuracy is: 0.7746362433862434\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "# Instantiate the model\n",
    "weight = {\n",
    "    1:1.05,\n",
    "    0:1\n",
    "}\n",
    "log = LogisticRegression(class_weight = weight)\n",
    "\n",
    "kf = KFold(n_splits=6) \n",
    "score = cross_val_score(log, diabetes[features], diabetes['Outcome'], cv=kf, scoring='accuracy')\n",
    "print(score)\n",
    "print(\"The mean accuracy is:\",score.mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using KFold cross validation with slightly more weight placed on a positive check for Diabetes, with 6 folds, we get an accuracy of around 77-78%.\n",
    "\n",
    "We'll train this model on the entire data-set, and then enter in a custom values to see the probabilities of Diabetes assigned to us."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.90168024, 0.09831976]])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weight = {\n",
    "    1:1.05,\n",
    "    0:1\n",
    "}\n",
    "\n",
    "log_model = LogisticRegression(class_weight = weight)\n",
    "log_model.fit(diabetes[features], diabetes['Outcome'])\n",
    "log_model.predict_proba(np.array(diabetes[features].iloc[0]).reshape(1, -1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Glucose           84.0\n",
       "BloodPressure     82.0\n",
       "SkinThickness     31.0\n",
       "Insulin          125.0\n",
       "BMI               38.2\n",
       "Age               23.0\n",
       "Name: 1, dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes[features].iloc[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.951025775184868"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testing1 = [120,10,100,190,50,60]\n",
    "features\n",
    "\n",
    "log_model.predict_proba(np.array(testing1).reshape(1, -1))[0,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'Age']"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we'll define a function receives inputs for the 6 feature columns, and then returns a Diabetes risk index."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------\n",
      "Health Indicator Analysis\n",
      "-------------------------\n",
      "You are probably in good health, keep it up.\n",
      "Your Diabetes Risk Index is 11.45/50.\n"
     ]
    }
   ],
   "source": [
    "def diabetes_risk_prediction(glucose, bp, skinthickness, insulin, bmi, age):\n",
    "    indicator_list = [glucose, bp, skinthickness, insulin, bmi, age]\n",
    "    predictions = log_model.predict_proba(np.array(indicator_list).reshape(1, -1))\n",
    "    risk = predictions[0,1]\n",
    "    print(\"-\"*len(\"Health Indicator Analysis\"))\n",
    "    print(\"Health Indicator Analysis\")\n",
    "    print(\"-\"*len(\"Health Indicator Analysis\"))   \n",
    "    if risk < 0.3:\n",
    "        print(\"You are probably in good health, keep it up.\")\n",
    "\n",
    "        #print(\"-\"*len(\"You are probably in good health, keep it up\"))\n",
    "    elif risk > 0.7:\n",
    "        print(\"See a doctor as soon as you can and listen to their recommendations. You might be on the way to developing diabetes if you don't change your lifestyle.\")\n",
    "        \n",
    "    elif risk > 0.9:\n",
    "        print(\"Go to a hospital right away. Odds are high you have diabetes.\")\n",
    "    else:\n",
    "        print(\"You should be alright for the most part, but take care not to let your health slip.\")\n",
    "    return print(\"Your Diabetes Risk Index is {:.2f}/50.\".format(risk*0.5*100))\n",
    "    \n",
    "diabetes_risk_prediction(90,80,30,120,60,50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll expand slightly on the function above. Now, it will not only receive all those inputs, but also re-instantiate the logistics regression and model and re-fit the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2713386277845828"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def diabetes_risk_prediction2(glucose, bp, skinthickness, insulin, bmi, age):\n",
    "    weight = {\n",
    "    1:1.05,\n",
    "    0:1}\n",
    "\n",
    "    log_model2 = LogisticRegression(class_weight = weight)\n",
    "    log_model2.fit(diabetes[features], diabetes['Outcome'])\n",
    "    \n",
    "    indicator_list = [glucose, bp, skinthickness, insulin, bmi, age]\n",
    "    predictions = log_model2.predict_proba(np.array(indicator_list).reshape(1, -1))\n",
    "    risk = predictions[0,1]\n",
    "    return risk\n",
    "\n",
    "diabetes_risk_prediction2(100,80,30,120,30,45) #Test array."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Taking User Inputs and generating a Risk Index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Enter your Blood Glucose Level (mg/DL):\n",
      "100\n",
      "Enter your Blood Pressure (mmHg):\n",
      "60\n",
      "Enter your Skin Thickness (mm):\n",
      "20\n",
      "Enter your Insulin Level (mg/DL):\n",
      "130\n",
      "Enter your BMI:\n",
      "27\n",
      "Enter your Age (Years):\n",
      "25\n",
      "Health Indicator Analysis\n",
      "-------------------------\n",
      "You are probably in good health, keep it up.\n",
      "Your Diabetes Risk Index is 7.92/50.\n"
     ]
    }
   ],
   "source": [
    "while True:\n",
    "    try:\n",
    "        while True:\n",
    "            Glucose = int(input(\"Enter your Blood Glucose Level (mg/DL):\\n\"))\n",
    "            if Glucose > 0:\n",
    "                    break\n",
    "            else:\n",
    "                print(\"Please enter a valid value\")\n",
    "        break\n",
    "    except (ValueError, NameError):# if the input is invalid \n",
    "        print(\"Invalid input! Please choose again.\") #prompts user to input valid action\n",
    "        continue\n",
    "\n",
    "while True:\n",
    "    try:\n",
    "        while True:\n",
    "            BloodPressure = int(input(\"Enter your Blood Pressure (mmHg):\\n\"))\n",
    "            if BloodPressure > 0:\n",
    "                    break\n",
    "            else:\n",
    "                print(\"Please enter a valid value\")\n",
    "        break\n",
    "    except (ValueError, NameError):# if the input is invalid \n",
    "        print(\"Invalid input! Please choose again.\") #prompts user to input valid action\n",
    "        continue\n",
    "        \n",
    "while True:\n",
    "    try:\n",
    "        while True:\n",
    "            SkinThickness = int(input(\"Enter your Skin Thickness (mm):\\n\"))\n",
    "            if SkinThickness > 0:\n",
    "                    break\n",
    "            else:\n",
    "                print(\"Please enter a valid value\")\n",
    "        break\n",
    "    except (ValueError, NameError):# if the input is invalid \n",
    "        print(\"Invalid input! Please choose again.\") #prompts user to input valid action\n",
    "        continue\n",
    "        \n",
    "while True:\n",
    "    try:\n",
    "        while True:\n",
    "            Insulin = int(input(\"Enter your Insulin Level (mg/DL):\\n\"))\n",
    "            if Insulin > 0:\n",
    "                    break\n",
    "            else:\n",
    "                print(\"Please enter a valid value\")\n",
    "        break\n",
    "    except (ValueError, NameError):# if the input is invalid \n",
    "        print(\"Invalid input! Please choose again.\") #prompts user to input valid action\n",
    "        continue\n",
    "        \n",
    "\n",
    "while True:\n",
    "    try:\n",
    "        while True:\n",
    "            BMI = int(input(\"Enter your BMI:\\n\"))\n",
    "            if BMI > 0:\n",
    "                    break\n",
    "            else:\n",
    "                print(\"Please enter a valid value\")\n",
    "        break\n",
    "    except (ValueError, NameError):# if the input is invalid \n",
    "        print(\"Invalid input! Please choose again.\") #prompts user to input valid action\n",
    "        continue    \n",
    "\n",
    "while True:\n",
    "    try:\n",
    "        while True:\n",
    "            Age = int(input(\"Enter your Age (Years):\\n\"))\n",
    "            if Age > 0:\n",
    "                    break\n",
    "            else:\n",
    "                print(\"Please enter a valid value\")\n",
    "        break\n",
    "    except (ValueError, NameError):# if the input is invalid \n",
    "        print(\"Invalid input! Please choose again.\") #prompts user to input valid action\n",
    "        continue\n",
    "        \n",
    "diabetes_risk_prediction(Glucose,BloodPressure,SkinThickness,Insulin,BMI,Age)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Future Work"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some future improvements to improve our predictive accuracy could be:\n",
    "- Use an error metric other than accuracy score, like f1 score or precision.\n",
    "- Use domain knowledge (acquired from a Doctor or more qualified person) to make better choices about feature engineering.\n",
    "- Doing the prediction with algorithms like Neural Networks instead. \n",
    "- Using a larger data-set - ideally one from Singapore."
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}