Assignment_2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Name: Taylor D. Gabatino"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import stats\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy as sp\n",
    "import scipy.stats as stats\n",
    "import pandas as pd\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Brief Overview \n",
    "\n",
    "- The goal of this assignment is to practice some essential probability and hypothesis testing sections.\n",
    "\n",
    "The assignment consists of two parts: 1- generative models and simulation; and 2- data analysis, interpretation and hypothesis testing. In both problems, you are given bits of information that guide you through solving the problem step-by-step before you are asked to repeat the analysis on a different question.\n",
    "\n",
    "##### Instructions on answering the questions\n",
    "\n",
    "1. The only Python modules you can use in this assignment are those imported for you in the top cell. They are `matplotlib`, `numpy`, `scipy`, and `pandas`\n",
    "\n",
    "2. Your graphs must have titles and axes labels.\n",
    "  The plot size can be set using `plt.figure(figsize=(a,b))`, where (a,b) are the plot dimensions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PROBLEM 1\n",
    "----"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Modeling Revenue from Memberships on a Training Website\n",
    "\n",
    "* We are interested in modeling sales revenue for a membership-based training website over 2022. In summary, the website charges a monthly fee ($39.99) and enrollees can take as long as they need to complete their training. Students can quit the platform at any time without completing the training.\n",
    "\n",
    "We know that:\n",
    "  * Women finish training faster on average than men\n",
    "  * Women quit less often on average than men \n",
    "  * Students have a small chance of dropping each month. As students get close to the end of the training, the probability of dropping decays (decreases), i.e., the longer a student has paid, the less likely they are to give up on their investment and leave the platform.\n",
    "  \n",
    "  * The site has seen a steady annual increase in memberships over the past five years. The average yearly increase was 13%, and the standard deviation was 1.4%. Memberships are expected to increase by the same percentage next year.\n",
    "   * In 2021, 90,316 people enrolled on the training site\n",
    "  \n",
    "* We have access to some of the data from previous years. The data is stored in the file `memberships_info.csv` and can be downloaded from the following [URL](https://www.dropbox.com/s/evaaiu1mevw6t3j/memberships_info.csv?dl=1)\n",
    "\n",
    "   \n",
    "The data contains the following columns:\n",
    "\n",
    "| Column    | Description                                    |\n",
    "|-----------|------------------------------------------------|\n",
    "| member_id | unique identifier of the member                |\n",
    "| gender    | M for male and F for female                    |\n",
    "| status    | C for completed and Q for quit                 |\n",
    "| mem_duration  | Total duration of the membership in Months |\n",
    "  \n",
    "\n",
    "\n",
    "Your task is to develop a generative model, which can be used to estimate revenue for next year. Nevertheless, the model's parameters need to be identified and their underlying distributions hypothesized before their estimated values can be determined. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Reading the Data \n",
    "\n",
    "* Use `pandas` to read in the data into a variable called `sample_users`\n",
    "* Based on the data in the tables, what is the proportion of males who completed the training?\n"
   ]
  },
  {
   "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>member_id</th>\n",
       "      <th>gender</th>\n",
       "      <th>status</th>\n",
       "      <th>mem_duration</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7Q4IWMC7DE</td>\n",
       "      <td>F</td>\n",
       "      <td>C</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>IZO8chWgJZ</td>\n",
       "      <td>M</td>\n",
       "      <td>C</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>VKwFA5VINu</td>\n",
       "      <td>F</td>\n",
       "      <td>C</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EfDYcr3Pn8</td>\n",
       "      <td>F</td>\n",
       "      <td>C</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>lW1u20Myap</td>\n",
       "      <td>M</td>\n",
       "      <td>C</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130056</th>\n",
       "      <td>mJ911HAbHB</td>\n",
       "      <td>M</td>\n",
       "      <td>Q</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130057</th>\n",
       "      <td>87xWLwwKp2</td>\n",
       "      <td>M</td>\n",
       "      <td>Q</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130058</th>\n",
       "      <td>hjIO8alt8e</td>\n",
       "      <td>M</td>\n",
       "      <td>Q</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130059</th>\n",
       "      <td>fj0X08835V</td>\n",
       "      <td>F</td>\n",
       "      <td>Q</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130060</th>\n",
       "      <td>iI3PTPulc0</td>\n",
       "      <td>M</td>\n",
       "      <td>Q</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>130061 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         member_id gender status  mem_duration\n",
       "0       7Q4IWMC7DE      F      C            10\n",
       "1       IZO8chWgJZ      M      C             6\n",
       "2       VKwFA5VINu      F      C            15\n",
       "3       EfDYcr3Pn8      F      C            13\n",
       "4       lW1u20Myap      M      C            10\n",
       "...            ...    ...    ...           ...\n",
       "130056  mJ911HAbHB      M      Q             4\n",
       "130057  87xWLwwKp2      M      Q             3\n",
       "130058  hjIO8alt8e      M      Q            11\n",
       "130059  fj0X08835V      F      Q             3\n",
       "130060  iI3PTPulc0      M      Q             4\n",
       "\n",
       "[130061 rows x 4 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_users = pd.read_csv('https://www.dropbox.com/s/evaaiu1mevw6t3j/memberships_info.csv?dl=1')\n",
    "sample_users"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gender  status\n",
       "M       C         45976\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "male_users = sample_users.value_counts(['gender', 'status'])\n",
    "male_users = male_users.drop(male_users.index[[1,2,3]])\n",
    "male_users"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Generative Model for our Simulations\n",
    "\n",
    "Below is a description of the generative model we will use. Our model will simulate yearly sales in 4 steps:\n",
    "\n",
    "  1. Simulate a likely number of enrollees based on $\\mu=13$ and $\\sigma =1.4$ increase from last year (2021).\n",
    "  \n",
    "  2. You can simulate whether a particular enrollee will:\n",
    "  \n",
    "    2.1 Quit the training: In which case, generate a likely value representing the number of months of enrollment before quitting\n",
    "      \n",
    "    2.2 Complete the training: in which case, generate a likely value representing the number of months of enrollment before completing the training\n",
    "\n",
    "  3. Compute the revenue from memberships based on the total number of months of enrollment simulated for each member.\n",
    "\n",
    "* This strategy is shown in the figure below"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"https://www.dropbox.com/s/xvvof9g36m7pthq/generative_model_diagram.png?dl=1\" alt=\"drawing\" style=\"width:700px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Estimating the 95% Confidence Interval of the Simulation\n",
    "\n",
    "* We are not interested in one value, but rather the interval in which the revenue falls 95% of the time. Therefore, we need to repeat the experiment above a large number of times (say 5,000) to derive the 95% confidence interval.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  1. Generating the Number of Enrollees\n",
    "\n",
    " * As mentioned earlier, the company that provides the training expects the increase in enrollments in 2022 to be comparable to annual increases in the past. The increase has, therefore, a mean $\\mu=13$% and a standard deviation of $\\sigma =1.4$% \n",
    " * Given that last year's number of enrollees was 90,316, use the Gaussian (normal) distribution to sample a likely number of enrollees for 2022. This number will represent the number of enrollees in a single iteration of the generative model\n",
    "\n",
    "* Save the value in a variable called `nb_enrollees`\n",
    "  * Recall that you will need to convert the resulting value to an `int` since the gaussian sample generated is a continuous values  \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of predicted enrollees are:  100973 and the type is:  int64\n"
     ]
    }
   ],
   "source": [
    "# Analyze the normal distribution using a mean of 13 % = 0.13, Standard Deviation = 1.4\n",
    "normal_dist = sp.stats.norm(.13, 1.4)\n",
    "total_num = 0\n",
    "# Experiment repeated a large number of times (5,000) to determine the confidence interval\n",
    "for _ in range(5000):\n",
    "    total_num = total_num + normal_dist.rvs(1)\n",
    "x = (1 + (total_num/5000)) * 90316\n",
    "d = np.int_(x)\n",
    "nb_enrollees = d[0]\n",
    "print('The number of predicted enrollees are: ', nb_enrollees, 'and the type is: ', nb_enrollees.dtype)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute the Proportion of People who Quit Training in the Population of Enrollees\n",
    "\n",
    "* Use the data in `sample_users` to build a 98% confidence interval of the proportion of enrollees who quit the training before completing. You can generate a single instance of the rate of people who quit by dividing by the number of  people who quit (`status == \"Q\"` in the data) by the number of people who complete the program (`status == \"C\"` in the data)\n",
    "\n",
    "* Will build a 98% confidence interval by computing the proportion in 10,000 bootstrap iteration of the data\n",
    "\n",
    "* Assing the the the confidence interval tuple to a variable called `quit_prop_conf_int`\n",
    "\n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a confidence interval tuple and assign it to a variable called quit_prop_conf_int\n",
    "quit_prop_conf_int = ()\n",
    "# Tuple of the instance people quit\n",
    "quit_num = ()\n",
    "\n",
    "# Range is generated by computing the proportion in 10,000 bootstrap iteration of the data\n",
    "for _ in range (10000):\n",
    "    # Creation of a single instance o the rate of people who quit (status == Q / status == C)\n",
    "    quit_num = (sample_users.status.value_counts()['Q'] / sample_users.status.value_counts()['C'])\n",
    "    # Populate this value to compute the proportion in the data\n",
    "    quit_prop_conf_int = quit_prop_conf_int + (quit_num, )\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "### 2. Generating a Radom Quitting Probability\n",
    "\n",
    "Approximately 98% of the time, the quitting parameter is within the 98% confidence interval. In order to incorporate this uncertainty in our model, we will randomly pick a quitting probability from this confidence interval and use it to estimate the number of people who quit each in iteration our our simualtion\n",
    " \n",
    "  * Use `np.random.uniform(min, max)` to randomly pick a value in the interval.\n",
    "  * ex. if the bounds of your quit_prop_conf_int are (0.2, 0.3), a possible value is \n",
    "  \n",
    "```python  \n",
    "  np.random.uniform(0.2, 0.3)\n",
    "  >>> 0.225\n",
    "  # this is equivalent to \n",
    "  # np.random.uniform(quit_prop_conf_int[0], quit_prop_conf_int[1])\n",
    "```\n",
    "\n",
    "Save the value in a variable called `quit_prop`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The quitting probability is:  0.5625007508499622\n"
     ]
    }
   ],
   "source": [
    "# Randomly pick a value in the interval, bounded by quit_prop_conf_int\n",
    "quit_prop = np.random.uniform(quit_prop_conf_int[0], quit_prop_conf_int[1])\n",
    "print('The quitting probability is: ', quit_prop)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Simulating the Number for People Who Quit and Those Who Complete the Training\n",
    "\n",
    "* Use the variables `nb_enrollees` and `quit_param` to determine the number of people who will complete the training and the number of enrollees who will quit.\n",
    "\n",
    "* Save the value in `nb_enrollees_finish` and `nb_enrollees_quit` \n",
    "\n",
    "  * Recall that the resulting values are integers not floats.\n",
    "  * Verify that\n",
    "```\n",
    "nb_enrollees_finish + nb_enrollees_quit == nb_enrollees\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of employees that quit are:  56797\n",
      "The number of enrollees that finish are:  44176\n",
      "The number of enrollees are:  100973\n"
     ]
    }
   ],
   "source": [
    "# The number of enrolleees that are determined to quit\n",
    "nb_enrollees_quit = round(nb_enrollees * quit_prop)\n",
    "# The amount of enrollees that completed the training\n",
    "nb_enrollees_finish = nb_enrollees - nb_enrollees_quit\n",
    "val = nb_enrollees_finish + nb_enrollees_quit\n",
    "print('The number of employees that quit are: ', nb_enrollees_quit)\n",
    "print('The number of enrollees that finish are: ', nb_enrollees_finish)\n",
    "print('The number of enrollees are: ',val)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Determining the Number of Payments Enrollees Make Before Quitting\n",
    "\n",
    "* Here we will be modeling the number of months that a person remains enrolled before eventually quitting\n",
    "\n",
    "* Recall that the probability of dropping declines as the students get closer to completing the training\n",
    "  * A person is more likely to quit in the first month than in any other month \n",
    "  * The probability of unenrolling during month $t$ is always higher than the probability of unenrolling during month $t+1$\n",
    "  * The longer the student has paid, the less likely they are to drop out\n",
    "  \n",
    " \n",
    "* The illustrations A, B and C below are all valid representations of the probability distribution that satisfy the constraints set above.\n",
    "  * For illustration purposes, the function is represented in the range [1, 10].\n",
    "\n",
    "<img src=\"https://www.dropbox.com/s/ys12ubh0mvmyji5/decay.png?dl=1\" alt=\"drawing\" style=\"width:1400px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Identifying the Most Likely Parameters\n",
    "\n",
    "* This three figure above (A, B and C) all represent exponential probability distributions\n",
    "  https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.expon.html\n",
    " \n",
    "* While we haven't covered this distribution in class, all we need to know is that this distribution has two parameters:\n",
    "  * `loc`: represents the value at which the function has its maximum\n",
    "  * `scale`: represents the \"inverse rate of decay\"\n",
    "    * the larger the scale, the slower the function is to reach a probability of 0.\n",
    "  \n",
    "* In the example above, all three figures were generated with a `loc` of 1 since enrollees are most likely to quit during the first month than in any other month\n",
    "* Figures A, B and C have different scale parameters; decays (`scale`) of `0.2`, `1` and `200` respectively\n",
    "\n",
    "* Examples of how to use the exponential probability distribution (`expon`) are given below\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fce505cc350>]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1584x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(22,4))\n",
    "x = np.linspace(0, 10, 1000)\n",
    "\n",
    "y_dist_1 = stats.expon(1, 1)\n",
    "y_dist_2 = stats.expon(1, 4)\n",
    "\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(x, y_dist_1.pdf(x), label=\"$a=0.2$\")\n",
    "\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(x, y_dist_2.pdf(x), label=\"$a=1$\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2.31382319, 1.25261612, 1.04335183, 3.92837972, 8.75095603,\n",
       "       4.72863857, 4.99939378])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sampling 7 values from stats.expon with `loc=1` and `scale=2`\n",
    "\n",
    "stats.expon(1, 2).rvs(7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Identifying the Scale Parameter that Maximizes the Likelihood (ML)\n",
    "\n",
    "* We will set `loc=1` since enrollees are likely to quit during the first month. However, what value of `scale` maximizes the log-likelihood of our data?\n",
    "  \n",
    "  \n",
    "* Before we answer this question, compute the log-likelihood of this dataset using a `scale` value of 5\n",
    " * Recall that the obsersed data here is the number of months a person remains enrolled before eventually quitting\n",
    "   * Make sure that your data is the `mem_duration` for individuals with `status==\"Q\"`\n",
    "\n",
    "* Also, to compute the log-likelihood of the data [2,1,2,3,4] and `location=1` and `scale=5`, we write:\n",
    "\n",
    "\n",
    "```python\n",
    "    log_pdfs = stats.expon(1, 4).logpdf([2,1,2,3,4])\n",
    "    sum(log_pdfs)\n",
    "    # or simply \n",
    "    # sum(stats.expon(1, 4).logpdf([2,1,2,3,4]))\n",
    "```\n",
    "\n",
    "* Note that above we chain operations by first building the distribution using `stats.expon(1, 4)` and, which we subsequently use to call the method `logpdf([2,1,2,3,4])`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-127237.50193603487"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mem duration is checked for status == \"Q\"\n",
    "num = sample_users[sample_users['status'] == 'Q']\n",
    "# Change the value to a numpy array\n",
    "sample_num = num.mem_duration.to_numpy()\n",
    "# Compute the log-likelihood of the dataset using a scale value of 5\n",
    "sum(stats.expon(1, 5).logpdf(sample_num.tolist()))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Estimating Scale Using the Maximum Likelihood\n",
    "\n",
    "* Compute and plot the log-likelihood of `scale` values between 1 and 50 given the data\n",
    "\n",
    "* Your graph should look like the one below.\n",
    "\n",
    "  * The X-axis represents possible scale params\n",
    "  * The Y-axis represents the log-likelihood of your data\n",
    "\n",
    "<img src=\"https://www.dropbox.com/s/qo5t4qck68pbads/ML.png?dl=1\" alt=\"drawing\" style=\"width:1400px;\"/>\n",
    "\n",
    "* Notes:\n",
    "  * I have marked the graph to indicate where likelihood is maximal\n",
    "  * The x-axis ticks are intentionally missing. Make sure yours are not.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fceb13b3790>]"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1728x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "arr = []\n",
    "val = np.linspace(1,20,50)\n",
    "nums = 1\n",
    "\n",
    "for _ in range(50):\n",
    "    arr.append(sum(stats.expon(1,nums).logpdf(sample_num)))\n",
    "    nums = nums + 1\n",
    "\n",
    "# Plotting \n",
    "plt.figure(figsize = (24, 8))\n",
    "plt.title('Maximum Likelihood Scale')\n",
    "plt.xlabel('Scale')\n",
    "plt.ylabel('Log-Likelihood')\n",
    "plt.plot(val, arr)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The log-likelihood of scale is:  0.25\n"
     ]
    }
   ],
   "source": [
    "scale = x[arr.index(max(arr))]\n",
    "print('The log-likelihood of scale is: ', scale)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### 4. Simulating Users Unenrolling from the Training\n",
    "\n",
    "* Simulate the number of months users who quit are likely to have stayed on the platform before quitting.\n",
    "\n",
    "* This can be done by sampling from the exponential using the `loc=1` and the `scale` parameter identified above. E.g.,. if your scale is 4, you can use: \n",
    "\n",
    "```python \n",
    "    scale =4\n",
    "    nb_users = 1\n",
    "    stats.expon(1, scale).rvs(nb_users)\n",
    "    array([ 2.68001742])\n",
    "    \n",
    "```\n",
    "\n",
    "* The above value tells us the user we simulated stayed on for 2.6 months before quitting. \n",
    "  *  To simulate multiple users, simply change the value of nb_users\n",
    "\n",
    "Since users pay monthly, you can round the values up to the nearest integer\n",
    "  *I.e., use `np.ceil` to determine the ceiling\n",
    "\n",
    "* Simulate the number of months enrolled for all users in `nb_enrollees_quit`  \n",
    " * Save the value to a variable called `months_paid_before_quitting`.   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of months that were paid before quitting is:  6.0\n"
     ]
    }
   ],
   "source": [
    "# Using a scale of 4 (This value can change depending on the simulation)\n",
    "scale = 4\n",
    "nb_users = 1\n",
    "# Simulate the number of months enrolled for all users in nb_enrollees quit\n",
    "months_paid_before_quitting = ((stats.expon(1,scale).rvs(nb_enrollees_quit)))\n",
    "# Save the value to a variable called months_paid_before_quitting\n",
    "months_paid_before_quitting = np.ceil(np.mean(months_paid_before_quitting))\n",
    "# Printed value\n",
    "print('The number of months that were paid before quitting is: ',months_paid_before_quitting)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modeling the Membership Duration for Enrollees who Complete the Training\n",
    "\n",
    "* We will use a similar approach to derive the duration of enrollment for users who completed the training\n",
    "\n",
    "  * Here we are not concerned by the gender difference.\n",
    "    \n",
    "* Plot a KDE of the duration of enrollment for enrollees who finished the training. \n",
    "  * Your graph should look like the following. \n",
    "\n",
    "<img src=\"https://www.dropbox.com/s/it164wz91kr44cz/kde.png?dl=1\" alt=\"drawing\" style=\"width:600px;\"/>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fcea1a1fe10>]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get the value from those who completed the training\n",
    "completed_training = sample_users[sample_users['status'] == 'C']\n",
    "# Get the mean of those who completed training\n",
    "mean = np.mean(completed_training.mem_duration)\n",
    "# Get the standard deviation of those that completed training\n",
    "std = np.std(completed_training.mem_duration)\n",
    "x = np.random.normal(mean, std, nb_enrollees_finish)\n",
    "# Get the Gaussian\n",
    "kde_evaluate = sp.stats.gaussian_kde(x, bw_method = 0.4)\n",
    "# Evaluate and arange the axis\n",
    "axis = np.arange(0, 25, 1)\n",
    "# Get the Kernel Densitu\n",
    "kernel_density = kde_evaluate(axis)\n",
    "\n",
    "'''\n",
    "The following code is to plot the KDE of the duration of enrollment for enrolleees who finished the training\n",
    "'''\n",
    "plt.figure(figsize=(18,8))\n",
    "plt.title('Membership Duration Model for Those Who Completed Training')\n",
    "plt.xlabel('Months')\n",
    "plt.ylabel('Percentage')\n",
    "plt.plot(axis, kernel_density)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visually identifying mean training completion time\n",
    "\n",
    "* Based on the KDE above, how long does it take, on average, to complte the training?\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Based on the KDE above, on average, it takes approximately 10 months to complete the training. We can also derive this answer programmatically, by evaluating the `mean` of this curve. This can be depicted in the following cell. \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "On average, it takes about 10.0  months to complete the training\n"
     ]
    }
   ],
   "source": [
    "print('On average, it takes about', np.ceil(mean), ' months to complete the training')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modeling the Membership Duration for Enrollees who Complete the Training - Cont'd \n",
    "\n",
    "\n",
    "* Searching online for similar-looking probability distributions, you find that your KDE resembles a gamma distribution. \n",
    "  \n",
    "   https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.gamma.html\n",
    "\n",
    "* The gamma distribution takes a single parameter $a$ called the shape parameter.\n",
    "  * See the documentation for `stats.gamma` in the `scipy` package\n",
    "\n",
    "* Below are a couple of examples of usage of `stats.gamma`. The first was generated with $a=2$ and the second with $a=8$\n",
    "  \n",
    "<img src=\"images/two_gammas.png\" alt=\"drawing\" style=\"width:1000px;\"/>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Computing the Maximum Likelihood Value of the Parameter $a$\n",
    "\n",
    "* Use the Maximum Likelihood to identify the value of the parameter $a$ that maximizes the log-likelihood \n",
    "    \n",
    "  * Compute and plot the log-likelihood for the values between 1 and 25\n",
    "  \n",
    "    * The X-axis represents possible distribution params\n",
    "    * The Y-axis represents the log-likelihood of your data\n",
    "  * Recall that tour data consists of the `mem_duration` for users where `status == \"C\"`\n",
    "  \n",
    "* Notes:\n",
    "  * I clearly marked the graph to indicate where likelihood is maximal\n",
    "  * The x-axis ticks are intentionally missing. Make sure yours are not.\n",
    "  * You can  use the `logpdf` directly on the `gamma(x)` object \n",
    "  \n",
    "<img src=\"https://www.dropbox.com/s/dahd99mz258curq/ML_gamma.png?dl=1\" alt=\"drawing\" style=\"width:800px;\"/>\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fced1daf890>]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1584x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Get the value from those who completed the training\n",
    "completed_training = sample_users[sample_users['status'] == 'C']\n",
    "sample_completed_training = completed_training.mem_duration.to_numpy()\n",
    "plotted_val = sp.stats.gamma.logpdf(np.linspace(0,20,25), 9, 0, 1)\n",
    "\n",
    "'''\n",
    "The following code is to plot the ML Value of the Parameter a\n",
    "'''\n",
    "plt.figure(figsize=(22,7))\n",
    "plt.title('Maximum Likelihood Value of the Paramter a')\n",
    "plt.xlabel('Duration')\n",
    "plt.ylabel('Log-Likelihood')\n",
    "plt.locator_params(axis=\"x\", nbins=25)\n",
    "plt.plot(np.linspace(0,20,25), plotted_val)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Computing the Maximum Likelihood Value of the Parameter $a$ - Cont'd\n",
    "\n",
    "\n",
    "* What is the value that maximizes the likelihood?\n",
    "* Does this value align with what you observed after inspecting rhe KDE graph above?\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "The value that maximizes the likehood of the parameter is 8. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "The value that maximizes the likelihood does not align with the value that is observed from the KDE graph generated and observed. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Simulating Users Completing the Training\n",
    "\n",
    "\n",
    "* Using the gamma distribution and the parameter $a$ determined above, simulate the number of months it takes users to finish the training. To simulate the number of months a user may have remained unrolled before completing the training we simply sample from the gamma using the parameter $a$ identified above.\n",
    "\n",
    "* E.g., for $a = 5$, you can use: \n",
    "\n",
    "```python \n",
    "    scale =4\n",
    "    nb_users = 1\n",
    "    stats.gamma(1, months_paid_before_completion).rvs(nb_users)\n",
    "    array([ 5.112342])\n",
    "    \n",
    "```\n",
    "\n",
    "* According to the value above, the simulated student stayed on for 5.1 months before quitting \n",
    "  * Change the value of `nb_users` to simulate multiple users\n",
    "\n",
    "* Since users pay monthly, round up the values to the nearest integer\n",
    "  * I.e., compute the ceiling using `np.ceil`\n",
    "\n",
    "* Simulate the number of months enrolled for all users in `nb_enrollees_finish`  \n",
    " * Save the values to a variable called `months_paid_before_finishing`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of months paid before finishing is:  10.0\n"
     ]
    }
   ],
   "source": [
    "num = np.mean(np.ceil(stats.gamma(1, 8).rvs(nb_enrollees_finish)))\n",
    "months_paid_before_finishing = np.ceil(num)\n",
    "print('The number of months paid before finishing is: ', months_paid_before_finishing)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6. Simulating Revenue for a Single Iteration\n",
    "\n",
    "* Compute the total revenue for this iteration as the total number of months paid by users who completed the training and those who quit, times the monthly fee.\n",
    "  * The total number of months of enrollement for users who completed the training is saved in `months_paid_before_finishing`\n",
    "  * The total number of months of enrollement for users who did not complete the training is saved in `months_paid_before_quitting`\n",
    "  * The monthly fee is 39.99\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The total revenue from the users who finished is:  399.90000000000003\n",
      "The total revenue from the users who quit is:  239.94\n"
     ]
    }
   ],
   "source": [
    "tot_rev_finished = months_paid_before_finishing *  39.99\n",
    "tot_rev_quitting = months_paid_before_quitting *  39.99\n",
    "print('The total revenue from the users who finished is: ', tot_rev_finished)\n",
    "print('The total revenue from the users who quit is: ', tot_rev_quitting)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build a 95% Confidence Interval for the Revenue\n",
    "\n",
    "* Repeat the simulation above 5000 times to build a confidence interval. The steps you'll need to repeat are those described in text cells with the following headers:\n",
    "    \n",
    "  1. Generating the Number of Enrollees\n",
    "  2. Generating a Radom Quitting Probability \n",
    "  3. Simulating the Number for People Who Quit and Those Who Complete the Training \n",
    "  4. Simulating Users Unenrolling from the Training\n",
    "  5. Simulating Users Completing the Training\n",
    "  6. Simulating Revenue for a Single Iteration\n",
    "\n",
    "\n",
    "    \n",
    "* The computation may take a few minutes to complete"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Create an array of the confidence interv al to place a min and max value in\n",
    "conf_interval = []\n",
    "# Repeat the simulation 5000 times\n",
    "for _ in range(5000):\n",
    "    # Generate the number of enrollees\n",
    "    num_enrollees = np.random.uniform(90000, 100000)\n",
    "    # Generate a random quitting probability\n",
    "    nb_enrollees_quit = num_enrollees * quit_prop\n",
    "    nb_enrollees_quit = round(nb_enrollees_quit)\n",
    "    nb_enrollees_finish = num_enrollees - nb_enrollees_quit\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Taking into account gender differences \n",
    "\n",
    "* Repeat the analysis above while taking into consideration the differences between males and females. To do so, you'll need two additional parameters:\n",
    "\n",
    "  * the proportion of males/females in the population of enrollees who quit the training, call it `male_female_prop_quit`\n",
    "     * This is the parameter that you'll use to decide whether the person who unenrolls is a Male or Female\n",
    "\n",
    "  * The proportion of males to females in the population of enrollees who complete the training, call it `male_female_prop_finish`\n",
    "    * This is the parameter that you'll use to decide whether the person who unenrolls is a man or a woman.\n",
    "\n",
    "* Rather than having a single parameter `scale` for the inverse decay of the exponential distribution, you will now have two parameters, one for males and a second for females\n",
    "* Rather than having a single shape parameter $a$ for gamma distribution, you will now have two parameters, one for males and a second for females\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Proportion of female:  0.7051366091562221\n",
      "Proportion of male:  0.2948633908437779\n"
     ]
    }
   ],
   "source": [
    "# Create a data structure that holds the males and the females\n",
    "males = []\n",
    "females = []\n",
    "# Separate the males and females using the data from sample_users\n",
    "prop_female = sample_users.gender.value_counts()['F'] / sample_users.gender.value_counts()['M']\n",
    "# To find the alternate, get the difference between these values\n",
    "prop_male = 1 - prop_female\n",
    "# Print out statements to check if they add up correctly\n",
    "print('Proportion of female: ',prop_female)\n",
    "print('Proportion of male: ',prop_male)\n",
    "# Simulate for the 5000 range\n",
    "for _ in range(5000):\n",
    "    # Simulate the number of enrollees\n",
    "    num_enrollees = np.random.uniform(90000, 100000)\n",
    "    # Simulate the number of enrollees who quit\n",
    "    nb_enrollees_quit = round(num_enrollees * quit_prop)\n",
    "    # Male proportion quit\n",
    "    male_prop_quit = nb_enrollees_quit * prop_male\n",
    "    # Female proportion quit\n",
    "    female_prop_quit = nb_enrollees_quit * prop_female\n",
    "    # Simulate the number of those who completed the training\n",
    "    nb_enrollees_finish = num_enrollees - nb_enrollees_quit\n",
    "    # Male proportion quit\n",
    "    male_prop_finish = nb_enrollees_finish * prop_male\n",
    "    # Female proportion quit\n",
    "    female_prop_finish = nb_enrollees_finish * prop_female\n",
    "    # Calculate the revenue associatied for males who quit\n",
    "    male_revenue_finished = male_prop_finish * months_paid_before_finishing * 39.99\n",
    "    male_revenue_quit = male_prop_quit * months_paid_before_quitting * 39.99\n",
    "    # Calculate the revenue associated for the females who quit\n",
    "    female_revenue_finished = female_prop_finish * months_paid_before_finishing * 39.99\n",
    "    female_revenue_quit = female_prop_quit * months_paid_before_quitting * 39.99\n",
    "    # Create the two parameters for the gamme distribution of males and females\n",
    "    males.append(male_revenue_finished + male_revenue_quit)\n",
    "    females.append(female_revenue_finished + female_revenue_quit)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Conclusion\n",
    "\n",
    "\n",
    "* Which do you think is the most accurate model, the one that takes into account the gender differences of the one that does not? Justify your answer.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Since the first model of this computation is based on computations prior to its augmentation, the most accurate model is the one that takes this into consideration. In conclusion, the accurate model is the one that does not take into account gender differences. \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PROBLEM 2\n",
    "----"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Average Medicare Payment and Hypothesis Testing \n",
    "\n",
    "* We have access to the average Medicare payment made by the Federal government to over 6,000 hospitals nationwide. Medicare payments are categorized according to `DRG Definitions`, which are diagnostic categories that describe the service provided. For instance, the average Medicare payment at  THE QUEENS MEDICAL CENTER for the diagnostic category \"917 - POISONING & TOXIC EFFECTS OF DRUGS W MCC\" was $15,305.21.\n",
    "\n",
    "* We are interested in comparing the average payment across states to see if they are similar or not, regardless of diagnostic category. Naturally, by different, we are interested in seeing whether the difference is statistically significant.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "\n",
    "### Loading  the data\n",
    "* The data is stored in `hospital_charge_data.csv` available at the following [URL](https://www.dropbox.com/s/zqf56enddz6mhlf/hospital_charge_data.csv?dl=1)\n",
    "\n",
    "\n",
    "* Load the data into a variable called `payment_data`\n",
    "  * Select only the columns `['DRG Definition', \"Provider Name\", 'Provider State', 'Average Medicare Payments']`\n",
    "  * Rename these columns as `['diagnostic', 'name', 'state', 'avg_medicare_payment']`\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "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>DRG Definition</th>\n",
       "      <th>Provider Id</th>\n",
       "      <th>Provider Name</th>\n",
       "      <th>Provider Street Address</th>\n",
       "      <th>Provider City</th>\n",
       "      <th>Provider State</th>\n",
       "      <th>Provider Zip Code</th>\n",
       "      <th>Hospital Referral Region Description</th>\n",
       "      <th>Total Discharges</th>\n",
       "      <th>Average Covered Charges</th>\n",
       "      <th>Average Total Payments</th>\n",
       "      <th>Average Medicare Payments</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>10001</td>\n",
       "      <td>SOUTHEAST ALABAMA MEDICAL CENTER</td>\n",
       "      <td>1108 ROSS CLARK CIRCLE</td>\n",
       "      <td>DOTHAN</td>\n",
       "      <td>AL</td>\n",
       "      <td>36301</td>\n",
       "      <td>AL - Dothan</td>\n",
       "      <td>91</td>\n",
       "      <td>32963.07</td>\n",
       "      <td>5777.24</td>\n",
       "      <td>4763.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>10005</td>\n",
       "      <td>MARSHALL MEDICAL CENTER SOUTH</td>\n",
       "      <td>2505 U S HIGHWAY 431 NORTH</td>\n",
       "      <td>BOAZ</td>\n",
       "      <td>AL</td>\n",
       "      <td>35957</td>\n",
       "      <td>AL - Birmingham</td>\n",
       "      <td>14</td>\n",
       "      <td>15131.85</td>\n",
       "      <td>5787.57</td>\n",
       "      <td>4976.71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>10006</td>\n",
       "      <td>ELIZA COFFEE MEMORIAL HOSPITAL</td>\n",
       "      <td>205 MARENGO STREET</td>\n",
       "      <td>FLORENCE</td>\n",
       "      <td>AL</td>\n",
       "      <td>35631</td>\n",
       "      <td>AL - Birmingham</td>\n",
       "      <td>24</td>\n",
       "      <td>37560.37</td>\n",
       "      <td>5434.95</td>\n",
       "      <td>4453.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>10011</td>\n",
       "      <td>ST VINCENT'S EAST</td>\n",
       "      <td>50 MEDICAL PARK EAST DRIVE</td>\n",
       "      <td>BIRMINGHAM</td>\n",
       "      <td>AL</td>\n",
       "      <td>35235</td>\n",
       "      <td>AL - Birmingham</td>\n",
       "      <td>25</td>\n",
       "      <td>13998.28</td>\n",
       "      <td>5417.56</td>\n",
       "      <td>4129.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>10016</td>\n",
       "      <td>SHELBY BAPTIST MEDICAL CENTER</td>\n",
       "      <td>1000 FIRST STREET NORTH</td>\n",
       "      <td>ALABASTER</td>\n",
       "      <td>AL</td>\n",
       "      <td>35007</td>\n",
       "      <td>AL - Birmingham</td>\n",
       "      <td>18</td>\n",
       "      <td>31633.27</td>\n",
       "      <td>5658.33</td>\n",
       "      <td>4851.44</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             DRG Definition  Provider Id  \\\n",
       "0  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC        10001   \n",
       "1  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC        10005   \n",
       "2  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC        10006   \n",
       "3  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC        10011   \n",
       "4  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC        10016   \n",
       "\n",
       "                      Provider Name     Provider Street Address Provider City  \\\n",
       "0  SOUTHEAST ALABAMA MEDICAL CENTER      1108 ROSS CLARK CIRCLE        DOTHAN   \n",
       "1     MARSHALL MEDICAL CENTER SOUTH  2505 U S HIGHWAY 431 NORTH          BOAZ   \n",
       "2    ELIZA COFFEE MEMORIAL HOSPITAL          205 MARENGO STREET      FLORENCE   \n",
       "3                 ST VINCENT'S EAST  50 MEDICAL PARK EAST DRIVE    BIRMINGHAM   \n",
       "4     SHELBY BAPTIST MEDICAL CENTER     1000 FIRST STREET NORTH     ALABASTER   \n",
       "\n",
       "  Provider State  Provider Zip Code Hospital Referral Region Description  \\\n",
       "0             AL              36301                          AL - Dothan   \n",
       "1             AL              35957                      AL - Birmingham   \n",
       "2             AL              35631                      AL - Birmingham   \n",
       "3             AL              35235                      AL - Birmingham   \n",
       "4             AL              35007                      AL - Birmingham   \n",
       "\n",
       "    Total Discharges    Average Covered Charges    Average Total Payments   \\\n",
       "0                  91                   32963.07                   5777.24   \n",
       "1                  14                   15131.85                   5787.57   \n",
       "2                  24                   37560.37                   5434.95   \n",
       "3                  25                   13998.28                   5417.56   \n",
       "4                  18                   31633.27                   5658.33   \n",
       "\n",
       "   Average Medicare Payments  \n",
       "0                    4763.73  \n",
       "1                    4976.71  \n",
       "2                    4453.79  \n",
       "3                    4129.16  \n",
       "4                    4851.44  "
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('https://www.dropbox.com/s/zqf56enddz6mhlf/hospital_charge_data.csv?dl=1')\n",
    "data.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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>diagnostic</th>\n",
       "      <th>name</th>\n",
       "      <th>state</th>\n",
       "      <th>avg_medicare_payment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>SOUTHEAST ALABAMA MEDICAL CENTER</td>\n",
       "      <td>AL</td>\n",
       "      <td>4763.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>MARSHALL MEDICAL CENTER SOUTH</td>\n",
       "      <td>AL</td>\n",
       "      <td>4976.71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>ELIZA COFFEE MEMORIAL HOSPITAL</td>\n",
       "      <td>AL</td>\n",
       "      <td>4453.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>ST VINCENT'S EAST</td>\n",
       "      <td>AL</td>\n",
       "      <td>4129.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>SHELBY BAPTIST MEDICAL CENTER</td>\n",
       "      <td>AL</td>\n",
       "      <td>4851.44</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 diagnostic                              name  \\\n",
       "0  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC  SOUTHEAST ALABAMA MEDICAL CENTER   \n",
       "1  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC     MARSHALL MEDICAL CENTER SOUTH   \n",
       "2  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC    ELIZA COFFEE MEMORIAL HOSPITAL   \n",
       "3  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC                 ST VINCENT'S EAST   \n",
       "4  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC     SHELBY BAPTIST MEDICAL CENTER   \n",
       "\n",
       "  state  avg_medicare_payment  \n",
       "0    AL               4763.73  \n",
       "1    AL               4976.71  \n",
       "2    AL               4453.79  \n",
       "3    AL               4129.16  \n",
       "4    AL               4851.44  "
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Select specific columns \n",
    "spec_columns = data[['DRG Definition', \"Provider Name\", 'Provider State', 'Average Medicare Payments']]\n",
    "# Rename the columns as necessary\n",
    "payment_data = spec_columns.rename(columns={'DRG Definition': 'diagnostic',\"Provider Name\": 'name', 'Provider State': 'state', 'Average Medicare Payments':'avg_medicare_payment'})\n",
    "# Inspection of the data\n",
    "payment_data.head(5)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Average Medicare Payment Value\n",
    "\n",
    "* Compute the mean average medicare payment value for Idaho (ID) and Hawaii (HI). \n",
    "\n",
    "* Based on the observed means, do you believe that there is a statistically significant difference in medicare payments between the two states? Justify your answer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hawaii Average:  avg_medicare_payment    10967.475045\n",
      "Name: HI, dtype: float64\n",
      "Idaho Average:  avg_medicare_payment    8461.977514\n",
      "Name: ID, dtype: float64\n",
      "The difference in Medicaid between Hawaii and Idaho are:  avg_medicare_payment    2505.497532\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# Write you code here\n",
    "mean_average = payment_data.groupby(['state']).mean()\n",
    "hawaii_average = mean_average.loc['HI']\n",
    "idaho_average = mean_average.loc['ID']\n",
    "\n",
    "print('Hawaii Average: ', hawaii_average)\n",
    "print('Idaho Average: ', idaho_average)\n",
    "print('The difference in Medicaid between Hawaii and Idaho are: ', hawaii_average - idaho_average)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on the observed means, I believe that there is a statistically significant difference in medicare payments between the two states. This is because on average, there seems to be approximately a $2505 difference between the two states when comparing the mean of the average medicare payment."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# FYI: Which are the Costliest Categories\n",
    "\n",
    "* Post-surgery infections are the categories with the highest Medicare payments cost,\n",
    "  * The most expensive category is `853 - INFECTIOUS & PARASITIC DISEASES W O.R. P....` \n",
    "    * there were 15 discharges in NY's  WESTCHESTER MEDICAL CENTER, with an average Medicare payment of $133,177. \n",
    "    \n",
    "    * there were 86 discharges in LA's CEDARS-SINAI MEDICAL CENTER with an average medicare payment of $87,437\n",
    "  \n",
    "* The overall cost of surgical site infection alone in the US is estimated at $3.5 to $10 billion annually."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Skewed Data\n",
    "\n",
    "* The payments data in Idaho and Hawaii are highly skewed and most payments are below $15k$, with some up to $60k$ in both states. See graphs below.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'avg_medicare_payment in Hawaii')"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1584x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(22,4))\n",
    "plt.subplot(1, 2, 1)\n",
    "_ = plt.hist(payment_data[payment_data[\"state\"] == \"ID\"][\"avg_medicare_payment\"])\n",
    "plt.title(\"avg_medicare_payment in ID\")\n",
    "plt.subplot(1, 2, 2)\n",
    "_ = plt.hist(payment_data[payment_data[\"state\"] == \"HI\"][\"avg_medicare_payment\"])\n",
    "plt.title(\"avg_medicare_payment in Hawaii\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Log-Transforming the Data\n",
    "\n",
    "\n",
    "* As discussed in class, data which follow bell-shaped distributions tend to be easier to work with in hypothesis testing.\n",
    "    * Bootstrapping is not recommended with skewed data.\n",
    "\n",
    "* Compute the `log` (base 10) of the `avg_medicare_payment` column and save the result to a new column called `log_avg_medicare_payment`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "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>diagnostic</th>\n",
       "      <th>name</th>\n",
       "      <th>state</th>\n",
       "      <th>avg_medicare_payment</th>\n",
       "      <th>log_avg_medicare_payment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>SOUTHEAST ALABAMA MEDICAL CENTER</td>\n",
       "      <td>AL</td>\n",
       "      <td>4763.73</td>\n",
       "      <td>3.677947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>MARSHALL MEDICAL CENTER SOUTH</td>\n",
       "      <td>AL</td>\n",
       "      <td>4976.71</td>\n",
       "      <td>3.696942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>ELIZA COFFEE MEMORIAL HOSPITAL</td>\n",
       "      <td>AL</td>\n",
       "      <td>4453.79</td>\n",
       "      <td>3.648730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>ST VINCENT'S EAST</td>\n",
       "      <td>AL</td>\n",
       "      <td>4129.16</td>\n",
       "      <td>3.615862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>039 - EXTRACRANIAL PROCEDURES W/O CC/MCC</td>\n",
       "      <td>SHELBY BAPTIST MEDICAL CENTER</td>\n",
       "      <td>AL</td>\n",
       "      <td>4851.44</td>\n",
       "      <td>3.685871</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 diagnostic                              name  \\\n",
       "0  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC  SOUTHEAST ALABAMA MEDICAL CENTER   \n",
       "1  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC     MARSHALL MEDICAL CENTER SOUTH   \n",
       "2  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC    ELIZA COFFEE MEMORIAL HOSPITAL   \n",
       "3  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC                 ST VINCENT'S EAST   \n",
       "4  039 - EXTRACRANIAL PROCEDURES W/O CC/MCC     SHELBY BAPTIST MEDICAL CENTER   \n",
       "\n",
       "  state  avg_medicare_payment  log_avg_medicare_payment  \n",
       "0    AL               4763.73                  3.677947  \n",
       "1    AL               4976.71                  3.696942  \n",
       "2    AL               4453.79                  3.648730  \n",
       "3    AL               4129.16                  3.615862  \n",
       "4    AL               4851.44                  3.685871  "
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "payment_data['log_avg_medicare_payment'] = np.log10(payment_data['avg_medicare_payment'])\n",
    "payment_data.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plotting the Hawaii KDE\n",
    "\n",
    "* Plot the `log_avg_medicare_payment` KDE for Hawaii \n",
    "    \n",
    "* Your curve should look like the one below\n",
    "\n",
    "* Notice that: \n",
    "  * The KDE is smooth (as opposed to having jagged peaks)\n",
    "   * The bandwidth was selected so that the range (x-axis) is mimized\n",
    "\n",
    "* In the example below, x- and y-axis labels are missing. Your final graph should be labeled appropriately and should have a title\n",
    "\n",
    "<img src=\"https://www.dropbox.com/s/i98rzkjiajz2by2/kde_hawaii.png?dl=1\" alt=\"drawing\" style=\"width:700px;\"/>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fce91452910>"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1584x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "hawaii_num = payment_data.state.value_counts()['HI']\n",
    "d1 = payment_data[['state', 'log_avg_medicare_payment']]\n",
    "\n",
    "# Get the mean values\n",
    "means = d1.groupby(['state'])\n",
    "HI_val = means.get_group('HI')\n",
    "HI_mean = np.mean(HI_val.log_avg_medicare_payment)\n",
    "# Get the standard deviation values\n",
    "states = d1.groupby(['state'])\n",
    "hawaii_state = states.get_group('HI')\n",
    "std = hawaii_state.log_avg_medicare_payment.std() \n",
    "# Generate the kernel density estimation\n",
    "x = np.arange(0, 10, .05)\n",
    "data = np.random.normal(HI_mean,std,num)\n",
    "kde = sp.stats.gaussian_kde(data, bw_method=2.5)\n",
    "y = kde.evaluate(x)\n",
    "\n",
    "'''\n",
    "Plotting the figure\n",
    "'''\n",
    "plt.figure(figsize=(22,8))\n",
    "plt.title('Plotting the Hawaii KDE: Log Average Medicare Payment')\n",
    "plt.xlabel('Average Log Payment')\n",
    "plt.ylabel('Density')\n",
    "plt.plot(x, y, label='Hawaii Log')\n",
    "plt.legend(['Hawaii Log Estimate'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Interpreting the Hawaii Emprical Distribution\n",
    "\n",
    "* According to the KDE plot above, what is the mean `avg_medicare_payment`? \n",
    "  * Plot the mean `avg_medicare_payment` and its density on your graph \n",
    "\n",
    "\n",
    "* Your graph should look like the one below     \n",
    "  * Note that lines in the example below are intentionally drawn at a value that may be different from the true mean  \n",
    "  * Make sure the vertical and horizontal lines in your graph match the mean and its density respectively\n",
    "  \n",
    "\n",
    "* Use the functions and `plt.vline` to plot the lines\n",
    "\n",
    "\n",
    "<img src=\"https://www.dropbox.com/s/boiakmsgxw69z1r/interpreting_kde.png?dl=1\" alt=\"drawing\" style=\"width:800px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the KDE plot above (not the sample, the one that is generated), the mean average medicare payment is approximately 4. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.LineCollection at 0x7fce71c1f050>"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1584x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the mean avg_medicare_payment and the density \n",
    "plt.figure(figsize=(22,8))\n",
    "plt.title('Plotting the Hawaii KDE: Log Average Medicare Payment')\n",
    "plt.xlabel('Average Log Payment')\n",
    "plt.ylabel('Density')\n",
    "plt.plot(x, y, label='Hawaii Log')\n",
    "plt.legend(['Hawaii Log Estimate'])\n",
    "# Use the functions plt.vline to plot the lines\n",
    "plt.vlines(HI_mean,0,.545, color='r')\n",
    "plt.hlines(.545, 0, HI_mean, color='r')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Interpreting the Hawaii Emprical Distribution\n",
    "\n",
    "* According to the `KDE` plot above, what is the probability of observing an average payment greater than $25,000\n",
    "\n",
    "  * The *KDE* object works in a similar fashion to any of the probability distributions in `scipy`\n",
    "  \n",
    "    * Note that what you are asked to compute here is not the density but the probability avg-payments > $25k\n",
    "    \n",
    "    * Since the total probability under the curve sums to 1, then the `avg_medicare_payment`  > $25k should be less than 1.\n",
    "    \n",
    "   * Also, recall that your payment values are log-transformed\n",
    "\n",
    "* Use `matplotlib`'s `fill_between` to paint the area under the curve representing that area. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PolyCollection at 0x7fce913e2a10>"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1584x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Generate a log value of something greater than 25000\n",
    "avg_payments = np.log10(25000)\n",
    "\n",
    "# Plot the figure based on previous generated fiure\n",
    "plt.figure(figsize=(22,8))\n",
    "plt.title('Plotting the Hawaii KDE: Log Average Medicare Payment')\n",
    "plt.xlabel('Average Log Payment')\n",
    "plt.ylabel('Density')\n",
    "plt.plot(x, y, label='Hawaii Log')\n",
    "plt.legend(['Hawaii Log Estimate'])\n",
    "# Fills the values in between\n",
    "plt.fill_between(avg_payments, y, color='r')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " ### Plotting Both KDEs   \n",
    "\n",
    "* Plot the Hawaii and Idaho `log_avg_medicare_payment` distributions on the same plot\n",
    "\n",
    "* Your plot should look like the following\n",
    "\n",
    "<img src=\"https://www.dropbox.com/s/tej47hhhxc18tdd/both_kdes.png?dl=1\" alt=\"drawing\" style=\"width:700px;\"/>\n",
    "\n",
    "\n",
    "* Label the x-axis and y-axis and choose a title for your plot.\n",
    "\n",
    "According to your plot, do you believe that the average payment across both states is different? Explain your answer.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hypothesis Testing\n",
    "\n",
    "* Is there a statistically significant difference between Hawaii and Idaho in terms of log_avg_medicare_payment? Here are the hypotheses we will test:\n",
    "\n",
    "* Null Hypothesis ($H_0$)\n",
    "  * The mean values across both state are the same and any differenc is due to solely sampling\n",
    "\n",
    "* Alternative Hypothesis ($H_A$)\n",
    "  * The mean values across both state are different\n",
    "    * The observed difference is not due to chance\n",
    "    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "### Exracting the Data\n",
    "\n",
    "* First, we will combine the average Medicaid payments from both states into a single `numpy` `array` called `log_avg_medicare_payment_hi_id`. This will allow us to work with the data more easily.\n",
    "\n",
    "  * Your `numpy` array should contain 993 entries\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "442\n",
      "551\n",
      "993\n"
     ]
    }
   ],
   "source": [
    "# Getting the payment value data\n",
    "payments_data = payment_data[['state', 'log_avg_medicare_payment']]\n",
    "# Grabbing the specific states payment data\n",
    "hawaii = payments_data[payments_data['state'] == 'HI']\n",
    "idaho = payments_data[payments_data['state'] == 'ID']\n",
    "# Conversion of the data to numpy arrays to concatenate\n",
    "hawaii_data = hawaii.log_avg_medicare_payment.to_numpy()\n",
    "idaho_data = idaho.log_avg_medicare_payment.to_numpy()\n",
    "#Checking individual values\n",
    "print(len(hawaii_data))\n",
    "print(len(idaho_data))\n",
    "# Combining the average medicaid payments from both states into a single numpy array\n",
    "log_avg_medicare_payment_hi_id = np.concatenate((hawaii_data, idaho_data))\n",
    "print(len(log_avg_medicare_payment_hi_id))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Computing the Observed Test Statistic\n",
    "\n",
    "* Since we are interested in the difference between both averages, i.e., avg_medicare_payment in Hawaii and Idaho, a good test statistic is, therefore, the difference between the sample means in both states. \n",
    "* Compute the observed test statistic and save it to a variable called `observed_test_stat`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The observed test statistic is:  0.12005870451646894\n"
     ]
    }
   ],
   "source": [
    "# Compute the difference between the sample means in both states\n",
    "HI_val = means.get_group('HI')\n",
    "HI_mean_val = np.mean(HI_val.log_avg_medicare_payment)\n",
    "ID_val = means.get_group('ID')\n",
    "ID_mean_val = np.mean(ID_val.log_avg_medicare_payment)\n",
    "# Save to the variable called observed_test_stat\n",
    "observed_test_stat = HI_mean_val - ID_mean_val\n",
    "print('The observed test statistic is: ', observed_test_stat)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Generating Data Under the Null Hypothesis: Single Iteration\n",
    "\n",
    "* Under the null hypothesis, we assume that Hawaii and Idaho data come from the same distribution, and therefore, have the same parameters\n",
    "\n",
    "* We can use permutations to simulate another set of samples from Hawaii and Idaho.\n",
    "\n",
    "To generate data for a single iteration we will:\n",
    "\n",
    "  1. Shuffle the values in `log_avg_medicare_payment_hi_id`\n",
    "  2. Select two new samples to represent Hawaii and Idaho. The samples must have the same sizes as in the original data.\n",
    "  3. Compute the difference between the mean `log_avg_medicare_payment` across both states.\n",
    "  4. Store the value in a variable called `perm_diff`\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The value for the permutations of the samples between these states is:  0.10903427021119283\n"
     ]
    }
   ],
   "source": [
    "# Shuffle the values in a log_avg_medicare_payment_hi_id\n",
    "np.random.shuffle(log_avg_medicare_payment_hi_id)\n",
    "# Select two new samples to represent Hawaii and Idaho, with the same sizes as orgiinal data\n",
    "new_HI = np.random.normal(HI_mean_val, std, 442)\n",
    "new_ID = np.random.normal(ID_mean_val, std, 551)\n",
    "# Compute the difference \n",
    "perm_diff = (np.mean(new_HI) - np.mean(new_ID))\n",
    "print('The value for the permutations of the samples between these states is: ', perm_diff)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Repeat the process described above 5,000 times. In each iteration save the difference in means between the two bootstrapped datasets to an array called `perm_diffs`\n",
    "* Plot the resulting distribution of the means \n",
    "  * You may use any plot type you deem appropriate for this use case\n",
    "* Use `plt.scatter` to plot in red the observed test statistic\n",
    "\n",
    "* The result should look \"somewhat\" like the graph below\n",
    "<img src=\"https://www.dropbox.com/s/doln0egikkbar52/diff_hi_id.png?dl=1\" alt=\"drawing\" style=\"width:700px;\"/>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Write you code here\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Computing the p-value for the test\n",
    "\n",
    "* Compute the p-value associated with your test statistic and state whether you have ground to reject the null hypothesis.\n",
    "  * Be sure to provide your `p-value` to justify your answer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Write you code here\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "P-values are determined be to accepting the null hypothesis when the value is large."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Hypothesis testing with values from DC\n",
    "\n",
    "* Repeat the hypothesis test above by testing Hawaii and DC. Use the first cells below, to run your code and plot the empirical bootstrap distribution and the second to answer whether the data are statistically significantly similar or now. Include the `p-value` to support your conclusion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Write you code here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "P-values are accepting the null hypothesis when the value is large. "
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "colab": {
   "collapsed_sections": [],
   "name": "PP_1_Introduction.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "hide_input": false,
  "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": "3.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}