Importance of Bayesian Testing .ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 347,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import datetime\n",
    "from scipy.stats import chi2_contingency, beta\n",
    "from IPython.display import Image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Experiment Definition\n",
    "\n",
    "We developed a new webpage and want to test it's effects on purchase conversion. As such we split our users evenly into 2 groups:\n",
    "1. **Control**: They get the old webpage\n",
    "2. **Treatment**: They get the new webpage\n",
    "\n",
    "Metric we want to track:\n",
    "$$\n",
    "\\text{purchase conversion} = \\frac{\\text{# Converted Users}}{\\text{# Exposed Users}}\n",
    "$$\n",
    "\n",
    "We have 3 weeks of logged exposure/conversion data. Let's define these terms:\n",
    "1. **Exposure**: A user is bucketed as control or treatment and sees their corresponding page for the first time in the experiment duration\n",
    "2. **Conversion**: An exposed user makes a purchase within 7 days of being first exposed\n",
    "\n",
    "Questions you should ask when setting up a test:\n",
    "- How do you think the experiment will fair?\n",
    "- Do we have actionable next steps laid out?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Data Collection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's use some A/B testing data: https://www.kaggle.com/saraabdelaal/abtestdata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 348,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('csv/ab_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "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>user_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>group</th>\n",
       "      <th>landing_page</th>\n",
       "      <th>converted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>265342</th>\n",
       "      <td>741320</td>\n",
       "      <td>2017-01-06 09:10:20.880418</td>\n",
       "      <td>treatment</td>\n",
       "      <td>new_page</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>180296</th>\n",
       "      <td>899025</td>\n",
       "      <td>2017-01-23 16:31:37.030115</td>\n",
       "      <td>control</td>\n",
       "      <td>old_page</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        user_id                   timestamp      group landing_page  converted\n",
       "265342   741320  2017-01-06 09:10:20.880418  treatment     new_page          0\n",
       "180296   899025  2017-01-23 16:31:37.030115    control     old_page          0"
      ]
     },
     "execution_count": 350,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each row is logged when user is exposed to a webpage.\n",
    "- **timestamp**: time the user is first exposed\n",
    "- **group**: bucket\n",
    "- **landing_page**: Which page are they seeing\n",
    "- **converted**: Initialized to 0. Changes to 1 if the user makes a purchase within 7 days of first exposure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 370,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of unique users in experiment: 290584\n",
      "Data collected for 21 days\n",
      "Landing pages to compare: ['old_page', 'new_page']\n",
      "Percentage of users in control: 50%\n"
     ]
    }
   ],
   "source": [
    "start_time = datetime.datetime.strptime(df['timestamp'].min(), '%Y-%m-%d %H:%M:%S.%f')\n",
    "end_time = datetime.datetime.strptime(df['timestamp'].max(), '%Y-%m-%d %H:%M:%S.%f')\n",
    "data_duration = (end_time - start_time).days\n",
    "\n",
    "print(f\"Number of unique users in experiment: {df['user_id'].nunique()}\")\n",
    "print(f\"Data collected for {data_duration} days\")\n",
    "print(f\"Landing pages to compare: {df['landing_page'].unique().tolist()}\")\n",
    "print(f\"Percentage of users in control: {round(df[df['group']=='control'].shape[0] * 100 / df.shape[0])}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Data Processing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "But there are some repeated exposures for some users"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "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>user_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>group</th>\n",
       "      <th>landing_page</th>\n",
       "      <th>converted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>29073</th>\n",
       "      <td>746755</td>\n",
       "      <td>2017-01-11 01:28:57.083669</td>\n",
       "      <td>control</td>\n",
       "      <td>new_page</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105487</th>\n",
       "      <td>722274</td>\n",
       "      <td>2017-01-19 01:46:53.093257</td>\n",
       "      <td>control</td>\n",
       "      <td>old_page</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>262554</th>\n",
       "      <td>722274</td>\n",
       "      <td>2017-01-09 21:21:23.638444</td>\n",
       "      <td>control</td>\n",
       "      <td>new_page</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>286566</th>\n",
       "      <td>746755</td>\n",
       "      <td>2017-01-05 03:40:08.457451</td>\n",
       "      <td>control</td>\n",
       "      <td>old_page</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        user_id                   timestamp    group landing_page  converted\n",
       "29073    746755  2017-01-11 01:28:57.083669  control     new_page          1\n",
       "105487   722274  2017-01-19 01:46:53.093257  control     old_page          0\n",
       "262554   722274  2017-01-09 21:21:23.638444  control     new_page          0\n",
       "286566   746755  2017-01-05 03:40:08.457451  control     old_page          0"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample = df[df['user_id'].isin([746755,722274])]\n",
    "sample"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get First timestamp for each username. In this dataset, it is the timestamp of first conversion, but in reality, make sure to do one of the following:\n",
    "1. Get timestamp of first exposure\n",
    "2. Remove users with multiple buckets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "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>user_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>group</th>\n",
       "      <th>landing_page</th>\n",
       "      <th>converted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>722274</td>\n",
       "      <td>2017-01-09 21:21:23.638444</td>\n",
       "      <td>control</td>\n",
       "      <td>new_page</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>746755</td>\n",
       "      <td>2017-01-05 03:40:08.457451</td>\n",
       "      <td>control</td>\n",
       "      <td>old_page</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id                   timestamp    group landing_page  converted\n",
       "0   722274  2017-01-09 21:21:23.638444  control     new_page          0\n",
       "1   746755  2017-01-05 03:40:08.457451  control     old_page          0"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 1. Get timestamp of first exposure\n",
    "first_conversion = sample.groupby('user_id')['timestamp'].min().to_frame().reset_index()\n",
    "sample = sample.merge(first_conversion, on=['user_id', 'timestamp'])\n",
    "sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    286690\n",
       "True       3894\n",
       "Name: user_id, dtype: int64"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counter = df['user_id'].value_counts()\n",
    "(counter > 1).value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3894 (1.34\\%) user_ids have been exposed to the old AND new page. It should be okay to remove them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 371,
   "metadata": {},
   "outputs": [],
   "source": [
    "#2. Remove users with multiple buckets\n",
    "valid_users = pd.DataFrame(counter[counter == 1].index, columns=['user_id'])\n",
    "df = df.merge(valid_users, on=['user_id'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 373,
   "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>user_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>group</th>\n",
       "      <th>landing_page</th>\n",
       "      <th>converted</th>\n",
       "      <th>week</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>129391</th>\n",
       "      <td>740710</td>\n",
       "      <td>2017-01-19 22:26:04.950909</td>\n",
       "      <td>control</td>\n",
       "      <td>old_page</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        user_id                   timestamp    group landing_page  converted  \\\n",
       "129391   740710  2017-01-19 22:26:04.950909  control     old_page          0   \n",
       "\n",
       "        week  \n",
       "129391     3  "
      ]
     },
     "execution_count": 373,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add week column to see the data as you would during experiment\n",
    "df['week'] = df['timestamp'].apply(lambda x: datetime.datetime.strptime(x, '%Y-%m-%d %H:%M:%S.%f').isocalendar()[1])\n",
    "df.sample()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 374,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    91380\n",
       "3    91056\n",
       "1    83745\n",
       "4    20509\n",
       "Name: week, dtype: int64"
      ]
     },
     "execution_count": 374,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['week'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Experiment: Frequentist Approach"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 381,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Treatment Conversion Rate: 11.873%\n",
      "Control Conversion Rate: 12.017%\n",
      "Lift = -0.144%\n"
     ]
    }
   ],
   "source": [
    "# Get Stats\n",
    "NUM_WEEKS = 4 # Vary number to get experiment data at weekly points in time\n",
    "experiment_data = df[df['week'] <= NUM_WEEKS]\n",
    "control = experiment_data[experiment_data['group']=='control']\n",
    "treatment = experiment_data[experiment_data['group']=='treatment']\n",
    "\n",
    "control_conversion_perc = round(control['converted'].sum() * 100/ control['converted'].count(), 3)\n",
    "treatment_conversion_perc = round(treatment['converted'].sum() * 100/ treatment['converted'].count(), 3)\n",
    "lift = round(treatment_conversion_perc - control_conversion_perc, 3)\n",
    "\n",
    "print(f\"Treatment Conversion Rate: {treatment_conversion_perc}%\")\n",
    "print(f\"Control Conversion Rate: {control_conversion_perc}%\")\n",
    "print(f\"Lift = {lift}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Chi-Squared Test\n",
    "\n",
    "$$\n",
    "H_0: \\text{Control & Treatment are independent} \\\\\n",
    "H_A: \\text{Control & Treatment are not independent} \n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create Contingency Table for Chi Squared Test\n",
    "control_converted = control['converted'].sum()\n",
    "treatment_converted = treatment['converted'].sum()\n",
    "control_non_converted = control['converted'].count() - control_converted\n",
    "treatment_non_converted = treatment['converted'].count() - treatment_converted\n",
    "contingency_table = np.array([[control_converted, control_non_converted], \n",
    "                              [treatment_converted, treatment_non_converted]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 15944, 117010],\n",
       "       [ 15799, 117428]])"
      ]
     },
     "execution_count": 228,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "contingency_table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {},
   "outputs": [],
   "source": [
    "chi, p_value, _, _ = chi2_contingency(contingency_table, correction=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1.1276469852100297, 0.28827779822703054)"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chi, p_value"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since the p_value > 0.05, we cannot reject null hypothesis. Hence, we cannot conclude if there exists a relationship between the control and treatment groups."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "28.83% probability that a more extreme chi quare than 1.128 would have occured by chance\n"
     ]
    }
   ],
   "source": [
    "print(f\"{round(p_value * 100, 2)}% probability that a more extreme chi square than {round(chi, 3)} would have occured by chance\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "But this is tough to interpret. We would to say something about the actual maginitude of lift. Something like this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(We CANNOT say this) We are 28.83% confident that our lift = -0.133%\n"
     ]
    }
   ],
   "source": [
    "print(f\"(We CANNOT say this) We are {round(p_value * 100, 2)}% confident that our lift = {lift}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 247,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image(filename='../mics/b_v_f.png') "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Experiment: Bayesian Approach"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We want to input the _prior distribution_ and have the experiment update the parameters to create _posterier distributions_. Since these prior & posterior distributions will be used to sample Conversion Rate, we model them after beta distribtion.\n",
    "\n",
    "Let's create the prior beta distribtion from the first weeks of conversion data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [],
   "source": [
    "prior = df[(df['week'] == 1) & (df['group']=='control')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {},
   "outputs": [],
   "source": [
    "prior_means = []\n",
    "for i in range(10000):\n",
    "    prior_means.append(prior.sample(1000)['converted'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.106, 0.105, 0.137, 0.116, 0.123, 0.12, 0.098, 0.129, 0.122, 0.124]"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prior_means[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model Beta Distribtion from sample means\n",
    "prior_alpha, prior_beta, _, _ = beta.fit(prior_means, floc=0, fscale=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 432,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Treatment Conversion Rate: 11.909%\n",
      "Control Conversion Rate: 12.058%\n",
      "Lift = -0.012%\n"
     ]
    }
   ],
   "source": [
    "# Get Stats\n",
    "NUM_WEEKS = 4 # Vary number to get experiment data at weekly points in time\n",
    "experiment_data = df[(df['week'] > 1) & (df['week'] <= NUM_WEEKS)]\n",
    "control = experiment_data[experiment_data['group']=='control']\n",
    "treatment = experiment_data[experiment_data['group']=='treatment']\n",
    "\n",
    "control_conversion_perc = round(control['converted'].sum() * 100/ control['converted'].count(), 3)\n",
    "treatment_conversion_perc = round(treatment['converted'].sum() * 100/ treatment['converted'].count(), 3)\n",
    "lift = round((treatment_conversion_perc - control_conversion_perc) / control_conversion_perc , 3)\n",
    "\n",
    "print(f\"Treatment Conversion Rate: {treatment_conversion_perc}%\")\n",
    "print(f\"Control Conversion Rate: {control_conversion_perc}%\")\n",
    "print(f\"Lift = {lift}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 433,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Probability that treatment > control: 15.8%\n"
     ]
    }
   ],
   "source": [
    "control_converted = control['converted'].sum()\n",
    "treatment_converted = treatment['converted'].sum()\n",
    "control_non_converted = control['converted'].count() - control_converted\n",
    "treatment_non_converted = treatment['converted'].count() - treatment_converted\n",
    "\n",
    "# Update Prior parameters with experiment conversion rates\n",
    "posterior_control = beta(prior_alpha + control_converted, prior_beta + control_non_converted)\n",
    "posterior_treatment = beta(prior_alpha + treatment_converted, prior_beta + treatment_non_converted)\n",
    "\n",
    "# Sample from Posteriors\n",
    "control_samples = posterior_control.rvs(1000)\n",
    "treatment_samples = posterior_treatment.rvs(1000)\n",
    "probability = np.mean(treatment_samples > control_samples)\n",
    "print(f\"Probability that treatment > control: {probability * 100}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 434,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Control Posterior: Mean: 0.12056303255497915, Variance: 1.0333316272670703e-06\n",
      "Treatment Posterior: Mean: 0.1190944894489385, Variance: 1.0242362170011709e-06\n"
     ]
    }
   ],
   "source": [
    "(control_mu), (control_var) = posterior_control.stats()\n",
    "(treatment_mu), (treatment_var) = posterior_treatment.stats()\n",
    "print(f\"Control Posterior: Mean: {control_mu}, Variance: {control_var}\") \n",
    "print(f\"Treatment Posterior: Mean: {treatment_mu}, Variance: {treatment_var}\") "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can even make statements like the following which are actionable:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 388,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Probability that we are seeing a 2% lift: 0.4%\n"
     ]
    }
   ],
   "source": [
    "lift_percentage = (treatment_samples - control_samples) / control_samples\n",
    "print(f\"Probability that we are seeing a 2% lift: {np.mean((100 * lift_percentage) > 2) * 100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Advantages of Bayesian over Frequentist:\n",
    "- Results are more interpretable than the ones we got from the frequentist approach\n",
    "- We can interpret results at any point during the experiment. Don't need to wait for an arbitrary \"statsig\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### References\n",
    "\n",
    "[1] Good blog to understand Beta Distribution: http://varianceexplained.org/statistics/beta_distribution_and_baseball/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "0:00 Introduction\n",
    "0:56 Define the Experiment\n",
    "4:25 Data Collection\n",
    "6:20 Data Processing\n",
    "8:43 Experiment: The Frequentist Approach\n",
    "12:50 Experiment: The Bayesian Approach\n",
    "14:40 Bayesian: Generating Priors\n",
    "20:06 Bayesian: Generating Posteriors\n",
    "22:00 Interpreting results\n",
    "26:00 Bayesian Vs Frequentist"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}