notebooks/puffer-analysis.ipynb

{
    "cells": [
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# Puffer Analysis\n",
                "\n",
                "You use this notebook to explore the Puffer data evaluation results and/or recreate the following figures:\n",
                "\n",
                "**Main paper:**\n",
                "- Figure 1\n",
                "- Figure 6\n",
                "\n",
                "**Appendix:**\n",
                "- Figure 10\n",
                "- Figure 17\n",
                "- Figure 18\n",
                "- Figure 19\n",
                "- Figure 20\n",
                "\n",
                "For a simpler overview of the evaluation data, see the [demo notebook](demo.ipynb).\n",
                "\n",
                "## Prerequisites\n",
                "\n",
                "You must download and evaluate the puffer data, which you can do by running the following CLI commands:\n",
                "\n",
                "```bash\n",
                "./run.py puffer-download\n",
                "\n",
                "# Optional, but strongly recommended if you plan to run all experiments.\n",
                "./run.py puffer-preprocess\n",
                "\n",
                "./run.py puffer-analysis\n",
                "```\n",
                "\n",
                "_Warning_: The full dataset consists of several hundred GB of data and the download and evaluation may take several hours. Each day takes about 5-10 minutes each do download and analyze, depending on your setup.\n",
                "Preprocessing the data takes about 3 minutes per day (can be parallelized) and prepares the Puffer \"TTP\" model input/output from the dataset. If you do not preprocess the data, this step is done on the fly when running the analysis. If you also plan to run the replay experiments, we recommend preprocessing the data to save a lot of time. However, the full dataset, including preprocessed data, will take about 1TB of disk space.\n",
                "\n",
                "\n",
                "\n",
                "The experiments are structured by day and you may only download and evaluate a subset of data to experiment with the notebook.\n",
                "You can do this select experiments with glob patterns.\n",
                "For example, the most recent Memento deployment started in September 2022. To load only the data since them, you can run:\n",
                "\n",
                "```bash\n",
                "./run.py puffer-download \"2022-09-*\" \"2022-1*\" \"2023-*\"\n",
                "./run.py puffer-analysis \"2022-09-*\" \"2022-1*\" \"2023-*\"\n",
                "```\n",
                "\n",
                "See `REAMDE.md` for additional CLI options such as verbosity, running experiments in parallel, or scheduling jobs with Slurm."
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Imports and Settings"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 1,
            "metadata": {},
            "outputs": [],
            "source": [
                "import sys\n",
                "sys.path.append('..')  # To allow imports from parent directory\n",
                "%load_ext autoreload\n",
                "%autoreload 2"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 2,
            "metadata": {},
            "outputs": [],
            "source": [
                "from operator import itemgetter\n",
                "\n",
                "import experiment_helpers as eh\n",
                "import matplotlib.patches as patches\n",
                "import numpy as np\n",
                "import pandas as pd\n",
                "import seaborn as sns\n",
                "from experiment_helpers.data import Path\n",
                "from matplotlib import dates as mdates\n",
                "from matplotlib import pyplot as plt\n",
                "from matplotlib.legend_handler import HandlerTuple\n",
                "from scipy import stats\n",
                "\n",
                "# Load the config, including the local config.\n",
                "from config import Config as LocalConfig  # Load from file: ../config.py\n",
                "from experiments.puffer.config import PufferExperimentConfig\n",
                "\n",
                "config = PufferExperimentConfig.with_updates(LocalConfig)"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 3,
            "metadata": {},
            "outputs": [],
            "source": [
                "# Plot import and configuration.\n",
                "%matplotlib inline\n",
                "\n",
                "# Set theme and parameters\n",
                "eh.plot_utils.setup(latex=True)\n",
                "\n",
                "plotdir = Path(\"../plots/puffer-analysis\")\n",
                "plotdir.mkdir(exist_ok=True, parents=True)"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 4,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": [
                        "1414 evaluated days found.\n"
                    ]
                }
            ],
            "source": [
                "# Load the config, including the local config.\n",
                "from config import Config as LocalConfig  # File: ../config.py\n",
                "from experiments.puffer.config import PufferExperimentConfig\n",
                "config = PufferExperimentConfig.with_updates(LocalConfig)\n",
                "\n",
                "evaldir = config.output_directory / \"puffer-analysis\"\n",
                "print(f\"{len(list(evaldir.iterdir()))} evaluated days found.\")"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Important dates\n",
                "\n",
                "We need these points in time to select appropriate time ranges."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 5,
            "metadata": {},
            "outputs": [],
            "source": [
                "# Important dates for the analysis.\n",
                "# First day of the current deployed version of Memento\n",
                "deployment_start = pd.to_datetime(\"2022-09-19\")\n",
                "# First day of the previous deployment with a different version of Memento.\n",
                "old_deployment_start = pd.to_datetime(\"2022-05-01\")\n",
                "# The last day for which a daily-retrained Fugu model is available.\n",
                "last_fugu_day = pd.to_datetime(\"2022-10-05\")\n",
                "# The last day considered for the analysis.\n",
                "cutoff_day = pd.to_datetime(\"2023-07-17\")"
            ]
        },
        {
            "attachments": {},
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Constants\n",
                "\n",
                "Here we define importants days to filter the data as well as pretty display names and colors for the figures."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 6,
            "metadata": {},
            "outputs": [],
            "source": [
                "# Pretty tex names.\n",
                "#bba_name = r\"\\textsf{BBA}\"\n",
                "fugu_name = r\"\\textsf{Fugu}\"\n",
                "fugufeb_name = r\"\\textsf{Fugu} (static)\"\n",
                "memento_name = r\"\\textsf{Memento}\"\n",
                "memento_deterministic = r\"\\textsf{Memento} (T\\texttt{=}0)\"\n",
                "\n",
                "# Map dataframe names to pretty names.\n",
                "name_mapping = {\n",
                "    #'bba': bba_name,\n",
                "    'fugu-feb': fugufeb_name,\n",
                "    'fugu': fugu_name,\n",
                "    'memento': memento_name,\n",
                "    'memento-deterministic': memento_deterministic,\n",
                "}\n",
                "\n",
                "# A consistent order for ABRs across plots.\n",
                "abr_order = [  # Easier sorting\n",
                "    memento_name, memento_deterministic, fugu_name, fugufeb_name, #bba_name\n",
                "]\n",
                "abr_order_short = [  # Exlude Fugu (keep FuguFeb)\n",
                "    memento_name, memento_deterministic, fugufeb_name,\n",
                "]\n",
                "\n",
                "# A consistent color palette for ABRs across plots.\n",
                "base_palette = sns.color_palette(\"deep\")\n",
                "abr_palette = {\n",
                "    memento_name: base_palette[2],           # green\n",
                "    memento_deterministic: base_palette[3],  # red\n",
                "    fugu_name: base_palette[1],              # orange\n",
                "    fugufeb_name: base_palette[0],           # blue\n",
                "}"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# Helper functions\n",
                "\n",
                "- Weighted mean (so we can average over days with different amounts of data)\n",
                "- absolute difference and maximum percent difference for metric comparisons."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 7,
            "metadata": {},
            "outputs": [],
            "source": [
                "def weighted_mean(frame, columns, weightcolumn=\"weight\"):\n",
                "    \"\"\"Compute the weighted mean for a dataframe.\"\"\"\n",
                "    values = frame[columns].values\n",
                "    weights = frame[weightcolumn].values if weightcolumn else None\n",
                "    return weighted_mean_raw(values, weights)\n",
                "\n",
                "def weighted_mean_raw(values, weights=None):\n",
                "    if weights is None:\n",
                "        return np.mean(values, axis=0)\n",
                "    else:\n",
                "        weights = np.asarray(weights)\n",
                "        # Add dimensions in case values is 2d.\n",
                "        weights = \\\n",
                "            weights[(..., ) + (np.newaxis,) * (values.ndim - weights.ndim)]\n",
                "        return (values * weights).sum(axis=0) / weights.sum()\n",
                "\n",
                "def diff(x, y):\n",
                "    \"\"\"Compute the absolute difference of two values.\"\"\"\n",
                "    return (x - y)\n",
                "\n",
                "def maxpd(x, y):\n",
                "    \"\"\"Compute the maximum percentage difference of two values.\"\"\"\n",
                "    return 100 * (x - y) / np.maximum(np.abs(x), np.abs(y))"
            ]
        },
        {
            "attachments": {},
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Load raw data\n",
                "\n",
                "First, load the raw evaluation data.\n",
                "\n",
                "Over time, we deployed several version of Memento. For the following plots, we only keep the two most recent deployments. We keep past data from past ABRs, but discard outdated Memento versions to avoid confusion."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 8,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": [
                        "Days with results per ABR:\n"
                    ]
                },
                {
                    "data": {
                        "text/plain": [
                            "abr\n",
                            "\\textsf{Memento}                    415\n",
                            "\\textsf{Memento} (T\\texttt{=}0)     415\n",
                            "\\textsf{Fugu}                       871\n",
                            "\\textsf{Fugu} (static)             1156\n",
                            "Name: day, dtype: int64"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                },
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": [
                        "Days with results per ABR (only current Memento deployment):\n"
                    ]
                },
                {
                    "data": {
                        "text/plain": [
                            "abr\n",
                            "\\textsf{Memento}                   285\n",
                            "\\textsf{Memento} (T\\texttt{=}0)    285\n",
                            "\\textsf{Fugu}                       10\n",
                            "\\textsf{Fugu} (static)             285\n",
                            "Name: day, dtype: int64"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                },
                {
                    "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>abr</th>\n",
                            "      <th>metric</th>\n",
                            "      <th>value</th>\n",
                            "      <th>variable</th>\n",
                            "      <th>selection</th>\n",
                            "      <th>day</th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>0</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>mean</td>\n",
                            "      <td>17.377717</td>\n",
                            "      <td>ssim</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>1</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>std</td>\n",
                            "      <td>0.044979</td>\n",
                            "      <td>ssim</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>2</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>rms</td>\n",
                            "      <td>17.390574</td>\n",
                            "      <td>ssim</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>0.0</td>\n",
                            "      <td>4.360611</td>\n",
                            "      <td>ssim</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>4</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>0.001</td>\n",
                            "      <td>8.490398</td>\n",
                            "      <td>ssim</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "                      abr metric      value variable selection        day\n",
                            "0  \\textsf{Fugu} (static)   mean  17.377717     ssim       all 2023-02-09\n",
                            "1  \\textsf{Fugu} (static)    std   0.044979     ssim       all 2023-02-09\n",
                            "2  \\textsf{Fugu} (static)    rms  17.390574     ssim       all 2023-02-09\n",
                            "3  \\textsf{Fugu} (static)    0.0   4.360611     ssim       all 2023-02-09\n",
                            "4  \\textsf{Fugu} (static)  0.001   8.490398     ssim       all 2023-02-09"
                        ]
                    },
                    "execution_count": 8,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": [
                "_frames = []\n",
                "for daydir in evaldir.iterdir():\n",
                "    _resfile = daydir / \"results.csv.gz\"\n",
                "    if not _resfile.is_file():\n",
                "        continue\n",
                "    _frames.append(eh.data.read_csv(_resfile, parse_dates=['day']))\n",
                "\n",
                "raw_results = (\n",
                "    pd\n",
                "    .concat(_frames, ignore_index=True)\n",
                "    .assign(\n",
                "        # For each frame, apply name mapping and use a categorical\n",
                "        # datatype to ensure consistent ordering and to save space.\n",
                "        abr=lambda df: pd.Categorical(\n",
                "            df['abr'].replace(name_mapping), abr_order),\n",
                "    )\n",
                "    .dropna(subset=['abr'])  # Drop rows without relevant ABR.\n",
                "    # Remove Memento versions from old deployments, i.e. keep non-Memento\n",
                "    # for the whole range and Memento only since the current deployment.\n",
                "    .query(\"not ((abr in [@memento_name, @memento_deterministic]) and \"\n",
                "                \"(day < @old_deployment_start)) and (day < @cutoff_day)\")\n",
                ")\n",
                "\n",
                "\n",
                "## Remove an outlier for Fugu (static)\n",
                "# On a single day (2022-12-10), Fugu-Feb reports 80h of rebuffering.\n",
                "# After closer investigation, this is caused by a single stream reporting\n",
                "# almost all of this rebuffering time. We exclude this day from the analysis.\n",
                "raw_results = raw_results.query(\"day != @pd.to_datetime('2022-12-10')\")\n",
                "\n",
                "## Remve and outlier for Fugu (daily)\n",
                "# For similar reasons we exclude 2022-07-07. On this day the data publihsed\n",
                "# for Fugu (daily) shows 40% of stream-time spent stalled, more than 20x the\n",
                "# next highest value. We exclude this day to not skew the results.\n",
                "raw_results = raw_results.query(\"day != @pd.to_datetime('2022-07-07')\")\n",
                "\n",
                "# Quick check how much data we loaded.\n",
                "print(\"Days with results per ABR:\")\n",
                "display(\n",
                "    raw_results\n",
                "    .groupby('abr')\n",
                "    ['day'].nunique()\n",
                ")\n",
                "\n",
                "print(\"Days with results per ABR (only current Memento deployment):\")\n",
                "display(\n",
                "    raw_results\n",
                "    .query(\"day >= @deployment_start\")\n",
                "    .groupby('abr')\n",
                "    ['day'].nunique()\n",
                ")\n",
                "\n",
                "# Also, let's have a peek at the data structure. It's a long-form frame.\n",
                "raw_results.head()"
            ]
        },
        {
            "attachments": {},
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "The data is stored in [long (also called narrow) format][1]. The context of each value is given by:\n",
                "- `day`: The evaluated day.\n",
                "- `abr` The evaluated algorithm.\n",
                "- `selection`: We evaluate three subsets of data.\n",
                "  - `all`: All data.\n",
                "  - `stalled`: Only results of sessions experiencing stalls.\n",
                "  - `unstalled`: Only results of sessions not experiencing stalls.\n",
                "- `variable`: Which aspect of the data is evaluated. Some important variables are:\n",
                "  - `ssim`: Image quality of all video chunks.\n",
                "  - `stream`: Time info for the stream, e.g. playtime and rebuffering time.\n",
                "- `metric`: For each variable, we compute several aggregate statistics. E.g. for SSIM, we evaluate mean, std, and several percentiles.\n",
                "\n",
                "Next, we select the `(variable, metric)` combinations we are going to plot and give them a nice name for visualization in a single column called `varmetric`.\n",
                "\n",
                "[1]: https://en.wikipedia.org/wiki/Wide_and_narrow_data"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 9,
            "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>abr</th>\n",
                            "      <th>value</th>\n",
                            "      <th>selection</th>\n",
                            "      <th>day</th>\n",
                            "      <th>varmetric</th>\n",
                            "      <th>days_running</th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>0</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>17.377717</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Mean SSIM</td>\n",
                            "      <td>143</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>18</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>17.447852</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Mean SSIM</td>\n",
                            "      <td>143</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>36</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>16.802479</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Mean SSIM</td>\n",
                            "      <td>143</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>144</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>751833.729000</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Time played</td>\n",
                            "      <td>143</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>145</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>1926.820000</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Time stalled</td>\n",
                            "      <td>143</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "                                 abr          value selection        day  \\\n",
                            "0             \\textsf{Fugu} (static)      17.377717       all 2023-02-09   \n",
                            "18                  \\textsf{Memento}      17.447852       all 2023-02-09   \n",
                            "36   \\textsf{Memento} (T\\texttt{=}0)      16.802479       all 2023-02-09   \n",
                            "144           \\textsf{Fugu} (static)  751833.729000       all 2023-02-09   \n",
                            "145           \\textsf{Fugu} (static)    1926.820000       all 2023-02-09   \n",
                            "\n",
                            "        varmetric  days_running  \n",
                            "0       Mean SSIM           143  \n",
                            "18      Mean SSIM           143  \n",
                            "36      Mean SSIM           143  \n",
                            "144   Time played           143  \n",
                            "145  Time stalled           143  "
                        ]
                    },
                    "execution_count": 9,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": [
                "# Mapping from (variable, variable metric) to high level metric label.\n",
                "metrics = {\n",
                "    # Number of sessions and video chunks\n",
                "    (\"session\", \"count\"): \"\\# Sessions\",\n",
                "    (\"chunk\", \"count\"): \"\\# Chunks\",\n",
                "    # Stream time.\n",
                "    (\"stream\", \"streamtime\"): \"Time streamed\",  # Play + buffer time.\n",
                "    (\"stream\", \"playtime\"): \"Time played\",\n",
                "    (\"stream\", \"rebuftime\"): \"Time stalled\",\n",
                "    (\"stream\", \"stalled\"): \"Percent stalled\",  # Time stalled / Time streamed\n",
                "    # Image quality (SSIM).\n",
                "    (\"ssim\", \"mean\"): \"Mean SSIM\",\n",
                "    # Prediction scores.\n",
                "    (\"logscore_0\", \"mean\"): \"Score Mean\",\n",
                "    (\"logscore_0\", '0.5'): \"Score Median\",\n",
                "    (\"logscore_0\", '0.01'): \"Score Tail\",\n",
                "}\n",
                "# Consistent sorting of metrics.\n",
                "metric_order = list(metrics.values())\n",
                "\n",
                "# Metrics with linear scale. We will look at percentage differences below,\n",
                "# but these are only meaningful for these metrics.\n",
                "metrics_linear = [\n",
                "    '\\# Sessions', '\\# Chunks',\n",
                "    'Time streamed', 'Time played', 'Time stalled', 'Percent stalled',\n",
                "    'Median SSIM', 'Mean SSIM',\n",
                "    'Median $\\Delta$SSIM', 'Mean $\\Delta$SSIM',\n",
                "]\n",
                "\n",
                "filtered_results = (\n",
                "    raw_results\n",
                "    .assign(\n",
                "        varmetric=lambda df:\n",
                "            df[['variable', 'metric']].apply(tuple, axis=1).map(metrics),\n",
                "        days_running=lambda df: (df['day'] - deployment_start).dt.days,\n",
                "    )\n",
                "    .dropna(subset=[\"varmetric\"])  # Drop unsued variable/metric combos\n",
                "    .drop(columns=['metric', 'variable'])  # Drop individual columns.\n",
                ")\n",
                "\n",
                "# Let's have a look again.\n",
                "filtered_results.head()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# Aggregate statistics per ABR\n",
                "\n",
                "First of all, let's aggregate the statistics per ABR over the whole duration.\n",
                "We also include a measure of \"stream-years\" similar to the original Puffer paper"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 10,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": [
                        "Aggregate stats since the start of the current deployment.\n"
                    ]
                },
                {
                    "data": {
                        "text/html": [
                            "<div>\n",
                            "<style scoped>\n",
                            "    .dataframe tbody tr th:only-of-type {\n",
                            "        vertical-align: middle;\n",
                            "    }\n",
                            "\n",
                            "    .dataframe tbody tr th {\n",
                            "        vertical-align: top;\n",
                            "    }\n",
                            "\n",
                            "    .dataframe thead th {\n",
                            "        text-align: right;\n",
                            "    }\n",
                            "</style>\n",
                            "<table border=\"1\" class=\"dataframe\">\n",
                            "  <thead>\n",
                            "    <tr style=\"text-align: right;\">\n",
                            "      <th></th>\n",
                            "      <th>\\# Days</th>\n",
                            "      <th>\\# Sessions</th>\n",
                            "      <th>Stream-years</th>\n",
                            "      <th>Mean SSIM</th>\n",
                            "      <th>Time stalled [\\%]</th>\n",
                            "      <th>Time stalled [h]</th>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>abr</th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Memento}</th>\n",
                            "      <td>285</td>\n",
                            "      <td>740532.0</td>\n",
                            "      <td>10.636713</td>\n",
                            "      <td>17.080865</td>\n",
                            "      <td>0.204200</td>\n",
                            "      <td>190.784362</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Memento} (T\\texttt{=}0)</th>\n",
                            "      <td>285</td>\n",
                            "      <td>749037.0</td>\n",
                            "      <td>10.762023</td>\n",
                            "      <td>16.350492</td>\n",
                            "      <td>0.208638</td>\n",
                            "      <td>197.236081</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Fugu}</th>\n",
                            "      <td>10</td>\n",
                            "      <td>28445.0</td>\n",
                            "      <td>0.397015</td>\n",
                            "      <td>17.025307</td>\n",
                            "      <td>0.182109</td>\n",
                            "      <td>6.349260</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Fugu} (static)</th>\n",
                            "      <td>285</td>\n",
                            "      <td>746357.0</td>\n",
                            "      <td>10.696841</td>\n",
                            "      <td>17.111312</td>\n",
                            "      <td>0.237257</td>\n",
                            "      <td>222.997122</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "                                 \\# Days  \\# Sessions  Stream-years  \\\n",
                            "abr                                                                   \n",
                            "\\textsf{Memento}                     285     740532.0     10.636713   \n",
                            "\\textsf{Memento} (T\\texttt{=}0)      285     749037.0     10.762023   \n",
                            "\\textsf{Fugu}                         10      28445.0      0.397015   \n",
                            "\\textsf{Fugu} (static)               285     746357.0     10.696841   \n",
                            "\n",
                            "                                 Mean SSIM  Time stalled [\\%]  \\\n",
                            "abr                                                             \n",
                            "\\textsf{Memento}                 17.080865           0.204200   \n",
                            "\\textsf{Memento} (T\\texttt{=}0)  16.350492           0.208638   \n",
                            "\\textsf{Fugu}                    17.025307           0.182109   \n",
                            "\\textsf{Fugu} (static)           17.111312           0.237257   \n",
                            "\n",
                            "                                 Time stalled [h]  \n",
                            "abr                                                \n",
                            "\\textsf{Memento}                       190.784362  \n",
                            "\\textsf{Memento} (T\\texttt{=}0)        197.236081  \n",
                            "\\textsf{Fugu}                            6.349260  \n",
                            "\\textsf{Fugu} (static)                 222.997122  "
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "SECONDS_PER_HOUR = 3600\n",
                "SECONDS_PER_YEAR = 31556952\n",
                "\n",
                "def aggregate_stats(frame):\n",
                "    \"\"\"Aggregate stats per abr.\"\"\"\n",
                "    full = frame.query(\"selection == 'all'\")\n",
                "    stalled = frame.query(\"selection == 'stalled'\")\n",
                "\n",
                "    n_days = (\n",
                "        full\n",
                "        .groupby('abr')\n",
                "        ['day']\n",
                "        .nunique()\n",
                "        .to_frame('\\# Days')\n",
                "    )\n",
                "\n",
                "    n_sessions = (\n",
                "        full\n",
                "        .query(\"varmetric == '\\# Sessions'\")\n",
                "        .groupby('abr')\n",
                "        ['value']\n",
                "        .sum()\n",
                "        .to_frame('\\# Sessions')\n",
                "    )\n",
                "\n",
                "    average_ssim = (\n",
                "        full\n",
                "        .query(\"varmetric == 'Mean SSIM'\")\n",
                "        .groupby('abr')\n",
                "        ['value']\n",
                "        .mean()\n",
                "        .to_frame('Mean SSIM')\n",
                "    )\n",
                "\n",
                "    stream_years = (\n",
                "        full\n",
                "        .query(\"varmetric == 'Time played'\")\n",
                "        .groupby(\"abr\")\n",
                "        [\"value\"]\n",
                "        .sum()\n",
                "        .to_frame(\"Stream-years\")\n",
                "        / SECONDS_PER_YEAR  # Scale to stream-years.\n",
                "    )\n",
                "\n",
                "    def _sum(df):\n",
                "        \"\"\"Sum by ABR.\"\"\"\n",
                "        return df.groupby(['abr'])['value'].sum()\n",
                "\n",
                "    percent_time_stalled = (\n",
                "        (\n",
                "            # Time stalled.\n",
                "            _sum(full.query(\"varmetric == 'Time stalled'\")) /\n",
                "            _sum(full.query(\"varmetric == 'Time streamed'\"))\n",
                "            * 100\n",
                "        )\n",
                "        .to_frame(\"Time stalled [\\%]\")\n",
                "    )\n",
                "\n",
                "    hours_stalled = (\n",
                "        _sum(full.query(\"varmetric == 'Time stalled'\")) / SECONDS_PER_HOUR\n",
                "    ).to_frame(\"Time stalled [h]\")\n",
                "\n",
                "    return pd.concat([\n",
                "        n_days, n_sessions, stream_years,\n",
                "        average_ssim,\n",
                "        #percent_sessions_stalled,\n",
                "        percent_time_stalled,\n",
                "        hours_stalled\n",
                "    ], axis=1)\n",
                "\n",
                "\n",
                "print(\"Aggregate stats since the start of the current deployment.\")\n",
                "aggregate = aggregate_stats(filtered_results.query(\"day >= @deployment_start\"))\n",
                "display(aggregate)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "Also, compute the overall difference (in percent) between Memento and Fugu (static).\n",
                "\n",
                "We can only compare against a static version of Fugu, as the version of Fugu retrained daily from random samples was discontinued only a couple of days into the current deployment, too little to draw any conclusions.\n",
                "We will analyze the daily-retrained version compared to static Fugu afterwards.\n",
                "\n",
                "The default Memento has about the the same image quality (onlt 0.8% worse) as the dailt retrained Fugu, but users spend about 14% less of their stream time stalled: 0.2 of total stream time vs 0.23%. Overall, stalls are rare, they are the \"tail\" of the distribution.\n",
                "\n",
                "\n",
                "(Differences are shown in percent)"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 11,
            "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>\\# Days</th>\n",
                            "      <th>\\# Sessions</th>\n",
                            "      <th>Stream-years</th>\n",
                            "      <th>Mean SSIM</th>\n",
                            "      <th>Time stalled [\\%]</th>\n",
                            "      <th>Time stalled [h]</th>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>abr</th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Memento}</th>\n",
                            "      <td>0.0</td>\n",
                            "      <td>-0.780458</td>\n",
                            "      <td>-0.562111</td>\n",
                            "      <td>-0.177934</td>\n",
                            "      <td>-13.933230</td>\n",
                            "      <td>-14.445370</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Memento} (T\\texttt{=}0)</th>\n",
                            "      <td>0.0</td>\n",
                            "      <td>0.357793</td>\n",
                            "      <td>0.605669</td>\n",
                            "      <td>-4.446301</td>\n",
                            "      <td>-12.062667</td>\n",
                            "      <td>-11.552185</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "                                 \\# Days  \\# Sessions  Stream-years  \\\n",
                            "abr                                                                   \n",
                            "\\textsf{Memento}                     0.0    -0.780458     -0.562111   \n",
                            "\\textsf{Memento} (T\\texttt{=}0)      0.0     0.357793      0.605669   \n",
                            "\n",
                            "                                 Mean SSIM  Time stalled [\\%]  \\\n",
                            "abr                                                             \n",
                            "\\textsf{Memento}                 -0.177934         -13.933230   \n",
                            "\\textsf{Memento} (T\\texttt{=}0)  -4.446301         -12.062667   \n",
                            "\n",
                            "                                 Time stalled [h]  \n",
                            "abr                                                \n",
                            "\\textsf{Memento}                       -14.445370  \n",
                            "\\textsf{Memento} (T\\texttt{=}0)        -11.552185  "
                        ]
                    },
                    "execution_count": 11,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": [
                "maxpd(\n",
                "    aggregate.loc[[memento_name, memento_deterministic]],\n",
                "    aggregate.loc[fugufeb_name].values\n",
                ")"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "Now compare Fugu (daily) and fugu_static during the whole time when both algorithms were deployed.\n",
                "Users that got assigned the daily-retrained model only spent 4% smaller fraction of their stream-time stalled overall, with almost identical image quality.\n",
                "\n",
                "(Differences are shown in percent)"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 12,
            "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>\\# Days</th>\n",
                            "      <th>\\# Sessions</th>\n",
                            "      <th>Stream-years</th>\n",
                            "      <th>Mean SSIM</th>\n",
                            "      <th>Time stalled [\\%]</th>\n",
                            "      <th>Time stalled [h]</th>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>abr</th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Fugu}</th>\n",
                            "      <td>871</td>\n",
                            "      <td>2089921.0</td>\n",
                            "      <td>28.197023</td>\n",
                            "      <td>17.063813</td>\n",
                            "      <td>0.238645</td>\n",
                            "      <td>591.269051</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Fugu} (static)</th>\n",
                            "      <td>871</td>\n",
                            "      <td>2154876.0</td>\n",
                            "      <td>29.501597</td>\n",
                            "      <td>17.036189</td>\n",
                            "      <td>0.248907</td>\n",
                            "      <td>645.293300</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "                        \\# Days  \\# Sessions  Stream-years  Mean SSIM  \\\n",
                            "abr                                                                     \n",
                            "\\textsf{Fugu}               871    2089921.0     28.197023  17.063813   \n",
                            "\\textsf{Fugu} (static)      871    2154876.0     29.501597  17.036189   \n",
                            "\n",
                            "                        Time stalled [\\%]  Time stalled [h]  \n",
                            "abr                                                          \n",
                            "\\textsf{Fugu}                    0.238645        591.269051  \n",
                            "\\textsf{Fugu} (static)           0.248907        645.293300  "
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                },
                {
                    "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>\\# Days</th>\n",
                            "      <th>\\# Sessions</th>\n",
                            "      <th>Stream-years</th>\n",
                            "      <th>Mean SSIM</th>\n",
                            "      <th>Time stalled [\\%]</th>\n",
                            "      <th>Time stalled [h]</th>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>abr</th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "      <th></th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>\\textsf{Fugu}</th>\n",
                            "      <td>0.0</td>\n",
                            "      <td>-3.014327</td>\n",
                            "      <td>-4.422047</td>\n",
                            "      <td>0.161888</td>\n",
                            "      <td>-4.122889</td>\n",
                            "      <td>-8.372046</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "               \\# Days  \\# Sessions  Stream-years  Mean SSIM  \\\n",
                            "abr                                                            \n",
                            "\\textsf{Fugu}      0.0    -3.014327     -4.422047   0.161888   \n",
                            "\n",
                            "               Time stalled [\\%]  Time stalled [h]  \n",
                            "abr                                                 \n",
                            "\\textsf{Fugu}          -4.122889         -8.372046  "
                        ]
                    },
                    "execution_count": 12,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": [
                "both_fugus = (  # Days for with we have both Fugu and FuguFeb.\n",
                "    set(filtered_results.query(\"abr == @fugu_name\")['day']).intersection(\n",
                "        filtered_results.query(\"abr == @fugufeb_name\")['day']\n",
                "    )\n",
                ")\n",
                "\n",
                "fugu_aggregate = aggregate_stats(\n",
                "    filtered_results\n",
                "    .query(\"(day in @both_fugus) and (abr in [@fugu_name, @fugufeb_name])\")\n",
                ").dropna()\n",
                "display(fugu_aggregate)\n",
                "maxpd(\n",
                "    fugu_aggregate.loc[[fugu_name]],\n",
                "    fugu_aggregate.loc[fugufeb_name].values\n",
                ")"
            ]
        },
        {
            "attachments": {},
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# ECDFs of ABR differences [Fig. 6 (a, c, d)]\n",
                "\n",
                "In addition to the aggregated performance, we can look at the ECDF of daily differences to show the performance in more detail.\n",
                "As some days have much more data than others, we prepare the total stream time each day as weights (video data plus rebuffering, i.e. the total time users spent on stream)."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 13,
            "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>abr</th>\n",
                            "      <th>value</th>\n",
                            "      <th>selection</th>\n",
                            "      <th>day</th>\n",
                            "      <th>varmetric</th>\n",
                            "      <th>days_running</th>\n",
                            "      <th>abr_weight</th>\n",
                            "      <th>value_feb</th>\n",
                            "      <th>abr_weight_feb</th>\n",
                            "      <th>value_diff</th>\n",
                            "      <th>value_diff_rel</th>\n",
                            "      <th>weight</th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>0</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>17.377717</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Mean SSIM</td>\n",
                            "      <td>143</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>17.377717</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>0.000000</td>\n",
                            "      <td>0.000000</td>\n",
                            "      <td>1507521.098</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>18</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>17.447852</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Mean SSIM</td>\n",
                            "      <td>143</td>\n",
                            "      <td>832863.011</td>\n",
                            "      <td>17.377717</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>0.070135</td>\n",
                            "      <td>0.401968</td>\n",
                            "      <td>1586623.560</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>36</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>16.802479</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Mean SSIM</td>\n",
                            "      <td>143</td>\n",
                            "      <td>982426.118</td>\n",
                            "      <td>17.377717</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>-0.575238</td>\n",
                            "      <td>-3.310205</td>\n",
                            "      <td>1736186.667</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>144</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>751833.729000</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Time played</td>\n",
                            "      <td>143</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>751833.729000</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>0.000000</td>\n",
                            "      <td>0.000000</td>\n",
                            "      <td>1507521.098</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>145</th>\n",
                            "      <td>\\textsf{Fugu} (static)</td>\n",
                            "      <td>1926.820000</td>\n",
                            "      <td>all</td>\n",
                            "      <td>2023-02-09</td>\n",
                            "      <td>Time stalled</td>\n",
                            "      <td>143</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>1926.820000</td>\n",
                            "      <td>753760.549</td>\n",
                            "      <td>0.000000</td>\n",
                            "      <td>0.000000</td>\n",
                            "      <td>1507521.098</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "                                 abr          value selection        day  \\\n",
                            "0             \\textsf{Fugu} (static)      17.377717       all 2023-02-09   \n",
                            "18                  \\textsf{Memento}      17.447852       all 2023-02-09   \n",
                            "36   \\textsf{Memento} (T\\texttt{=}0)      16.802479       all 2023-02-09   \n",
                            "144           \\textsf{Fugu} (static)  751833.729000       all 2023-02-09   \n",
                            "145           \\textsf{Fugu} (static)    1926.820000       all 2023-02-09   \n",
                            "\n",
                            "        varmetric  days_running  abr_weight      value_feb  abr_weight_feb  \\\n",
                            "0       Mean SSIM           143  753760.549      17.377717      753760.549   \n",
                            "18      Mean SSIM           143  832863.011      17.377717      753760.549   \n",
                            "36      Mean SSIM           143  982426.118      17.377717      753760.549   \n",
                            "144   Time played           143  753760.549  751833.729000      753760.549   \n",
                            "145  Time stalled           143  753760.549    1926.820000      753760.549   \n",
                            "\n",
                            "     value_diff  value_diff_rel       weight  \n",
                            "0      0.000000        0.000000  1507521.098  \n",
                            "18     0.070135        0.401968  1586623.560  \n",
                            "36    -0.575238       -3.310205  1736186.667  \n",
                            "144    0.000000        0.000000  1507521.098  \n",
                            "145    0.000000        0.000000  1507521.098  "
                        ]
                    },
                    "execution_count": 13,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": [
                "\n",
                "# Add weights (time streamed per day) for each ABR.\n",
                "# First get the weights per day, then add them as a separate column.\n",
                "weight = (\n",
                "    filtered_results\n",
                "    .query(\"varmetric == 'Time streamed'\")\n",
                "    .set_index([\"abr\", \"day\", \"selection\"])\n",
                "    [\"value\"]\n",
                "    .to_frame(\"abr_weight\")\n",
                ")\n",
                "weighted = filtered_results.join(weight, on=weight.index.names)\n",
                "\n",
                "# Add Fugufeb comparison.\n",
                "fugufeb_values = (\n",
                "    weighted\n",
                "    .query(\"abr == @fugufeb_name\")\n",
                "    .set_index( [\"day\", \"selection\", \"varmetric\"] )\n",
                "    [[\"value\", \"abr_weight\"]]\n",
                ")\n",
                "eval_results_full = (\n",
                "    weighted\n",
                "    .join(fugufeb_values, on=fugufeb_values.index.names, rsuffix=\"_feb\")\n",
                "    .assign(\n",
                "        # Absolute and relative difference (max percent difference).\n",
                "        value_diff=lambda df: diff(df['value'], df['value_feb']),\n",
                "        value_diff_rel=lambda df: maxpd(df['value'], df['value_feb']),\n",
                "        weight=lambda df: df['abr_weight'] + df['abr_weight_feb'],\n",
                "    )\n",
                ")\n",
                "\n",
                "# Separate by selection, all and stalled only.\n",
                "comparison = eval_results_full.query(\"selection == 'all'\")\n",
                "comparison_stalled = eval_results_full.query(\"selection == 'stalled'\")\n",
                "\n",
                "# For the ECDF plots for Memento in deployment, select the correct subset.\n",
                "comparison_deployment = (\n",
                "    comparison\n",
                "    .query(\"(day >= @deployment_start)\"\n",
                "           \" and (abr in [@memento_name, @memento_deterministic]) \")\n",
                ")\n",
                "\n",
                "# Peek.\n",
                "comparison.head()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Figure 6 (a): Logscore Improvement"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 14,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 360.455x360.455 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "_median_and_tail = (\n",
                "    comparison_deployment\n",
                "    .query(\"varmetric == ['Score Median', 'Score Tail']\")\n",
                "    .pivot(\n",
                "        index=[\"abr\", \"day\", \"selection\", 'weight'],\n",
                "        columns=\"varmetric\",\n",
                "        values=\"value_diff\"\n",
                "    )\n",
                "    .reset_index()\n",
                ")\n",
                "\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=_median_and_tail,\n",
                "    hue=\"abr\",\n",
                "    hue_order=[memento_name, memento_deterministic],\n",
                "    palette=abr_palette,\n",
                "    **eh.plot_utils.opts_square_small(is_grid=True)\n",
                ")\n",
                "\n",
                "grid.map_dataframe(\n",
                "    sns.ecdfplot, x='Score Median', weights=\"weight\",\n",
                ")\n",
                "\n",
                "grid.add_legend(title=None)\n",
                "\n",
                "\n",
                "grid.map_dataframe(\n",
                "    sns.ecdfplot, x='Score Tail', weights=\"weight\",\n",
                "    linestyle='--',\n",
                ")\n",
                "\n",
                "grid.set(\n",
                "    ylabel=\"\",\n",
                "    xlabel=\"Logscore Improvement\",\n",
                "    xlim=(-1, 3),\n",
                "    xticks=np.arange(-1, 3 + 1, 1),\n",
                ")\n",
                "\n",
                "# Touch up legend and add annotations.\n",
                "legend = plt.gca().legend([\n",
                "    f\"{memento_name}\",\n",
                "    r\"Mem.(T\\texttt{=}0)\",\n",
                "    #\"Median\",\n",
                "    #\"1st P.\",\n",
                "], loc=\"lower right\", handlelength=1, handletextpad=0.5,\n",
                "    bbox_to_anchor=(1.02, -0.02),\n",
                ")\n",
                "\n",
                "plt.gca().text(\n",
                "    0.55, 0.92, \"better\",\n",
                "    ha=\"center\", va=\"center\", rotation=0, size=15,\n",
                "    bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1))\n",
                "\n",
                "plt.gca().text(\n",
                "    -0.55, 0.5, \"Median\",\n",
                "    ha=\"center\", va=\"center\", rotation=85, size=15,\n",
                "    #bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1)\n",
                ")\n",
                "plt.gca().text(\n",
                "    1.25, 0.5, \"1st Perc.\",\n",
                "    ha=\"center\", va=\"center\", rotation=85, size=15,\n",
                "    #bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1)\n",
                ")\n",
                "\n",
                "plt.axvline(0, color='gray', linestyle='-', zorder=0)\n",
                "plt.savefig(plotdir / \"ecdf_logscore_diff.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# Figure 14: Logscore Improvement in comparison to Fugu (daily)\n",
                "\n",
                "While we cannot compare directly to a daily-retrained version of Fugu for lack of data, we can look at the logscore difference between Fugu (static) and Fugu (daily retrained from random samples of the last two weeks) in the past.\n",
                "\n",
                "We can see that they have an almost identical median score, but retraining from random samples does not yield consistent tail score improvements.\n",
                "For about 20% of days, the tail score is even _worse_ than Fugu(static), and for the remaining days the tail gains are significantly smaller than for Memento.\n",
                "\n",
                "While the relationship between an ML model in a control loop with the overall QoE is hard to quantify, this may explain why we see considerably less improvement in stalls when using a model retrained from random samples."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 15,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 360.455x360.455 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "_median_and_tail_fugu = (\n",
                "    comparison\n",
                "    .query(\"(abr == @fugu_name) and day in @both_fugus\")\n",
                "    .dropna(subset=['value_diff'])\n",
                "    .query(\"varmetric == ['Score Median', 'Score Tail']\")\n",
                "    .pivot(\n",
                "        index=[\"abr\", \"day\", \"selection\", 'weight'],\n",
                "        columns=\"varmetric\",\n",
                "        values=\"value_diff\"\n",
                "    )\n",
                "    .reset_index()\n",
                ")\n",
                "\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=pd.concat([_median_and_tail, _median_and_tail_fugu]),\n",
                "    hue=\"abr\",\n",
                "    hue_order=[memento_name, memento_deterministic, fugu_name],\n",
                "    palette=abr_palette,\n",
                "    **eh.plot_utils.opts_square_small(is_grid=True)\n",
                ")\n",
                "\n",
                "grid.map_dataframe(\n",
                "    sns.ecdfplot, x='Score Median', weights=\"weight\",\n",
                ")\n",
                "\n",
                "grid.add_legend(title=None)\n",
                "\n",
                "\n",
                "grid.map_dataframe(\n",
                "    sns.ecdfplot, x='Score Tail', weights=\"weight\",\n",
                "    linestyle='--',\n",
                ")\n",
                "\n",
                "grid.set(\n",
                "    ylabel=\"\",\n",
                "    xlabel=\"Logscore Improvement\",\n",
                "    xlim=(-1, 3),\n",
                "    xticks=np.arange(-1, 3 + 1, 1),\n",
                ")\n",
                "\n",
                "# Touch up legend and add annotations.\n",
                "legend = plt.gca().legend([\n",
                "    f\"{memento_name}\",\n",
                "    r\"Mem.(T\\texttt{=}0)\",\n",
                "    \"Past Fugu\",\n",
                "], loc=\"lower right\", handlelength=1, handletextpad=0.5,\n",
                "    bbox_to_anchor=(1.02, -0.02),\n",
                ")\n",
                "\n",
                "plt.gca().text(\n",
                "    0.55, 0.92, \"better\",\n",
                "    ha=\"center\", va=\"center\", rotation=0, size=15,\n",
                "    bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1))\n",
                "\n",
                "plt.gca().text(\n",
                "    -0.55, 0.5, \"Median\",\n",
                "    ha=\"center\", va=\"center\", rotation=85, size=15,\n",
                "    #bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1)\n",
                ")\n",
                "plt.gca().text(\n",
                "    1.25, 0.5, \"1st Perc.\",\n",
                "    ha=\"center\", va=\"center\", rotation=85, size=15,\n",
                "    #bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1)\n",
                ")\n",
                "\n",
                "plt.axvline(0, color='gray', linestyle='-', zorder=0)\n",
                "plt.savefig(plotdir / \"ecdf_logscore_diff_with_past_fugu.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Figure 6 (c): Relative SSIM improvement"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 16,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 360.455x360.455 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "grid = sns.displot(\n",
                "    kind='ecdf',\n",
                "    data=comparison_deployment.query(\"varmetric == 'Mean SSIM'\"),\n",
                "    x='value_diff_rel',\n",
                "    weights=\"weight\",\n",
                "    hue=\"abr\", hue_order=[memento_name, memento_deterministic], palette=abr_palette,\n",
                "    **eh.plot_utils.opts_square_small()\n",
                ").set(\n",
                "    ylabel=\"\",\n",
                "    xlabel=\"SSIM Improvement [\\%]\",\n",
                "    xlim=(-10, 2),\n",
                "    xticks=np.arange(-10, 2 + 1, 2)\n",
                ")\n",
                "grid.legend.set_title(None)\n",
                "\n",
                "sns.move_legend(\n",
                "    plt.gca(), \"upper left\", title=None,\n",
                "    handlelength=1, handletextpad=0.5, #bbox_to_anchor=(1.02, -0.02),\n",
                ")\n",
                "\n",
                "\n",
                "plt.axvline(0, color='gray', linestyle='-', zorder=0)\n",
                "\n",
                "plt.gca().text(\n",
                "    0, 0.1, \"better\",\n",
                "    ha=\"center\", va=\"center\", rotation=0, size=15,\n",
                "    bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1))\n",
                "\n",
                "plt.savefig(plotdir / \"ecdf_ssim_rel.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Figure 6 (d): Relative stall improvement"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 17,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 360.455x360.455 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "stalltimediff = (\n",
                "    comparison_deployment.query(\"varmetric == 'Percent stalled'\")\n",
                "    # Flip such that less = better\n",
                "    .assign(value_diff_rel=lambda df: -1*df['value_diff_rel'])\n",
                ")\n",
                "\n",
                "grid = sns.displot(\n",
                "    kind='ecdf',\n",
                "    data=stalltimediff,\n",
                "    x='value_diff_rel',\n",
                "    weights=\"weight\",\n",
                "    hue=\"abr\", hue_order=[memento_name, memento_deterministic], palette=abr_palette,\n",
                "    **eh.plot_utils.opts_square_small()\n",
                ").set(\n",
                "    ylabel=\"ECDF\",\n",
                "    xlabel=\"Stall Improvement [\\%]\",\n",
                "    xlim=(-100, 100),\n",
                ")\n",
                "sns.move_legend(\n",
                "    plt.gca(), \"upper left\", title=None,\n",
                "    handlelength=1, handletextpad=0.5, bbox_to_anchor=(-0.02, 1.02),\n",
                ")\n",
                "\n",
                "plt.gca().text(\n",
                "    0, 0.1, \"better\",\n",
                "    ha=\"center\", va=\"center\", rotation=0, size=15,\n",
                "    bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1))\n",
                "\n",
                "plt.axvline(0, color='gray', linestyle='-', zorder=0)\n",
                "plt.savefig(plotdir / \"ecdf_stall_rel.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "Where exactly do the lines cross zero? That is, on which fraction of days is Memento worse?"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 18,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "text/plain": [
                            "abr\n",
                            "\\textsf{Memento}                   0.428022\n",
                            "\\textsf{Memento} (T\\texttt{=}0)    0.342407\n",
                            "Name: weight, dtype: float64"
                        ]
                    },
                    "execution_count": 18,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": [
                "# Where do the lines cross 0?\n",
                "((\n",
                "    stalltimediff.query(\"value_diff_rel <= 0\")\n",
                "    .groupby(\"abr\")[\"weight\"].sum()\n",
                "\n",
                ") / (\n",
                "    stalltimediff\n",
                "     .groupby(\"abr\")[\"weight\"].sum()\n",
                ")).dropna()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# Timeseries plots [Fig. 1, 18, 19, 20]\n",
                "\n",
                "Timeseries plots with a rolling mean and a (also rolling) 90% confidence interval."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 19,
            "metadata": {},
            "outputs": [],
            "source": [
                "def rolling(frame, days):\n",
                "    \"\"\"Prepare a frame so a rolling apply can be achieved by grouping by day.\n",
                "\n",
                "    Only needed because pandas.rolling.apply only works per-column.\n",
                "\n",
                "    Concretely, copy the frame and increase the day, so that in the end a given\n",
                "    day contains data from the past `days` frames.\n",
                "    \"\"\"\n",
                "    daycol = frame['day']\n",
                "    last_day = daycol.max()\n",
                "    copies = [frame.assign(day=daycol + pd.Timedelta(days=ind))\n",
                "              for ind in range(days)]\n",
                "    # Merge and remove days shifted into the future.\n",
                "    return (\n",
                "        pd\n",
                "        .concat(copies)\n",
                "        [lambda df: df['day'] <= last_day]\n",
                "        .reset_index(drop=True)\n",
                "    )\n",
                "\n",
                "def bootstrap(\n",
                "    data, metric=None,\n",
                "    col=\"value\", comp_col=\"value_feb\",\n",
                "    weight=\"abr_weight\",  weight_comp=\"abr_weight_feb\",\n",
                "    estimator=weighted_mean_raw, difference=maxpd,\n",
                "    n_resamples=1000, confidence_level=0.9, random_state=42,\n",
                "    format=False,\n",
                "    ):\n",
                "    \"\"\"Bootstrap the difference between two ABRs.\"\"\"\n",
                "    if metric is not None:\n",
                "        data = data[lambda df: df['metric'] == metric]\n",
                "\n",
                "    # Inputs for bootstrap: col, comp_col, weight\n",
                "    if (weight is not None) and (weight_comp is not None):\n",
                "        arrays = (\n",
                "            data[col].values, data[comp_col].values,\n",
                "            data[weight].values, data[weight_comp].values\n",
                "        )\n",
                "    else:\n",
                "        arrays = (data[col].values, data[comp_col].values,)\n",
                "\n",
                "    def _diff(*arrays):\n",
                "        \"\"\"Aggregate the arrays.\"\"\"\n",
                "        if (weight is not None) and (weight_comp is not None):\n",
                "            x1, x2, w1, w2 = arrays\n",
                "        else:\n",
                "            x1, x2 = arrays\n",
                "            w1, w2 = None, None\n",
                "\n",
                "        if difference is None:\n",
                "            return estimator(x1, w1)\n",
                "        return difference(estimator(x1, w1), estimator(x2, w2))\n",
                "\n",
                "    _est = _diff(*arrays)\n",
                "    if len(arrays[0]) == 1:\n",
                "        # Cannot bootstrap with only one day; return the single value.\n",
                "        _ci = (_est, _est)\n",
                "    else:\n",
                "        _ci = stats.bootstrap(\n",
                "            arrays, _diff, method='percentile',\n",
                "            n_resamples=n_resamples,\n",
                "            confidence_level=confidence_level,\n",
                "            random_state=random_state\n",
                "        ).confidence_interval\n",
                "\n",
                "    if format:\n",
                "        return f\"{_est:.2f} [{_ci[0]:.2f}, {_ci[1]:.2f}]\"\n",
                "    return (_est, *_ci)\n",
                "\n",
                "\n",
                "def plot_timeseries(\n",
                "    data=None,\n",
                "    x=\"day\",\n",
                "    window=14,\n",
                "    metric=None,\n",
                "    ax=None,\n",
                "    color=None,\n",
                "    alpha=0.2,\n",
                "    label=None,\n",
                "    bootstrap=bootstrap,\n",
                "    bootstrap_kwargs=None,\n",
                "    color_code=False,\n",
                "    xdate=True,\n",
                "    invert=tuple(),\n",
                "    styles=None,\n",
                "    **plot_kwargs\n",
                "):\n",
                "    \"\"\"Plot a rolling average with bootstrapped confidence intervals.\"\"\"\n",
                "    if ax is None:\n",
                "        ax = plt.gca()\n",
                "    if color is None:\n",
                "        color = ax._get_lines.get_next_color()\n",
                "    if bootstrap_kwargs is None:\n",
                "        bootstrap_kwargs = {}\n",
                "    if styles is not None and label in styles:\n",
                "        plot_kwargs.update(linestyle=styles[label])\n",
                "\n",
                "    if metric is not None:\n",
                "        data = data.query(\"metric == @metric\")\n",
                "    data = data.sort_values(by=\"day\")\n",
                "    assert x == 'day'\n",
                "\n",
                "    # Compute the rolling estimates.\n",
                "    results = (\n",
                "        rolling(data, window)\n",
                "        .groupby(\"day\")\n",
                "        .apply(bootstrap, **bootstrap_kwargs)\n",
                "    )\n",
                "    x = results.index.values\n",
                "    if not xdate:\n",
                "        # Use #days instead of datetime.\n",
                "        x = (x - x[0]) / np.timedelta64(1, \"D\")\n",
                "\n",
                "    y = results.map(itemgetter(0)).values\n",
                "    low = results.map(itemgetter(1)).values\n",
                "    high = results.map(itemgetter(2)).values\n",
                "\n",
                "    if label in invert:\n",
                "        # Invert the sign so that lower values mean better.\n",
                "        y, low, high = -y, -high, -low\n",
                "\n",
                "    if not color_code:\n",
                "        if window > 1:\n",
                "            ax.fill_between(x, low, high, color=color, alpha=alpha)\n",
                "        ax.plot(x, y, color=color, **plot_kwargs, label=label)\n",
                "    else:\n",
                "        green = sns.color_palette(\"deep\")[2]\n",
                "        red = sns.color_palette(\"deep\")[3]\n",
                "\n",
                "        if window > 1:\n",
                "            # Increase: red, everything above 0\n",
                "            ilow = np.where(low >= 0, low, 0)\n",
                "            ihigh = np.where(high >= 0, high, 0)\n",
                "            ax.fill_between(x, ilow, ihigh, color=red, alpha=alpha)\n",
                "            # Decrease: green, everything below 0\n",
                "            dlow = np.where(low <= 0, low, 0)\n",
                "            dhigh = np.where(high <= 0, high, 0)\n",
                "            ax.fill_between(x, dlow, dhigh, color=green, alpha=alpha)\n",
                "\n",
                "        # Find crossing points for smooth color transition.\n",
                "        # See: https://stackoverflow.com/a/70850749\n",
                "        x1, x2 = np.stack([x[:-1],  x[1:]])[:,np.diff(y < 0)]\n",
                "        y1, y2 = np.stack([y[:-1], y[1:]])[:,np.diff(y < 0)]\n",
                "        xf = x1 + -y1 * (x2 - x1) / (y2 - y1)\n",
                "\n",
                "        i = np.searchsorted(x, xf)\n",
                "        x0 = np.insert(x, i, xf)\n",
                "        y0 = np.insert(y, i, 0)\n",
                "\n",
                "        above = y0.copy()\n",
                "        above[above < 0] = np.nan\n",
                "        below = y0.copy()\n",
                "        below[below > 0] = np.nan\n",
                "\n",
                "        ax.plot(x0, above, color=red, **plot_kwargs, label=label)\n",
                "        ax.plot(x0, below, color=green, **plot_kwargs, label=label)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Figure 1: Fugu (daily) vs Fugu (static)\n",
                "\n",
                "The SSIM difference is very small, so we add a little zoom-in window to show that there is in fact some difference, but insignificant.\n",
                "\n",
                "Note: computing the rolling confidence intervals takes some minutes."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 20,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x450.568 with 3 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "both_fugus = (  # Days for with we have both Fugu and FuguFeb.\n",
                "    set(comparison.query(\"abr == @fugu_name\")['day']).intersection(\n",
                "        comparison.query(\"abr == @fugufeb_name\")['day']\n",
                "    )\n",
                ")\n",
                "\n",
                "firstpage_selection = comparison.query(\n",
                "    \"(day in @both_fugus) & (abr in [@fugu_name])\"\n",
                "    \"& (varmetric in ['Mean SSIM', 'Percent stalled'])\"\n",
                ").dropna()\n",
                "\n",
                "# Plot with a 1-month rolling window.\n",
                "window=31\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=firstpage_selection,\n",
                "    row=\"varmetric\", row_order=[\"Mean SSIM\", \"Percent stalled\"],\n",
                "    hue=\"abr\", palette=abr_palette,\n",
                "    aspect=eh.plot_utils.default_aspect * 2,\n",
                "    height=eh.plot_utils.default_height / 2,\n",
                "    sharey=False, despine=False,\n",
                ").map_dataframe(\n",
                "    plot_timeseries,\n",
                "    window=window,\n",
                "    color_code=True,\n",
                ").set(\n",
                "    title=None,\n",
                "    ylim=(-100, 100),\n",
                "    yticks=[-100, -50, 0, 50, 100],\n",
                "    xlabel=\"\",\n",
                ")\n",
                "\n",
                "# Manual adjustments\n",
                "fig = plt.gcf()\n",
                "\n",
                "# SSIM axis\n",
                "fig.axes[0].set(\n",
                "    #ylim=(-100, 100),\n",
                "    ylabel=\"SSIM [\\%]\",\n",
                ")\n",
                "fig.axes[0].set_title(\n",
                "    \"Relative Difference (Daily-retrained $-$ Static)\",\n",
                "    pad=10,\n",
                ")\n",
                "# Stall axis\n",
                "fig.axes[1].set(\n",
                "    #xlabel=\"\",\n",
                "    #ylim=(-100, 100),\n",
                "    ylabel=\"Time Stalled [\\%]\",\n",
                "\n",
                ")\n",
                "fig.axes[1].xaxis.set_major_locator(mdates.MonthLocator(bymonth=(1, 7)))\n",
                "fig.axes[1].xaxis.set_major_formatter(mdates.DateFormatter(\"%b %y\"))\n",
                "\n",
                "green = sns.color_palette(\"deep\")[2]\n",
                "red = sns.color_palette(\"deep\")[3]\n",
                "\n",
                "mini_lim=(-3, 3)\n",
                "height=np.abs(mini_lim[1]) + np.abs(mini_lim[0])\n",
                "mini_start = pd.to_datetime(\"2020-10-01\")\n",
                "mini_dur = pd.to_timedelta(\"30D\")\n",
                "\n",
                "\n",
                "_edgec = \"gray\"\n",
                "\n",
                "\n",
                "miniax = plt.axes((0.35, 0.75, 0.2, 0.12))\n",
                "for spine in miniax.spines.values():\n",
                "    spine.set_edgecolor(_edgec)\n",
                "\n",
                "rect = patches.Rectangle(\n",
                "    (mini_start, mini_lim[0]), mini_dur, height, linewidth=1,\n",
                "    edgecolor=_edgec, facecolor='none', zorder=10\n",
                ")\n",
                "fig.axes[0].add_patch(rect)\n",
                "\n",
                "\n",
                "_start = mini_start - pd.to_timedelta(f\"{window}D\")\n",
                "_end = mini_start + mini_dur + pd.to_timedelta(f\"{window}D\")\n",
                "\n",
                "plot_timeseries(\n",
                "    data=firstpage_selection.query(\n",
                "        \"(varmetric == 'Mean SSIM') and (day >= @_start) and (day <= @_end)\"\n",
                "    ),\n",
                "    ax=miniax,\n",
                "    window=window,\n",
                "    color_code=True,\n",
                ")\n",
                "miniax.set(\n",
                "    xlabel=None, ylabel=None,\n",
                "    xlim=(mini_start, mini_start + mini_dur),\n",
                "    ylim=mini_lim,\n",
                "    yticks=[],\n",
                "    xticks=[],\n",
                ")\n",
                "fig.axes[0].annotate(\n",
                "    \"\", xy=(mini_start, mini_lim[1]),\n",
                "    xytext=(0.295, 0.915), textcoords=\"axes fraction\",\n",
                "    arrowprops=dict(arrowstyle=\"-\", color=_edgec),\n",
                ")\n",
                "fig.axes[0].annotate(\n",
                "    \"\", xy=(mini_start + mini_dur, mini_lim[0]),\n",
                "    xytext=(0.535, 0.561), textcoords=\"axes fraction\",\n",
                "    arrowprops=dict(arrowstyle=\"-\", color=_edgec),\n",
                ")\n",
                "\n",
                "\n",
                "\n",
                "fig.align_ylabels()\n",
                "#fig.tight_layout()\n",
                "plt.savefig(plotdir / \"firstpage_puffer.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "### Figure 17: Logscores Improvement with Memento over time\n",
                "\n",
                "For Memento, and for Memento (T=0). We use a two-week sliding window as the time interval is shorter than before.\n",
                "\n",
                "Logscores are not a linear scale, so we plot the absolute difference."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 21,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x300.379 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "current_logscores = comparison.query(\n",
                "    \"(day >= @deployment_start) & (abr == @memento_name)\"\n",
                "    \" & (varmetric in ['Score Median', 'Score Tail'])\"\n",
                ").dropna()\n",
                "\n",
                "# Nicer xticks.\n",
                "n_days = (current_logscores.day.max() - current_logscores.day.min()).days\n",
                "tickfreq = 7 * 3  # every 3 weeks\n",
                "xticks = np.arange(0, n_days + 1, tickfreq)\n",
                "xticklabels = np.arange(len(xticks)) * 3\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=current_logscores,\n",
                "    hue=\"varmetric\", hue_order=['Score Median', 'Score Tail'],\n",
                "    **eh.plot_utils.opts_short(is_grid=True)\n",
                ").map_dataframe(\n",
                "    plot_timeseries,\n",
                "    window=14,\n",
                "    bootstrap_kwargs=dict(n_resamples=1000, difference=diff),\n",
                "    xdate=False,\n",
                "    styles={\"Score Tail\": \"--\"},\n",
                ").set(\n",
                "    title=None,\n",
                "    ylim=(-1, 2.5),\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlim=(0, n_days),\n",
                "    ylabel=\"Logscore Improvement\",\n",
                ")\n",
                "\n",
                "# Nicer legend.\n",
                "fig = plt.gcf()\n",
                "h, l = fig.axes[0].get_legend_handles_labels()\n",
                "fig.axes[0].legend(\n",
                "    title=None, handles=h, labels=[\"Median\", \"1st Percentile\"],\n",
                "    loc=\"upper right\",\n",
                ")\n",
                "\n",
                "plt.savefig(plotdir / \"logscore_improvement.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 22,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x300.379 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "current_logscores_t0 = comparison.query(\n",
                "    \"(day >= @deployment_start) & (abr == @memento_deterministic)\"\n",
                "    \" & (varmetric in ['Score Median', 'Score Tail'])\"\n",
                ")#.dropna(subset=\"value\")\n",
                "\n",
                "n_days = (current_logscores_t0.day.max() - current_logscores_t0.day.min()).days\n",
                "tickfreq = 7 * 3  # every 3 weeks\n",
                "xticks = np.arange(0, n_days + 1, tickfreq)\n",
                "xticklabels = np.arange(len(xticks)) * 3\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=current_logscores_t0,\n",
                "    hue=\"varmetric\", hue_order=['Score Median', 'Score Tail'],\n",
                "    **eh.plot_utils.opts_short(is_grid=True)\n",
                ").map_dataframe(\n",
                "    plot_timeseries,\n",
                "    window=14,\n",
                "    bootstrap_kwargs=dict(n_resamples=1000, difference=diff),\n",
                "    xdate=False,\n",
                "    styles={\"Score Tail\": \"--\"},\n",
                ").set(\n",
                "    title=None,\n",
                "    ylim=(-1, 2.5),\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlim=(0, n_days),\n",
                "    ylabel=\"Logscore Improvement\",\n",
                ")\n",
                "\n",
                "# Nicer legend.\n",
                "fig = plt.gcf()\n",
                "h, l = fig.axes[0].get_legend_handles_labels()\n",
                "fig.axes[0].legend(\n",
                "    title=None, handles=h, labels=[\"Median\", \"1st Percentile\"],\n",
                "    loc=\"upper right\",\n",
                ")\n",
                "\n",
                "plt.savefig(plotdir / \"logscore_improvement_deterministic.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "### Figure 18a and 19a: SSIM over time; raw (14) and smoothed (15)\n",
                "\n",
                "For Memento and Memento (T=0) in the current deployment.\n",
                "For the smoothed case, again with 14-day rolling window and CI."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 23,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x300.379 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "ssim_over_time = comparison.query(\n",
                "    \"(day >= @deployment_start) & (abr in @abr_order_short)\"\n",
                "    #\" & (day not in @worst_days)\"\n",
                "    \" & (varmetric in ['Mean SSIM'])\"\n",
                ").dropna()\n",
                "\n",
                "n_days = (ssim_over_time.day.max() - ssim_over_time.day.min()).days\n",
                "tickfreq = 7 * 3  # every 3 weeks\n",
                "xticks = np.arange(0, n_days + 1, tickfreq)\n",
                "xticklabels = np.arange(len(xticks)) * 3\n",
                "\n",
                "\n",
                "grid = sns.relplot(\n",
                "    kind='line',\n",
                "    data=ssim_over_time,\n",
                "    x=\"days_running\", y=\"value\",\n",
                "    hue=\"abr\", hue_order=abr_order_short, palette=abr_palette,\n",
                "    **eh.plot_utils.opts_short()\n",
                ").set(\n",
                "    title=None,\n",
                "    #ylim=(-100, 100),\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlim=(0, n_days),\n",
                "    ylabel=\"SSIM [dB]\",\n",
                ")\n",
                "\n",
                "plt.legend(title=None)\n",
                "plt.savefig(plotdir / \"ssim.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 24,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x300.379 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "since_bugfix = ssim_over_time.query(\n",
                "    \"(day >= @deployment_start) & (abr in @abr_order_short)\"\n",
                "    #\" & (day not in @worst_days)\"\n",
                "    \" & (varmetric in ['Mean SSIM'])\"\n",
                ").dropna()\n",
                "\n",
                "n_days = (ssim_over_time.day.max() - ssim_over_time.day.min()).days\n",
                "tickfreq = 7 * 3  # every 3 weeks\n",
                "xticks = np.arange(0, n_days + 1, tickfreq)\n",
                "xticklabels = np.arange(len(xticks)) * 3\n",
                "\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=ssim_over_time,\n",
                "    hue=\"abr\", hue_order=abr_order_short, palette=abr_palette,\n",
                "    **eh.plot_utils.opts_short(is_grid=True)\n",
                ").map_dataframe(\n",
                "    plot_timeseries,\n",
                "    window=14,\n",
                "    bootstrap_kwargs=dict(n_resamples=1000, difference=None),\n",
                "    xdate=False,\n",
                ").set(\n",
                "    title=None,\n",
                "    #ylim=(0, 0.4),\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlim=(0, n_days),\n",
                "    ylabel=\"SSIM [dB]\",\n",
                ")\n",
                "plt.legend()\n",
                "\n",
                "plt.savefig(plotdir / \"ssim_smoothed.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "### Figure 18b and 19b: Stalls over time; raw (14) and smoothed (15)\n",
                "\n",
                "For Memento and Memento (T=0) in the current deployment.\n",
                "For the smoothed case, again with 14-day rolling window and CI."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 25,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x300.379 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "stalls_over_time = comparison.query(\n",
                "    \"(day >= @deployment_start) & (abr in @abr_order_short)\"\n",
                "    #\" & (day not in @worst_days)\"\n",
                "    \" & (varmetric in ['Percent stalled'])\"\n",
                ").dropna()\n",
                "\n",
                "n_days = (stalls_over_time.day.max() - stalls_over_time.day.min()).days\n",
                "tickfreq = 7 * 3  # every 3 weeks\n",
                "xticks = np.arange(0, n_days + 1, tickfreq)\n",
                "xticklabels = np.arange(len(xticks)) * 3\n",
                "\n",
                "\n",
                "grid = sns.relplot(\n",
                "    kind='line',\n",
                "    data=stalls_over_time,\n",
                "    x=\"days_running\", y=\"value\",\n",
                "    hue=\"abr\", hue_order=abr_order_short, palette=abr_palette,\n",
                "    **eh.plot_utils.opts_short()\n",
                ").set(\n",
                "    title=None,\n",
                "    #ylim=(-100, 100),\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlim=(0, n_days),\n",
                "    ylabel=\"Stalled [\\%]\",\n",
                ")\n",
                "\n",
                "plt.legend(title=None, loc=\"upper center\")\n",
                "\n",
                "plt.savefig(plotdir / \"stalls.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 26,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x300.379 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "stalls_over_time = comparison.query(\n",
                "    \"(day >= @deployment_start) & (abr in @abr_order_short)\"\n",
                "    #\" & (day not in @worst_days)\"\n",
                "    \" & (varmetric in ['Percent stalled'])\"\n",
                ").dropna()\n",
                "\n",
                "n_days = (stalls_over_time.day.max() - stalls_over_time.day.min()).days\n",
                "tickfreq = 7 * 3  # every 3 weeks\n",
                "xticks = np.arange(0, n_days + 1, tickfreq)\n",
                "xticklabels = np.arange(len(xticks)) * 3\n",
                "\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=stalls_over_time,\n",
                "    hue=\"abr\", hue_order=abr_order_short, palette=abr_palette,\n",
                "    **eh.plot_utils.opts_short(is_grid=True)\n",
                ").map_dataframe(\n",
                "    plot_timeseries,\n",
                "    window=14,\n",
                "    bootstrap_kwargs=dict(n_resamples=1000, difference=None),\n",
                "    xdate=False,\n",
                ").set(\n",
                "    title=None,\n",
                "    ylim=(0, 0.6),\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlim=(0, n_days),\n",
                "    ylabel=\"Stalled [\\%]\",\n",
                ")\n",
                "plt.legend(loc=\"upper center\")\n",
                "\n",
                "plt.savefig(plotdir / \"stalls_smoothed.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "### Figure 13: Performance drop in previous deployment\n",
                "\n",
                "We found that settings T=0, i.e. not rejecting noise, can lead to significant degradation over time.\n",
                "\n",
                "This is noticeable primarily in a drop in image quality."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 27,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 637.2x300.379 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "last_dep = comparison.query(\n",
                "    \"(day >= @old_deployment_start) & (day <= @deployment_start)\"\n",
                "    \" & (abr in @abr_order_short)\"\n",
                "    #\" & (day not in @worst_days)\"\n",
                "    \" & (varmetric in ['Mean SSIM'])\"\n",
                ").dropna()\n",
                "\n",
                "n_days = (last_dep.day.max() - last_dep.day.min()).days\n",
                "tickfreq = 7 * 3  # every 3 weeks\n",
                "xticks = np.arange(0, n_days + 1, tickfreq)\n",
                "xticklabels = np.arange(len(xticks)) * 3\n",
                "\n",
                "\n",
                "grid = sns.FacetGrid(\n",
                "    data=last_dep,\n",
                "    hue=\"abr\", hue_order=abr_order_short, palette=abr_palette,\n",
                "    **eh.plot_utils.opts_short(is_grid=True)\n",
                ").map_dataframe(\n",
                "    plot_timeseries,\n",
                "    window=1,\n",
                "    bootstrap_kwargs=dict(n_resamples=1000, difference=None),\n",
                "    xdate=False,\n",
                ").set(\n",
                "    title=None,\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlim=(0, n_days),\n",
                "    ylabel=\"SSIM [dB]\",\n",
                ")\n",
                "plt.legend()\n",
                "\n",
                "plt.savefig(plotdir / \"failure_deterministic.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# 2D aggregate plots\n",
                "\n",
                "Plots with the same style as the Puffer paper, with Stalls on the x and SSIM on the y axis.\n",
                "\n",
                "To help compare the new results and put them in context,\n",
                "we add specific datapoints for Fugu (daily) and Fugu (static) for the days\n",
                "where both were deployed at the same time. This helps to highlight the differnece between random and density-based daily retraining.\n",
                "\n",
                "But first we need some helpers to actually plot this."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 28,
            "metadata": {},
            "outputs": [],
            "source": [
                "# We need a wide-form dataframe for these plots.\n",
                "comparison_wide = (\n",
                "    comparison\n",
                "    .pivot(\n",
                "        index=[\"abr\", \"day\", \"selection\"], columns=\"varmetric\", values=\"value\"\n",
                "    )\n",
                ").reset_index()\n",
                "\n",
                "both_fugus = (  # Days for with we have both Fugu and FuguFeb.\n",
                "    set(comparison.query(\"abr == @fugu_name\")['day']).intersection(\n",
                "        comparison.query(\"abr == @fugufeb_name\")['day']\n",
                "    )\n",
                ")\n",
                "\n",
                "comparison_with_past = pd.concat([\n",
                "    (\n",
                "        comparison_wide\n",
                "        .query(\"(day > @deployment_start) and (abr != @fugu_name)\")\n",
                "    ),\n",
                "    (\n",
                "        comparison_wide\n",
                "        .query(\"(day in @both_fugus) and (abr in [@fugu_name, @fugufeb_name])\")\n",
                "        .assign(abr = lambda df: [f\"Past {_abr}\" for _abr in df['abr']])\n",
                "    )\n",
                "])\n",
                "\n",
                "abr_palette_extended = {\n",
                "    **abr_palette,\n",
                "    f\"Past {fugu_name}\": sns.color_palette('dark')[1],\n",
                "    f\"Past {fugufeb_name}\": sns.color_palette('dark')[0],\n",
                "}\n",
                "\n",
                "\n",
                "def plot_2d(*, data, x, y, w=None, ax=None,\n",
                "            n_resamples=1000, confidence_level=0.9, random_state=42,\n",
                "            **kwargs):\n",
                "    if ax is None:\n",
                "        ax = plt.gca()\n",
                "\n",
                "    # Tuples for compatibility with bootstrap.\n",
                "    _x = (data[x].values,)\n",
                "    _y =( data[y].values,)\n",
                "    if w is not None:\n",
                "        _w = data[w].values\n",
                "        _x = (*_x, _w)\n",
                "        _y = (*_y, _w)\n",
                "\n",
                "    _x_est = weighted_mean_raw(*_x)\n",
                "    _x_ci = stats.bootstrap(\n",
                "        _x, weighted_mean_raw, method='percentile',\n",
                "        n_resamples=n_resamples,\n",
                "        confidence_level=confidence_level,\n",
                "        random_state=random_state\n",
                "    ).confidence_interval\n",
                "    _y_est = weighted_mean_raw(*_y)\n",
                "    _y_ci = stats.bootstrap(\n",
                "        _y, weighted_mean_raw, method='percentile',\n",
                "        n_resamples=n_resamples,\n",
                "        confidence_level=confidence_level,\n",
                "        random_state=random_state\n",
                "    ).confidence_interval\n",
                "\n",
                "    # A bit weird, the error bars must be relative, absolute values.\n",
                "    # Also they need shape (2, n) for the errorbar function.\n",
                "    _xerr = np.abs(np.array(_x_ci) - _x_est).reshape(2, 1)\n",
                "    _yerr = np.abs(np.array(_y_ci) - _y_est).reshape(2, 1)\n",
                "\n",
                "    ax.errorbar(_x_est, _y_est, xerr=_xerr, yerr=_yerr, **kwargs)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "### Figure 17a: 2D performance"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 29,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 400x400 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "grid = sns.FacetGrid(\n",
                "    data=comparison_with_past,\n",
                "    hue=\"abr\", palette=abr_palette_extended,\n",
                "    despine=False,\n",
                "    height=4,\n",
                "    legend_out=False,\n",
                ").map_dataframe(\n",
                "    plot_2d,\n",
                "    x=\"Percent stalled\", y=\"Mean SSIM\", w=\"Time streamed\",\n",
                ").add_legend(title=\"\", loc=\"lower center\").set(\n",
                "    xlim=(0.30, 0.0),\n",
                "    ylim=(0, 17.5),\n",
                ")\n",
                "\n",
                "ax = plt.gca()\n",
                "ax.text(\n",
                "    0.89, 0.89, \"better\", transform=ax.transAxes,\n",
                "    ha=\"center\", va=\"center\", rotation=45, size=15,\n",
                "    bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1))\n",
                "\n",
                "plt.savefig(plotdir / \"aggregated_2d.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "### Figure 17b: 2D performance, zoomed in."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 30,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 400x400 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "grid = sns.FacetGrid(\n",
                "    data=comparison_with_past,\n",
                "    hue=\"abr\", palette=abr_palette_extended,\n",
                "    despine=False,\n",
                "    height=4,\n",
                "    legend_out=False,\n",
                ").map_dataframe(\n",
                "    plot_2d,\n",
                "    x=\"Percent stalled\", y=\"Mean SSIM\", w=\"Time streamed\",\n",
                ").add_legend(title=\"\", loc=\"lower center\").set(\n",
                "    xlim=(0.3, 0.15),\n",
                "    ylim=(14, 17.5),\n",
                ")\n",
                "\n",
                "ax = plt.gca()\n",
                "ax.text(\n",
                "    0.89, 0.89, \"better\", transform=ax.transAxes,\n",
                "    ha=\"center\", va=\"center\", rotation=45, size=15,\n",
                "    bbox=dict(boxstyle=\"rarrow,pad=0.2\", fc=\"lightgray\", ec=\"gray\", lw=1))\n",
                "\n",
                "\n",
                "plt.savefig(plotdir / \"aggregated_2d_zoom.pdf\")\n",
                "plt.show()"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "# Density Increase and Retraining\n",
                "\n",
                "Finally, we investigate how often Memento retrained in deployment, i.e. how often the threshold for the relative coverage increase was reached."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 31,
            "metadata": {},
            "outputs": [],
            "source": [
                "all_stats = []\n",
                "for deploykey, deploydir in config.memento_model_directories.items():\n",
                "    if not Path(deploydir).is_absolute():\n",
                "        # If it's relative, it's relative to the base dir, not to `notebooks/`.`\n",
                "        deploydir = f'../{deploydir}'\n",
                "\n",
                "    for model_index in range(5):\n",
                "        modeldir = Path(deploydir) / str(model_index)\n",
                "        for daydir in modeldir.iterdir():\n",
                "            try:\n",
                "                day = pd.to_datetime(daydir.name)\n",
                "            except ValueError:\n",
                "                continue # logs, checkpoint\n",
                "\n",
                "            try:\n",
                "                _stats = eh.data.read_json(\n",
                "                    daydir / f\"stats_{model_index}.json.gz\")\n",
                "            except FileNotFoundError:\n",
                "                try:\n",
                "                    _stats = eh.data.read_pickle(\n",
                "                        daydir / f\"stats_{model_index}.pickle.gz\")\n",
                "                except FileNotFoundError:\n",
                "                    continue\n",
                "            _stats.update(model=model_index, day=day, abr=deploykey)\n",
                "            all_stats.append(_stats)\n",
                "\n",
                "DAY = pd.Timedelta(days=1)\n",
                "statframe = (\n",
                "    pd\n",
                "    .DataFrame\n",
                "    .from_records(all_stats)\n",
                "    # Add iteration counter\n",
                "    .assign(\n",
                "        iteration=lambda df: (\n",
                "        ((df['day'] - df['day'].min()) / DAY ).astype(int) + 1),\n",
                "        # For each frame, apply name mapping and use a categorical\n",
                "        # datatype to ensure consistent ordering and to save space.\n",
                "        abr=lambda df: pd.Categorical(\n",
                "            df['abr'].replace(name_mapping), abr_order),\n",
                "    )\n",
                "    .sort_values(by='day')\n",
                ")"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "Retraining days for Memento."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 32,
            "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>abr</th>\n",
                            "      <th>day</th>\n",
                            "      <th>day_since_start</th>\n",
                            "    </tr>\n",
                            "  </thead>\n",
                            "  <tbody>\n",
                            "    <tr>\n",
                            "      <th>524</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>2022-09-20</td>\n",
                            "      <td>1</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>157</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>2022-09-22</td>\n",
                            "      <td>3</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>181</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>2022-09-26</td>\n",
                            "      <td>7</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>605</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>2022-09-27</td>\n",
                            "      <td>8</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>480</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>2023-01-13</td>\n",
                            "      <td>116</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>402</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>2023-04-22</td>\n",
                            "      <td>215</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>227</th>\n",
                            "      <td>\\textsf{Memento}</td>\n",
                            "      <td>2023-04-27</td>\n",
                            "      <td>220</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3286</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-09-20</td>\n",
                            "      <td>1</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3563</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-09-21</td>\n",
                            "      <td>2</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3756</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-09-25</td>\n",
                            "      <td>6</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3571</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-09-26</td>\n",
                            "      <td>7</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3226</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-09-27</td>\n",
                            "      <td>8</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3451</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-10-03</td>\n",
                            "      <td>14</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3196</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-10-14</td>\n",
                            "      <td>25</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3456</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-11-09</td>\n",
                            "      <td>51</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3414</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-11-10</td>\n",
                            "      <td>52</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3266</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2022-11-13</td>\n",
                            "      <td>55</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3642</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2023-01-02</td>\n",
                            "      <td>105</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3391</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2023-01-13</td>\n",
                            "      <td>116</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3534</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2023-01-16</td>\n",
                            "      <td>119</td>\n",
                            "    </tr>\n",
                            "    <tr>\n",
                            "      <th>3259</th>\n",
                            "      <td>\\textsf{Memento} (T\\texttt{=}0)</td>\n",
                            "      <td>2023-01-17</td>\n",
                            "      <td>120</td>\n",
                            "    </tr>\n",
                            "  </tbody>\n",
                            "</table>\n",
                            "</div>"
                        ],
                        "text/plain": [
                            "                                  abr        day  day_since_start\n",
                            "524                  \\textsf{Memento} 2022-09-20                1\n",
                            "157                  \\textsf{Memento} 2022-09-22                3\n",
                            "181                  \\textsf{Memento} 2022-09-26                7\n",
                            "605                  \\textsf{Memento} 2022-09-27                8\n",
                            "480                  \\textsf{Memento} 2023-01-13              116\n",
                            "402                  \\textsf{Memento} 2023-04-22              215\n",
                            "227                  \\textsf{Memento} 2023-04-27              220\n",
                            "3286  \\textsf{Memento} (T\\texttt{=}0) 2022-09-20                1\n",
                            "3563  \\textsf{Memento} (T\\texttt{=}0) 2022-09-21                2\n",
                            "3756  \\textsf{Memento} (T\\texttt{=}0) 2022-09-25                6\n",
                            "3571  \\textsf{Memento} (T\\texttt{=}0) 2022-09-26                7\n",
                            "3226  \\textsf{Memento} (T\\texttt{=}0) 2022-09-27                8\n",
                            "3451  \\textsf{Memento} (T\\texttt{=}0) 2022-10-03               14\n",
                            "3196  \\textsf{Memento} (T\\texttt{=}0) 2022-10-14               25\n",
                            "3456  \\textsf{Memento} (T\\texttt{=}0) 2022-11-09               51\n",
                            "3414  \\textsf{Memento} (T\\texttt{=}0) 2022-11-10               52\n",
                            "3266  \\textsf{Memento} (T\\texttt{=}0) 2022-11-13               55\n",
                            "3642  \\textsf{Memento} (T\\texttt{=}0) 2023-01-02              105\n",
                            "3391  \\textsf{Memento} (T\\texttt{=}0) 2023-01-13              116\n",
                            "3534  \\textsf{Memento} (T\\texttt{=}0) 2023-01-16              119\n",
                            "3259  \\textsf{Memento} (T\\texttt{=}0) 2023-01-17              120"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                },
                {
                    "data": {
                        "text/plain": [
                            "abr\n",
                            "\\textsf{Memento}                    7\n",
                            "\\textsf{Memento} (T\\texttt{=}0)    14\n",
                            "\\textsf{Fugu}                       0\n",
                            "\\textsf{Fugu} (static)              0\n",
                            "Name: retrain, dtype: int64"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "training_current_deployment = (\n",
                "    statframe.query(\n",
                "        \"(model==0) and (abr in [@memento_name, @memento_deterministic])\"\n",
                "        \" and (day >= @deployment_start) and (day < @cutoff_day)\"\n",
                "    )\n",
                "    .assign(day_since_start=lambda df: (df['day'] - df['day'].min()).dt.days)\n",
                ")\n",
                "\n",
                "# After bugfix:\n",
                "display(\n",
                "    training_current_deployment\n",
                "    .loc[lambda df: df['retrain'], [\"abr\", 'day', 'day_since_start']]\n",
                "    .sort_values(by=[\"abr\", 'day_since_start'])\n",
                ")\n",
                "\n",
                "display(\n",
                "    training_current_deployment\n",
                "    .groupby(\"abr\")\n",
                "    ['retrain']\n",
                "    .sum()\n",
                ")"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Figure 6 (b): Retraining"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": 33,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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",
                        "text/plain": [
                            "<Figure size 360.455x360.455 with 1 Axes>"
                        ]
                    },
                    "metadata": {},
                    "output_type": "display_data"
                }
            ],
            "source": [
                "n_days = training_current_deployment['day_since_start'].max()\n",
                "tickfreq = 6   # every n weeks\n",
                "xticks = np.array(range(0, n_days, 7 * tickfreq))\n",
                "xticklabels = np.arange(len(xticks)) * tickfreq\n",
                "\n",
                "# For simplicity, show only model 0. The other have the same pattern\n",
                "# (uncomment to verify)\n",
                "xrange = training_current_deployment['day_since_start'].agg(['min', 'max'])\n",
                "grid = sns.relplot(\n",
                "    kind='line',\n",
                "    data=training_current_deployment,\n",
                "    x='day_since_start',\n",
                "    y='coverage_increase',\n",
                "    hue=\"abr\", hue_order=[memento_name, memento_deterministic],\n",
                "    palette=abr_palette,\n",
                "    style=\"abr\", style_order=[memento_name, memento_deterministic],\n",
                "    **eh.plot_utils.opts_square_small()\n",
                ").set(\n",
                "    xlim=xrange,\n",
                "    xticks=xticks[1:],\n",
                "    xticklabels=xticklabels[1:],\n",
                "    xlabel=\"Weeks in Deployment\",\n",
                "    ylim=(0, 0.4),\n",
                "    yticks=[0, 0.1, 0.2, 0.3],\n",
                "    ylabel=\"Relative Coverage Increase\"\n",
                ")\n",
                "\n",
                "ax = plt.gca()\n",
                "\n",
                "# On top, put markers for retraining events.\n",
                "sns.scatterplot(\n",
                "    ax=ax,\n",
                "    data=training_current_deployment.query(\"retrain\"),\n",
                "    x='day_since_start',\n",
                "    y='coverage_increase',\n",
                "    hue=\"abr\", hue_order=[memento_name, memento_deterministic],\n",
                "    style=\"abr\", style_order=[memento_name, memento_deterministic],\n",
                "    size=10, edgecolor='k', zorder=42,\n",
                "    palette=abr_palette,\n",
                ")\n",
                "\n",
                "\n",
                "# Manual adjustments\n",
                "h, l = ax.get_legend_handles_labels()\n",
                "ax.legend(\n",
                "    title=None,\n",
                "    handles=[h[0], h[1], (h[2], h[3])],\n",
                "    labels=[\"Memento\", \"Memento (T=0)\", \"Retraining\"],\n",
                "    # Using this map we can put both retraining symbols next to each other.\n",
                "    handler_map={tuple: HandlerTuple(ndivide=None)}\n",
                ")\n",
                "\n",
                "# Force Memento (first line plotted) to be on top.\n",
                "ax.get_children()[0].set_zorder(10)\n",
                "\n",
                "# Add line for threshold\n",
                "ax.plot(xrange, [0.1, 0.1], 'k--')\n",
                "ax.annotate(\"Retraining Threshold\", xy=[xrange.max() * 0.99, 0.1],\n",
                "            xytext=(0, 30), textcoords='offset points',\n",
                "            ha=\"right\", va=\"bottom\",\n",
                "            arrowprops=dict(facecolor='black', width=2, headwidth=8,))\n",
                "\n",
                "plt.tight_layout()\n",
                "plt.savefig(plotdir / \"memento_coverage_increase.pdf\")\n",
                "plt.show()"
            ]
        }
    ],
    "metadata": {
        "interpreter": {
            "hash": "f78f8e3aabf3906a5ed8d5dc64bda1959ee71737771fe5fec1226ba5e4787360"
        },
        "kernelspec": {
            "display_name": "Python 3.9.7 64-bit ('env': venv)",
            "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": "3.10.11"
        },
        "orig_nbformat": 4
    },
    "nbformat": 4,
    "nbformat_minor": 2
}