sequential_ab_testing.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "sequential_ab_testing.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N7ER5fea_pip",
        "colab_type": "text"
      },
      "source": [
        "## Sequential A/B Testing\n",
        "- 此 Notebook 將介紹 Sequential A/B Testing\n",
        "- 主要參考資料是 Evan Miller 的文章（[連結](https://www.evanmiller.org/sequential-ab-testing.html)）\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G2HY58UZ0cLm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#@title Import Packages\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from scipy import stats\n",
        "import numpy as np\n",
        "import math\n",
        "from plotnine import *"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UV0AcGNEAJzW",
        "colab_type": "text"
      },
      "source": [
        "## 將會與標準假設檢定的固定樣本方法比較所需樣本數"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OlONmmerzJWK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#@title Calculate Sample Size for 'Fixed-sample-size-method'\n",
        "# ref: https://stats.stackexchange.com/questions/178568/calculate-sample-size-based-on-conversion-rate-minimum-detectable-effect-stati\n",
        "def get_sample_size(base_rate, absolute_diff, alpha=0.05, beta=0.2, if_one_tail=True):\n",
        "    p0 = base_rate\n",
        "    p1 = base_rate + absolute_diff\n",
        "    result = np.power(stats.norm.ppf(1 - alpha / (1.0 if if_one_tail else 2.)) + stats.norm.ppf(1 - beta), 2)\n",
        "    result *= (p0 * (1 - p0)) + (p1 * (1 - p1))\n",
        "    result = result / np.power(absolute_diff, 2)\n",
        "    return math.ceil(result)"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IkULxGvAATKk",
        "colab_type": "text"
      },
      "source": [
        "## Sequential A/B Testing 所需樣本數可以查表取得"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d_RRdD120YS4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "97c082d3-a653-4667-cd72-a7ecc5313c35"
      },
      "source": [
        "#@title Sample Size Reference Table for Sequential A/B Testing\n",
        "# ref: https://www.evanmiller.org/ab-testing/sequential.html\n",
        "seq_size_table = pd.DataFrame({\n",
        "    'sig': [0.05]*5,\n",
        "    'power': [0.8]*5,\n",
        "    'min_relative_effect': [0.1, 0.2, 0.3, 0.4, 0.5],\n",
        "    'n_total_success': [2922, 808, 391, 243, 170],\n",
        "    'n_success_ahead': [106, 56, 39, 31, 26],\n",
        "})\n",
        "seq_size_table"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>sig</th>\n",
              "      <th>power</th>\n",
              "      <th>min_relative_effect</th>\n",
              "      <th>n_total_success</th>\n",
              "      <th>n_success_ahead</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.1</td>\n",
              "      <td>2922</td>\n",
              "      <td>106</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.2</td>\n",
              "      <td>808</td>\n",
              "      <td>56</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.3</td>\n",
              "      <td>391</td>\n",
              "      <td>39</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.4</td>\n",
              "      <td>243</td>\n",
              "      <td>31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.5</td>\n",
              "      <td>170</td>\n",
              "      <td>26</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    sig  power  min_relative_effect  n_total_success  n_success_ahead\n",
              "0  0.05    0.8                  0.1             2922              106\n",
              "1  0.05    0.8                  0.2              808               56\n",
              "2  0.05    0.8                  0.3              391               39\n",
              "3  0.05    0.8                  0.4              243               31\n",
              "4  0.05    0.8                  0.5              170               26"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qWdMdAzDMBp8",
        "colab_type": "text"
      },
      "source": [
        "## Simple Sequential A/B Testing 規則（單尾檢定）\n",
        "1. 決定好你對實驗假設的顯著力、檢定力、以及期望的轉換率差異量之後，使用 [計算機](https://www.evanmiller.org/ab-testing/sequential.html) 或者查表，選出**轉換成功**樣本數大小 N\n",
        "2. 開始 A/B Testing\n",
        "3. 持續觀測實驗組（B 組）的轉換成功數量 $T$、以及控制組（A 組）轉換成功數量 $C$\n",
        "4. 如果 $T - C$ 超過 $2\\sqrt{N}$，宣告實驗組較好，結束實驗\n",
        "5. 如果第 4 點不成立，且 $T + C$ 累積到超過 $N$，無法宣告任一方較好，結束實驗"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_MavyP_e3D0r",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#@title Sequential A/B Testing\n",
        "def seq_ab_testing(base_rate=0.01, \n",
        "          true_relative_lift_effect=0, \n",
        "          n_total_success=808, \n",
        "          n_success_ahead=56):\n",
        "    step_size = 50\n",
        "    sample_size_per_group = 0\n",
        "    n_success_a, n_success_b = 0, 0\n",
        "    while True:\n",
        "        sample_size_per_group += step_size\n",
        "        n_success_a += np.random.binomial(n=step_size, p=base_rate)\n",
        "        n_success_b += np.random.binomial(n=step_size, p=base_rate*(1+true_relative_lift_effect))\n",
        "        if (n_success_b - n_success_a) >= n_success_ahead:\n",
        "            return {'sample_size_per_group': sample_size_per_group, 'n_success_a': n_success_a, 'n_success_b': n_success_b, 'if_b_win': True}\n",
        "        elif (n_success_a + n_success_b) >= n_total_success:\n",
        "            return {'sample_size_per_group': sample_size_per_group, 'n_success_a': n_success_a, 'n_success_b': n_success_b, 'if_b_win': False}\n",
        "        if sample_size_per_group >= 5000000: return None"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CV-bR0ca_et5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#@title Aggregate Multiple Testing Results\n",
        "def aggregate_test_results(df, fixed_sample_method_size, verbose=True):\n",
        "    if verbose: \n",
        "        print(f\"total trials: {len(df)} times\")\n",
        "        print(f\"treatment group wins ratio: {df['if_b_win'].sum()/len(df)*100:.1f}%\")\n",
        "        print(f\"sequential method sample size larger than fixed sample method, ratio: {(df['sample_size_per_group']>fixed_sample_method_size).sum()/len(df)*100:.1f}%\")\n",
        "        print(f\"average sequential method sample size relative to fixed sample method: {df['sample_size_per_group'].mean()/fixed_sample_method_size*100:.1f}%\")\n",
        "    return {'total_trials': len(df),\n",
        "         'treatment_group_win_ratio': df['if_b_win'].sum()/len(df),\n",
        "         'seq_method_sample_size_larger_ratio': (df['sample_size_per_group']>fixed_sample_method_size).sum()/len(df),\n",
        "         'avg_seq_method_sample_size_over_fixed_sample_method': df['sample_size_per_group'].mean()/fixed_sample_method_size,\n",
        "         'fixed_sample_method_size': fixed_sample_method_size,\n",
        "         'avg_n_total_success': (df['n_success_a'].sum()+df['n_success_b'].sum())/len(df)}"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ISy0s9XHMLeU",
        "colab_type": "text"
      },
      "source": [
        "## Sequential A/B Testing 有效性驗證\n",
        "\n",
        "### 刻意設計不同情境，觀察檢定方法是否判定 A/B 兩組有差異\n",
        "- A/A Test: A/B 兩組實際上並無差異\n",
        "- A/B Test: A/B 兩組有差異，差異量恰好為期望差異量\n",
        "- Blockbuster Test: A/B 兩組有差異，而且差異量是期望差異量的兩倍\n",
        "\n",
        "### 範例驗證結果\n",
        "- Sequential A/B Testing 的 Type-1 Error Rate 確實逼近 5%、符合假設的顯著力\n",
        "- Sequential A/B Testing 的 Type-2 Error Rate 確實逼近 80%、符合假設的檢定力\n",
        "- 說明此方法確實有效"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "efDPZrhc-pV4",
        "colab_type": "code",
        "cellView": "both",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "outputId": "58febcdd-a268-4e22-b6c6-2209f1353385"
      },
      "source": [
        "#@title Example\n",
        "np.random.seed(42)\n",
        "trial_times = 1000\n",
        "base_rate = 0.01\n",
        "expected_relative_lift_effect = 0.2\n",
        "expected_absolute_lift_effect = base_rate*expected_relative_lift_effect\n",
        "n_total_success = 808\n",
        "n_success_ahead = 56\n",
        "fixed_sample_method_size = get_sample_size(base_rate, absolute_diff=expected_absolute_lift_effect)\n",
        "aa_test_result = pd.DataFrame(seq_ab_testing(base_rate=base_rate, \n",
        "                        true_relative_lift_effect=0,\n",
        "                        n_total_success=n_total_success,\n",
        "                        n_success_ahead=n_success_ahead) for _ in range(trial_times))\n",
        "ab_test_result = pd.DataFrame(seq_ab_testing(base_rate=base_rate, \n",
        "                        true_relative_lift_effect=expected_relative_lift_effect,\n",
        "                        n_total_success=n_total_success,\n",
        "                        n_success_ahead=n_success_ahead) for _ in range(trial_times))\n",
        "blockbuster_test_result = pd.DataFrame(seq_ab_testing(base_rate=base_rate, \n",
        "                        true_relative_lift_effect=expected_relative_lift_effect*2,\n",
        "                        n_total_success=n_total_success,\n",
        "                        n_success_ahead=n_success_ahead) for _ in range(trial_times))\n",
        "print(f\"base_rate = {base_rate*100:.0f}%, expected_relative_lift_effect = {expected_relative_lift_effect*100:.0f}%\\n\")\n",
        "print(\"aa test\")\n",
        "aggregate_test_results(aa_test_result, fixed_sample_method_size)\n",
        "print()\n",
        "print(\"ab test\")\n",
        "aggregate_test_results(ab_test_result, fixed_sample_method_size)\n",
        "print()\n",
        "print(\"blockbuster test\")\n",
        "aggregate_test_results(blockbuster_test_result, fixed_sample_method_size)\n",
        "print()"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "base_rate = 1%, expected_relative_lift_effect = 20%\n",
            "\n",
            "aa test\n",
            "total trials: 1000 times\n",
            "treatment group wins ratio: 5.4%\n",
            "sequential method sample size larger than fixed sample method, ratio: 96.8%\n",
            "average sequential method sample size relative to fixed sample method: 118.6%\n",
            "\n",
            "ab test\n",
            "total trials: 1000 times\n",
            "treatment group wins ratio: 80.7%\n",
            "sequential method sample size larger than fixed sample method, ratio: 25.9%\n",
            "average sequential method sample size relative to fixed sample method: 76.3%\n",
            "\n",
            "blockbuster test\n",
            "total trials: 1000 times\n",
            "treatment group wins ratio: 99.7%\n",
            "sequential method sample size larger than fixed sample method, ratio: 0.1%\n",
            "average sequential method sample size relative to fixed sample method: 41.8%\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x-nBwfdYNl-h",
        "colab_type": "text"
      },
      "source": [
        "## 觀察基礎轉換率差異，對於 Sequential A/B Testing 優勢的影響"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5Xbb7zfsHWBL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#@title Main Experiment: Try Each Setting for 1,000 times\n",
        "np.random.seed(42)\n",
        "df = []\n",
        "base_rate_list = [0.01, 0.02, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5]\n",
        "trial_times = 1000\n",
        "\n",
        "for base_rate in base_rate_list:\n",
        "    for i in range(len(seq_size_table)):\n",
        "        sig, power, min_relative_effect, n_total_success, n_success_ahead = seq_size_table.iloc[i, :][['sig', 'power', 'min_relative_effect', 'n_total_success', 'n_success_ahead']]\n",
        "        expected_relative_lift_effect = min_relative_effect \n",
        "        expected_absolute_lift_effect = base_rate * expected_relative_lift_effect\n",
        "        fixed_sample_method_size = get_sample_size(base_rate, absolute_diff=expected_absolute_lift_effect, alpha=sig, beta=1-power)\n",
        "        aa_test_result = pd.DataFrame(seq_ab_testing(base_rate=base_rate, \n",
        "                                true_relative_lift_effect=0,\n",
        "                                n_total_success=n_total_success,\n",
        "                                n_success_ahead=n_success_ahead) for _ in range(trial_times))\n",
        "        ab_test_result = pd.DataFrame(seq_ab_testing(base_rate=base_rate, \n",
        "                                true_relative_lift_effect=expected_relative_lift_effect,\n",
        "                                n_total_success=n_total_success,\n",
        "                                n_success_ahead=n_success_ahead) for _ in range(trial_times))\n",
        "        blockbuster_test_result = pd.DataFrame(seq_ab_testing(base_rate=base_rate, \n",
        "                                true_relative_lift_effect=expected_relative_lift_effect*2,\n",
        "                                n_total_success=n_total_success,\n",
        "                                n_success_ahead=n_success_ahead) for _ in range(trial_times))\n",
        "        result = aggregate_test_results(aa_test_result, fixed_sample_method_size, verbose=False)\n",
        "        df.append({'base_rate': base_rate, 'test_type': 'aa test', **dict(seq_size_table.iloc[i, :], **result)})\n",
        "        result = aggregate_test_results(ab_test_result, fixed_sample_method_size, verbose=False)\n",
        "        df.append({'base_rate': base_rate, 'test_type': 'ab test', **dict(seq_size_table.iloc[i, :], **result)})\n",
        "        result = aggregate_test_results(blockbuster_test_result, fixed_sample_method_size, verbose=False)\n",
        "        df.append({'base_rate': base_rate, 'test_type': 'blockbuster test', **dict(seq_size_table.iloc[i, :], **result)})\n",
        "df = pd.DataFrame(df)"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KVwrptii92Og",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df = df.assign(fixed_method_expected_n_success = lambda df: (df['fixed_sample_method_size'] * df['base_rate'] * 2).round(0),\n",
        "         savings = lambda df: ((df['fixed_method_expected_n_success'] - df['avg_n_total_success']) / df['fixed_method_expected_n_success']).round(2))"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KYQzlPrkxTCv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 162
        },
        "outputId": "ae732526-42d7-4dd0-bee5-fc412809fb28"
      },
      "source": [
        "#@title Reproduce a Part of Evan's Sample Size Table\n",
        "(\n",
        "     df[['base_rate', 'test_type', 'sig', 'power', 'min_relative_effect', 'n_total_success', 'n_success_ahead', 'avg_n_total_success', 'fixed_method_expected_n_success', 'savings']]\n",
        "     .pipe(lambda x: x.loc[x['min_relative_effect'].isin({0.5, 0.2, 0.1}), :])\n",
        "     .pipe(lambda x: x.loc[x['test_type'] == 'ab test', :])\n",
        "     .pipe(lambda x: x.loc[x['base_rate'] == 0.01, :])\n",
        "     .sort_values(['test_type', 'sig', 'power', 'min_relative_effect'], ascending=False)\n",
        ")"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>base_rate</th>\n",
              "      <th>test_type</th>\n",
              "      <th>sig</th>\n",
              "      <th>power</th>\n",
              "      <th>min_relative_effect</th>\n",
              "      <th>n_total_success</th>\n",
              "      <th>n_success_ahead</th>\n",
              "      <th>avg_n_total_success</th>\n",
              "      <th>fixed_method_expected_n_success</th>\n",
              "      <th>savings</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>0.01</td>\n",
              "      <td>ab test</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.5</td>\n",
              "      <td>170.0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>121.828</td>\n",
              "      <td>122.0</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.01</td>\n",
              "      <td>ab test</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.2</td>\n",
              "      <td>808.0</td>\n",
              "      <td>56.0</td>\n",
              "      <td>570.441</td>\n",
              "      <td>673.0</td>\n",
              "      <td>0.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.01</td>\n",
              "      <td>ab test</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.1</td>\n",
              "      <td>2922.0</td>\n",
              "      <td>106.0</td>\n",
              "      <td>2070.550</td>\n",
              "      <td>2569.0</td>\n",
              "      <td>0.19</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    base_rate test_type  ...  fixed_method_expected_n_success  savings\n",
              "13       0.01   ab test  ...                            122.0     0.00\n",
              "4        0.01   ab test  ...                            673.0     0.15\n",
              "1        0.01   ab test  ...                           2569.0     0.19\n",
              "\n",
              "[3 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wakaGMUFN0n5",
        "colab_type": "text"
      },
      "source": [
        "## 在低基礎轉換率、且兩組實際有差異的情境，Sequential A/B Testing 有加速實驗完成的優勢"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_yXf770VxCPk",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 485
        },
        "outputId": "03e90f8e-7f05-4280-a795-e765d1096448"
      },
      "source": [
        "#@title Visualize the Advantage of Sequential A/B Testing\n",
        "plot_data = (df.assign(base_rate = lambda x: x['base_rate'].astype(str)))\n",
        "(\n",
        "    ggplot(data=plot_data, mapping=aes(x='min_relative_effect', y='savings', colour='base_rate'))\n",
        "    + geom_hline(yintercept=0)\n",
        "    + geom_line()\n",
        "    + geom_point()\n",
        "    + facet_wrap('~ test_type') \n",
        "    + coord_cartesian(ylim=(-0.6, 0.6))\n",
        ")"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<ggplot: (-9223363308132445436)>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_DxFrMsNOENI",
        "colab_type": "text"
      },
      "source": [
        "## 以「1.5 * 基礎轉換率 + 相對差異 < 36%」的經驗法則，確認是否適合用 Sequential A/B Testing\n",
        "#### 經驗法則視覺化：只有在灰色虛線左邊的範圍，Sequential A/B Testing 才有加速實驗完成的優勢"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u396BhYWVg8b",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 485
        },
        "outputId": "5c09b540-a38c-4704-920e-2cb9e33934ef"
      },
      "source": [
        "#@title Largest MDE with positive savings\n",
        "largest_mde = pd.DataFrame({'base_rate': df['base_rate'].unique().astype(str), \n",
        "               'largest_mde': 0.36 - 1.5 * df['base_rate'].unique()})\n",
        "largest_mde = largest_mde.loc[largest_mde['largest_mde'] > 0, :]\n",
        "plot_data = (df.pipe(lambda x: x.loc[(x['test_type'] == 'ab test') & x['base_rate'].isin(largest_mde['base_rate']), :])\n",
        "         .assign(base_rate = lambda x: x['base_rate'].astype(str)))\n",
        "(\n",
        "    ggplot(data=plot_data, mapping=aes(x='min_relative_effect', y='savings', colour='base_rate'))\n",
        "    + geom_hline(yintercept=0)\n",
        "    + geom_vline(data=largest_mde,\n",
        "           mapping=aes(xintercept='largest_mde'),\n",
        "           colour='grey',\n",
        "           linetype='dashed')\n",
        "    + geom_text(data=largest_mde.assign(largest_mde = (largest_mde['largest_mde']+0.08).round(2),\n",
        "                       largest_mde_label = (largest_mde['largest_mde'] * 100).astype(int).astype(str) + '%'),\n",
        "            mapping=aes(x='largest_mde',\n",
        "                  y=0.5,\n",
        "                  label='largest_mde_label'),\n",
        "           colour='grey')\n",
        "    + geom_line()\n",
        "    + geom_point()\n",
        "    + facet_wrap('~ base_rate') \n",
        "    + coord_cartesian(ylim=(-0.6, 0.6))\n",
        ")"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 5 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<ggplot: (-9223363308135527964)>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    }
  ]
}