knowledge/How_Fast_Would_You_Drive_In_Monaco.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How fast would you drive in Monaco?\n",
    "\n",
    "**Pietro Grandinetti -- January 21, 2021**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n",
    "In this article I am going to try out statistical inference techniques with the goal to draw insights from a pretty large dataset.\n",
    "\n",
    "I will touch on several subjects, from Python and PostgreSQL, to data sampling, hypothesis testing and inference, while driven by a business point of view. In other terms, I will put myself at the place of the Data Scientist who's been given the task to provide its business with data-driven, actionable insights.\n",
    "\n",
    "The content will thus be interesting for those of you who work on any of the following topics:\n",
    "- Python development\n",
    "- Database programming and optimization\n",
    "- Inferential Statistics\n",
    "- Data Science"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The data (and the story behind it)\n",
    "\n",
    "Raw data is the fuel behind the most important innovations of these days. It's a pity that many organizations keep their data secret. On the one hand, I understand their reasons; on the other, history has shown that open-source contributions drive innovation like, or even more than, competition.\n",
    "\n",
    "That's why we set on a mission to publish everything we do and make it reproducible. One of the first things we did was to take a complicated machinery developed by researchers at the Communication Systems Department\n",
    "of Sophia-Antipolis, France, that was already open-sourced and we made it more easily accessible and 100% reproducible.\n",
    "\n",
    "The result is... _a lot_ of data.\n",
    "\n",
    "This is a system that simulates road traffic in the Principality of Monaco for ten hours, starting from 4am. It basically is a fairly large network of roads (think of a graph) inside which vehicles of any type and pedestrians  move (think of objects moving from node to node of a graph).\n",
    "\n",
    "The key is that every vehicle is equipped with a GPS-like sensor. Therefore, at the end of the simulation, we can retrieve instantaneous positions and speed information for each vehicle. That sounds like a great dataset!\n",
    "\n",
    "We've done this already, put all data in CSV format as well as loaded them into a PostgreSQL database. Then (suprise!), we simply published the data on the internet. You can download all of them on your computer without even asking permission. And you can reproduce the experiment (or run your own) in a couple of clicks. Careful though, it's many gigabytes of data-- may take a while. I would recommend to first read through [this article](https://github.com/pgrandinetti/standard-traffic-data/blob/main/knowledge/Urban_Traffic_Data_Exploratory_Analysis.ipynb) that shows some interesting exploratory data analysis done over a subset of the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Settings\n",
    "\n",
    "The dataset is definitely great, though it presents an obvious obstacle to start with: it's a bit too large!\n",
    "\n",
    "In fact, I am not even sure how many records are there. For sure, it's fixed size, therefore I could theoretically download the CSV, wait for probably a few hours, then run a `wc -l` to know its size, wait several minutes and then I'd get the number. And then... what?\n",
    "\n",
    "I suspect the dataset contains about 100 million rows, but this information is useless. I also know it's a CSV of around 8 GB (I can see the file size in the browser when it asks me to confirm the download), but this information is not very useful either.\n",
    "\n",
    "I was not given a cluster of computers with a lot of memory. I am using my laptop with 8G of fast memory. Even if I download the entire file, how am I supposed to load it into memory?\n",
    "\n",
    "The same problem applies to the PostgreSQL database. For sure the database has a lot of advantages over the CSV file (and the unique disadvantage that you need to know a bit of SQL!), but it comes with problems too. First of all, to maintain it costs money. I was forbidden by my team to run a `select count(*)` which would consume a lot of memory and maybe even take down the CPU of the server that hosts the DB. In a real-world production environment, it may very well be that the system administration has disabled queries that take too much DB memory. Imagine if I were to take down the entire server!\n",
    "\n",
    "No reason to worry too much though. This is a fairly common situation for data scientists. Whether you were given 1 billion tweets, 500 million pictures, or 100 million payment transactions, you won't be able to analyze the dataset in its entirety.\n",
    "\n",
    "You and I need a hat."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Choose the right hat\n",
    "\n",
    "I need to wear the statistician hat to work on this task.\n",
    "\n",
    "The database will be the _population_ that I have to study. I will have to come up with _hypothesis_ about this population, driven by _samples_ and _reject_ (or not) them via statistical _testing_ and _evidence_.\n",
    "\n",
    "Statisticians never assume to know the entire population. Think about healthcare studies: when a company wants to evaluate the effectiveness of a new drug, they certainly don't assume to know how the entire world's population would react to it. They take a sample (volunteers usually), test the drug on this small sample and then make conclusions based on statistical evidence.\n",
    "\n",
    "This is the correct Data Science approach for large dataset, and it's the one I'll use."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**My approach: ask questions.**\n",
    "\n",
    "My approach to statistical analysis is to ask questions. These questions will drive sampling and light exploration, and then shape the hypotheses. Once I have some meaningful hypotheses, I will work on them statistically (via confidence intervals and/or hypothesis testing)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preliminaries\n",
    "\n",
    "Once for all, let me install and load all packages I need."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install pandas\n",
    "!pip install psycopg2-binary\n",
    "!pip install numpy\n",
    "!pip install scipy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import time\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy.stats as st\n",
    "import psycopg2\n",
    "\n",
    "import concurrent.futures"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I follow the best practice to have credentials and other environment variables in a `.env` file. Before I run this notebook, I do `$source .env` which is why now I can find them into `os.environ`. That's how I am going to initialize the database connection."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def db_conn():\n",
    "    return psycopg2.connect(\n",
    "        host=os.environ['DB_HOST'],\n",
    "        port=os.environ['DB_PORT'],\n",
    "        password=os.environ['DB_PASSWORD'],\n",
    "        user=os.environ['DB_USER'],\n",
    "        dbname=os.environ['DB_NAME']\n",
    "    )\n",
    "connection = db_conn()\n",
    "assert connection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That's it for the preliminaries! Let me know go ahead and talk about the cornerstone of this article, _the sampling procedure_."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sampling from the population"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_Sampling_ literally means to extract a few elements from the population. The most famous example of it is probably the [Urn problem](https://en.wikipedia.org/wiki/Urn_problem).\n",
    "\n",
    "With respect to urns and health studies I am in a advantageous position. Here the population sits in a database therefore I can take as many samples as I want. There are however a few gotchas to keep in mind:\n",
    "\n",
    "- I shouldn't rely on the fact that \"it's a database so I can look at the whole\". In fact, I cannot. Not only because the size is very large, but rather because it's unknown. Somebody may put more data in there, at any time.\n",
    "- I have be careful on the load I put on the database CPU. Hundreds, or low thousands of records should be fine.\n",
    "- The samples have to be taken as randomly as possible. As you know, it's not easy to have real randomness in computer systems.\n",
    "\n",
    "Luckily, PostgreSQL is here for the rescue.\n",
    "\n",
    "From version 9.5, PostgreSQL supports an operator called [TABLESAMPLE](https://www.postgresql.org/docs/9.5/sql-select.html). Like the name suggests, this allows to take a sample of a table.\n",
    "\n",
    "Few, very important things to keep in mind:\n",
    "\n",
    "- `TABLESAMPLE` can be applied either with `SYSTEM` or with `BERNOULLI` as sampling method. The first one is _very_ fast, because it's based on blocks/pages, but (and for the same reason) it's not really random. The latter is slower, but offer better randomization, and is the one I will use.\n",
    "- It requires a real number between 0 and 100, that represents the percentage of the table to sample. However, the result will not be the _exact_ percentage of rows, just a close number.\n",
    "- `TABLESAMPLE` is applied _before_ the `WHERE` clause. Hence, if I want to have 100 rows all respecting a certain condition, I have to sample more to try to get close to 100 after the `WHERE` is applied.\n",
    "\n",
    "This is a very brief overview and you should definitely read up the documentation for more details. There are some takeways for the experiments I will be working on.\n",
    "\n",
    "To start with, `TABLESAMPLE` can be _seeded_. This is common practice in numerical simulation, to initialize all pseudo-random algorithms with a seed, so that the experiment is reproducible. However, in this case I think it's unnecessary -- it doesn't fit well the inferential set-up, in my opinion. I will not seed the samples, hence keep in mind that when you run the notebook the results may vary slightly.\n",
    "\n",
    "The second point I want to make is about the percentage used to select the sample. I want to underline that the percentage doesn't offer guarantee on the total number of records in the database, nor it does about the number of rows I will get in output! PostgreSQL _tries_ to select that percentage of rows, but for efficiency reasons it may actually be a different value. In other words, if I use 0.1 and I get back 1000 records, it doesn't mean that there are 10,000 records in total. In fact, if I run the same command again I may get a slightly different number."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before continuing, let's see `TABLESAMPLE` in action with `BERNOULLI` sampling method. I will ask PostgreSQL to give me 0.001% of the population of vehicle IDs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 138 ms, sys: 22.6 ms, total: 160 ms\n",
      "Wall time: 2min 8s\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 82 entries, 0 to 81\n",
      "Data columns (total 2 columns):\n",
      " #   Column         Non-Null Count  Dtype  \n",
      "---  ------         --------------  -----  \n",
      " 0   timestep_time  82 non-null     float64\n",
      " 1   vehicle_id     69 non-null     object \n",
      "dtypes: float64(1), object(1)\n",
      "memory usage: 1.4+ KB\n"
     ]
    }
   ],
   "source": [
    "query = \"\"\"\n",
    "    SELECT timestep_time\n",
    "           , vehicle_id\n",
    "    FROM most_0400_1400_1_5\n",
    "    TABLESAMPLE BERNOULLI (0.0001)\n",
    "\"\"\"\n",
    "df = %time pd.read_sql(query, connection)\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I captured the execution time with the `%time` magic. As you can see, the _Wall time_, that is the time experienced by me waiting for output, is over 2 minutes. And this for fetching just ~80 records!\n",
    "\n",
    "Compare it with `TABLESAMPLE SYSTEM`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3.34 ms, sys: 11.9 ms, total: 15.3 ms\n",
      "Wall time: 187 ms\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 69 entries, 0 to 68\n",
      "Data columns (total 2 columns):\n",
      " #   Column         Non-Null Count  Dtype  \n",
      "---  ------         --------------  -----  \n",
      " 0   timestep_time  69 non-null     float64\n",
      " 1   vehicle_id     67 non-null     object \n",
      "dtypes: float64(1), object(1)\n",
      "memory usage: 1.2+ KB\n"
     ]
    }
   ],
   "source": [
    "query = \"\"\"\n",
    "    SELECT timestep_time\n",
    "           , vehicle_id\n",
    "    FROM most_0400_1400_1_5\n",
    "    TABLESAMPLE SYSTEM (0.0001)\n",
    "\"\"\"\n",
    "df = %time pd.read_sql(query, connection)\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this case I received the result after just 187 ms. As I said, however, `SYSTEM` fetches pages of data therefore providing not-so-great of randomization.\n",
    "\n",
    "Now, let's move on to running the experiments and collecting the results."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run and save the experimental data\n",
    "\n",
    "My idea is to run 10 sampling experiments on the population, while trying to get around 1,000 samples in each of them.\n",
    "\n",
    "Experiments are usually very expensive: you need to find volunteers, pay them, set up everything carefully because you might not get a second chance.\n",
    "I am definitely in a better position, because my experiments consist of queries sent to a database over the network. Still, I want to be careful and therefore will save the results in a file, so I can refer to that file later, without needing to query the database again.\n",
    "\n",
    "This can take awhile, as I've shown above with a single query. That's fine though. I still want to try to speed up things and therefore will execute the 10 experiments in a threads pool. Some care is required because I don't want to overstress the database. For what I know, maybe it only accepts one connection at the time!\n",
    "I think, however, that 10 threads won't be a problem and that, in the worst case, the database will queue the ten connections and respond sequentially. That's OK, I'll go grab some coffee in the meantime.\n",
    "\n",
    "The code below goes along this lines:\n",
    "\n",
    "- A thread pool runs the ten experiments.\n",
    "- Wait then collect the results in ten dataframes.\n",
    "- Append the experiment number to each dataframe in a new column.\n",
    "- Concatenate all dataframe in one, and save it onto a CSV file.\n",
    "\n",
    "There are two other points I want to make:\n",
    "\n",
    "First, there exists a [package](https://github.com/MagicStack/asyncpg) for async queries to PostgreSQL databases, which may be worth looking into. I haven't done it yet, as the efficiency of the data collection part is not very critical for this article.\n",
    "\n",
    "Second, on the same subject, the ThreadPoolExecutor may not be the best option for this task. Some test is needed, in my opinion. It's well known that ThreadPool is recommended to I/O-bound operations, and database queries surely fall into this bucket. However, to make things more complicated there's the `connection` object and the pandas `read_sql` function.\n",
    "\n",
    "I had initially created only one `connection` and passed it as argument of each thread, but this somehow made the whole execution sequential. This pattern is commonly used in Web Services (and API), where you definitely don't want to create a new connection for every piece of code that you run to handle an incoming request; it's better to create the connection once and then pass it around.\n",
    "\n",
    "This didn't work here, as if the `psycopg2` connection object queues the queries in the same connection internally; this is just my guess and I haven't verified it yet. Either way, I decided to make the threads so that each of them creates a new connection, but I also limited the number of workers in the pool to 5. If everything makes sense then the total time for running 10 experiments with 5 workers in a ThreadPool should be the same as running two single queries sequentially, which means ~4 minutes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def single_experiment():\n",
    "    query = \"\"\"\n",
    "        SELECT timestep_time\n",
    "               , vehicle_id\n",
    "               , vehicle_speed\n",
    "               , vehicle_x\n",
    "               , vehicle_y\n",
    "               , vehicle_z\n",
    "               , vehicle_type\n",
    "        FROM most_0400_1400_1_5        -- the largest table\n",
    "        TABLESAMPLE BERNOULLI (0.005)\n",
    "        WHERE vehicle_type is not NULL -- skip pedestrians\n",
    "        ;\n",
    "    \"\"\"\n",
    "    with db_conn() as conn:\n",
    "        df = pd.read_sql(query, conn)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 208 ms, sys: 62.5 ms, total: 270 ms\n",
      "Wall time: 4min 23s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "n_experiments = 10\n",
    "workers = 5\n",
    "with concurrent.futures.ThreadPoolExecutor(max_workers=workers) as executor:\n",
    "    futures = [executor.submit(single_experiment)\n",
    "               for _ in range(n_experiments)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "filep = '/home/pietro/statinf_10exp.csv'\n",
    "dframes = [f.result() for f in futures]\n",
    "for i, df in enumerate(dframes):\n",
    "    df['experiment'] = i + 1\n",
    "df_tot = pd.concat(dframes)\n",
    "df_tot.to_csv(filep, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of samples: 30608\n"
     ]
    }
   ],
   "source": [
    "print(f'Number of samples: {df_tot.shape[0]}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "OK. The sampled data is saved, so I can reference to it. It seems that I got much more data than I wanted: good.\n",
    "\n",
    "Now I can finally start the statistical analysis. For real!\n",
    "\n",
    "Here are a few functions that I will use to clean a bit some dataframe. If you are wondering how do I know that in advance... it's not in advance: I came back to this part of the notebook _after_ I was finished and polished the code and isolated some useful function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def step_2_time(step):\n",
    "    \"\"\"Converts timestamp in seconds from midnight to H:M:S string format.\"\"\"\n",
    "    return time.strftime('%H:%M:%S', time.gmtime(step))\n",
    "\n",
    "\n",
    "def drop_speed_below(df, label, value=1):\n",
    "    \"\"\"Drop rows from dataframe with valules in column `label` below `value`.\n",
    "    Returns a new dataframe.\"\"\"\n",
    "    rows = df[df[label] < value].index\n",
    "    return df.drop(rows)\n",
    "\n",
    "\n",
    "def get_experiment(df, experiment, col='experiment'):\n",
    "    \"\"\"Get all rows corresponding to a given experiment number.\n",
    "    Returns a new dataframe.\"\"\"\n",
    "    exp_i_rows = df[col] == experiment\n",
    "    return df.loc[exp_i_rows].copy()\n",
    "\n",
    "\n",
    "def convert_speed(df, label='vehicle_speed'):\n",
    "    \"\"\"Add two columns to the dataframe:\n",
    "        - `vehicle_speed [km/h]`\n",
    "        - `speed_int`\n",
    "       The `speed_int` is the speed value in km/h truncated to its integer part.\n",
    "       Modifies the same dataframe in place.\n",
    "    \"\"\"\n",
    "    df[label + ' [km/h]'] = df[label].apply(lambda x: x * 3.6)\n",
    "    df['speed_int'] = df[label + ' [km/h]'].apply(lambda x: int(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question: What is the average speed vehicles travel at?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The average speed of any vehicle in a city is standard, basic indicator of whether traffic is fluid or not. There are a lot of caveats to this statement, because the speed is heavily affected by the type of vehicle, the type of road and the time in the day. Despite these many variables, average speed is still a good indicator.\n",
    "\n",
    "The sampling method I used to get ten samples from the poulation is such that the samples don't have a pattern with respect to vehicle type, road type and timestamp. Thus, I can assume that the distribution of vehicles, roads and timestamps in their categories is a fair representation of the real (population's) distribution.\n",
    "\n",
    "Before anything else, I want to take a quick look at one of the experiments and the vehicles' speed values therein."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of samples in experiment 1: 3111\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>timestep_time</th>\n",
       "      <th>vehicle_id</th>\n",
       "      <th>vehicle_speed</th>\n",
       "      <th>vehicle_x</th>\n",
       "      <th>vehicle_y</th>\n",
       "      <th>vehicle_z</th>\n",
       "      <th>vehicle_type</th>\n",
       "      <th>experiment</th>\n",
       "      <th>vehicle_speed [km/h]</th>\n",
       "      <th>speed_int</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>43545.0</td>\n",
       "      <td>pedestrian_3-1_2290_tr</td>\n",
       "      <td>0.07</td>\n",
       "      <td>5386.15</td>\n",
       "      <td>2883.33</td>\n",
       "      <td>54.10</td>\n",
       "      <td>passenger1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.252</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>43545.0</td>\n",
       "      <td>pedestrian_3-1_3011_tr</td>\n",
       "      <td>0.02</td>\n",
       "      <td>4318.94</td>\n",
       "      <td>1494.53</td>\n",
       "      <td>44.16</td>\n",
       "      <td>passenger2b</td>\n",
       "      <td>1</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>43550.0</td>\n",
       "      <td>pedestrian_3-1_4234_tr</td>\n",
       "      <td>0.03</td>\n",
       "      <td>4102.45</td>\n",
       "      <td>1406.34</td>\n",
       "      <td>125.78</td>\n",
       "      <td>passenger2a</td>\n",
       "      <td>1</td>\n",
       "      <td>0.108</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>43550.0</td>\n",
       "      <td>pedestrian_GW1-1_202_tr</td>\n",
       "      <td>0.00</td>\n",
       "      <td>4721.84</td>\n",
       "      <td>1522.93</td>\n",
       "      <td>8.25</td>\n",
       "      <td>passenger2b</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>43555.0</td>\n",
       "      <td>pedestrian_GW1-1_217_tr</td>\n",
       "      <td>0.04</td>\n",
       "      <td>4562.51</td>\n",
       "      <td>2204.12</td>\n",
       "      <td>73.47</td>\n",
       "      <td>passenger2a</td>\n",
       "      <td>1</td>\n",
       "      <td>0.144</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   timestep_time               vehicle_id  vehicle_speed  vehicle_x  \\\n",
       "0        43545.0   pedestrian_3-1_2290_tr           0.07    5386.15   \n",
       "1        43545.0   pedestrian_3-1_3011_tr           0.02    4318.94   \n",
       "2        43550.0   pedestrian_3-1_4234_tr           0.03    4102.45   \n",
       "3        43550.0  pedestrian_GW1-1_202_tr           0.00    4721.84   \n",
       "4        43555.0  pedestrian_GW1-1_217_tr           0.04    4562.51   \n",
       "\n",
       "   vehicle_y  vehicle_z vehicle_type  experiment  vehicle_speed [km/h]  \\\n",
       "0    2883.33      54.10   passenger1           1                 0.252   \n",
       "1    1494.53      44.16  passenger2b           1                 0.072   \n",
       "2    1406.34     125.78  passenger2a           1                 0.108   \n",
       "3    1522.93       8.25  passenger2b           1                 0.000   \n",
       "4    2204.12      73.47  passenger2a           1                 0.144   \n",
       "\n",
       "   speed_int  \n",
       "0          0  \n",
       "1          0  \n",
       "2          0  \n",
       "3          0  \n",
       "4          0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1 = get_experiment(df_tot, 1)\n",
    "print(f'Number of samples in experiment 1: {df1.shape[0]}')\n",
    "convert_speed(df1)\n",
    "df1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean speed value: 3.91 km/h\n",
      "Speed standard deviation: 13.16 km/h\n",
      "Mean speed value (only > 1 km/h values): 38.54 km/h\n",
      "Speed standard deviation (only > 1 km/h values): 20.6 km/h\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=1, ncols=3, figsize=(20, 5))\n",
    "ax0 = axes[0]\n",
    "ax1 = axes[1]\n",
    "ax2 = axes[2]\n",
    "speed_col = 'vehicle_speed [km/h]'\n",
    "time_col = 'timestep_time'\n",
    "time_ax_label = 'time [H:M:S]'\n",
    "fig1 = 'Complete scatter plot of vehicles speed values'\n",
    "fig2 = 'Boxplot of vehicle speed values (> 1 km/h only)'\n",
    "fig3 = 'Complete boxplot of vehicle speed values'\n",
    "\n",
    "xticks = [20e3, 25e3, 30e3, 35e3, 40e3, 45e3]\n",
    "df1.plot.scatter(y=speed_col,\n",
    "                 x=time_col,\n",
    "                 xticks=xticks,\n",
    "                 ax=ax0);\n",
    "ax0.set_xlabel(time_ax_label);\n",
    "ax0.set_xticklabels([step_2_time(tick) for tick in xticks]);\n",
    "ax0.set_title(fig1);\n",
    "\n",
    "df_tmp = drop_speed_below(df1, speed_col)\n",
    "df_tmp.boxplot(speed_col, ax=ax1);\n",
    "ax1.set_title(fig2);\n",
    "\n",
    "df1.boxplot(speed_col, ax=ax2);\n",
    "ax2.set_title(fig3);\n",
    "\n",
    "mu = np.round(np.mean(df1[speed_col]), 2)\n",
    "dev = np.round(np.std(df1[speed_col]), 2)\n",
    "print(f'Mean speed value: {mu} km/h')\n",
    "print(f'Speed standard deviation: {dev} km/h')\n",
    "mu_1 = np.round(np.mean(df_tmp[speed_col]), 2)\n",
    "dev_1 = np.round(np.std(df_tmp[speed_col]), 2)\n",
    "print(f'Mean speed value (only > 1 km/h values): {mu_1} km/h')\n",
    "print(f'Speed standard deviation (only > 1 km/h values): {dev_1} km/h')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This very, very brief exploratory analysis reveals a few things worth noting:\n",
    "\n",
    "**There are a lot of time instants where a lot of vehicles have speed equal to zero!** This actually makes sense: in a city, the stop-and-go dynamics is the most common. That being said, it's important to keep this in mind because zeros will inevitably drive down the overall average, while not being necessarily bad, or wrong things.\n",
    "\n",
    "With a timeseries dataset I could go deeper into this and find alternatives that take this problem into account. I could, for instance, consider the average for each vehicle computed over a few consecutive time steps; the idea is that if the average was low for many consecutive time instants, then it would mean something's not working well; if it was for one time instant only (or few of them) then it's a normal occurrence.\n",
    "\n",
    "Even if the data I have is _technically_ a timeseries, the sampling procedure creates a lot of \"holes\" in it. There are many techniques to fill holes in timeseries (they go under the name of \"data inputation\" techniques); none of them is perfect, everyone is worth trying. But the data I have it has so many missing timestamps (due to the sampling) that it looks more like a random walk inside the population, rather than a timeseries. Hence, I will not try any data inputation, and focus solely on statistical inference.\n",
    "\n",
    "**On the other hand, non-zero values do make sense**. When I briefly looked at this sample distribution with a boxplot by first removing speed values lower than 1 km/h, then I can tell that such distribution is reasonable. This is the plot in the center of the figure above. For an explanation of boxplot I recommend the [Wikipedia page](https://en.wikipedia.org/wiki/Box_plot).\n",
    "\n",
    "**When including zero values, nothing makes sense anymore**. I added a third plot (far right) that is a boxplot of the whole sample (hence, not removing any speed value). The very high number of zero values makes the distribution so skewed that it doesn't offer any meaningul insight (the box is collapsed somewhere near the zero). This is a repetition of the concept expressed in the first point above.\n",
    "\n",
    "**The standard deviation is big anyway**. No matter if I include zero-valued speed samples or not, the standard deviation turns out to be a pretty big number. The so-called _coefficient of variation_, computed by dividing the standard deviation over the mean value, is much higher than 1 --hence very large. I already know that this will make my confidence interval less _confident_, but at the same time it's a good descriptor of the population."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This brief analysis was done on just one of the ten samples that I got. By using it I found that the average vehicle speed is 3.91 km/h (or 38.5 km/h if speed values lower than 1 km/h were excluded) with a standard deviation of 13.16 km/h (or 20.6 km/h).\n",
    "\n",
    "This sample contains 3111 samples. It's a very small number, as I know that the overall database contains more than 80 million datapoints. However, according to theory, we know that the sample mean is an estimator of the population mean (the true mean), and the sample standard deviation is an estimator for the population std (the true std), by a factor equal to `sqrt(3111)`.\n",
    "\n",
    "I could then stop here and start making inferential reasoning with this data. However, thanks to the Central Limit Theorem, we also know that the mean of the sample means is the better estimator of the true population mean, as it converges to the center of the population.\n",
    "\n",
    "In other words: every time I collect a sample (\"sample\" in the general sense, it can be a sample of 3111 datapoints), this comes with a sample mean. **Such mean is also a realization of a random variable**, for the very simple reason that I can keep collecting samples and compute the sample mean for each of them.\n",
    "The collection of all these sample means is the distribution of the sample means. And (finally) the Central Limit Theorem says that a lot of samples will make the average of the sample means distribution equal to the true population average.\n",
    "\n",
    "This was a lot of words simply to say that is definitely worth looking at the other 9 extracted samples, and to their average values, to see what's the distribution of the ten sampled means."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The next thing I want to do now is to take a look at the individual samples' mean values (for the speed). I am just curious to see if they all have similar values or there are outliers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of samples in experiment 1: 3111\n",
      "Number of samples in experiment 2: 3063\n",
      "Number of samples in experiment 3: 3021\n",
      "Number of samples in experiment 4: 3069\n",
      "Number of samples in experiment 5: 3051\n",
      "Number of samples in experiment 6: 3093\n",
      "Number of samples in experiment 7: 3109\n",
      "Number of samples in experiment 8: 3044\n",
      "Number of samples in experiment 9: 3053\n",
      "Number of samples in experiment 10: 2994\n",
      "Sample means (10 samples): [3.91, 3.97, 4.01, 3.84, 3.47, 3.87, 4.4, 3.83, 4.01, 4.1]\n",
      "Sample means (10 samples, only > 1 km/h values): [38.54, 37.89, 38.16, 38.0, 36.75, 38.43, 39.45, 38.92, 40.83, 40.8]\n"
     ]
    }
   ],
   "source": [
    "all_mu = []\n",
    "all_std = []\n",
    "all_mu_1 = []\n",
    "all_std_1 = []\n",
    "new_col = 'vehicle_speed [km/h]'\n",
    "speed_col = 'vehicle_speed'\n",
    "for i in range(1, 11):\n",
    "    df_ = get_experiment(df_tot, i)\n",
    "    print(f'Number of samples in experiment {i}: {df_.shape[0]}')\n",
    "    convert_speed(df_, label=speed_col)\n",
    "    mu_ = np.round(np.mean(df_[new_col]), 2)\n",
    "    all_mu.append(mu_)\n",
    "    dev_ = np.round(np.std(df_[new_col]), 2)\n",
    "    all_std.append(dev_)\n",
    "    df_tmp_ = drop_speed_below(df_, new_col)\n",
    "    #df_tmp_ = df_.loc[df_[new_col] > 1]\n",
    "    mu_1_ = np.round(np.mean(df_tmp_[new_col]), 2)\n",
    "    all_mu_1.append(mu_1_)\n",
    "    dev_1_ = np.round(np.std(df_tmp_[new_col]), 2)\n",
    "    all_std_1.append(dev_1_)\n",
    "    \n",
    "print(f'Sample means (10 samples): {all_mu}')\n",
    "print(f'Sample means (10 samples, only > 1 km/h values): {all_mu_1}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The values are consistent each with the others. This means that the _adjustment_ each new sample does to the previous ones, to steer the overall mean towards the real population's mean, is not very big. Even then, it's worth in my opinion to take a look at how confidence increases with the increase of the samples.\n",
    "\n",
    "To look at this, I will plot the distribution of the first sample and draw a line at its sample average. Then I will merge the first two samples, plot the resulting distribution and look at its average. Then, the same with the first three samples; and so on until I have merged all of them.\n",
    "\n",
    "What I expect is a slight change to the average everytime I add one more sample, and these small changes are such that eventually the average will be very close to the true (unknown) population's average."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Confused? Read this!\n",
    "\n",
    "As I was reading this notebook again, I realized this could be a good place to pause for a minute and clarify a few things. I suspect you may have a question: _What the heck is this guy doing with all these \"samples\", \"populations\", \"averages\" and whatnot?_\n",
    "\n",
    "In fact, I am going through a very simple statistical result, one that everybody knows well. I will now draw the comparison between what I am doing here and the most famous statistical benchmark: flipping a 2-sided coin.\n",
    "\n",
    "The experiment with a coin is very simple: you flip the coin ten times and you count how many tails vs heads you got. Thus, you run one experiment and you get 7 heads vs 3 tails. Weird, but can happen.\n",
    "\n",
    "If you stopped there you would conclude that flipping a coin has 70% probability to give head. Or, if we assign 1 to head and 0 to tail, you would conclude that the average value of the coin is 0.7.\n",
    "\n",
    "But you know that's just one experiment.\n",
    "\n",
    "So you do it again: 6 tails, 4 heads. Average 0.4.\n",
    "\n",
    "And again: 3 tails, 7 heads. Average 0.7.\n",
    "\n",
    "And again: 2 heads, 8 tails. Average 0.2.\n",
    "\n",
    "And again. And again. And ... again.\n",
    "\n",
    "Every time you run one experiment and compute the average of that experiment, that's just the sample average. If you collect all such sample averages, then you have another sample: the sample of the averages of the experiments.\n",
    "\n",
    "The crucial point (stated by the Central Limit Theorem) is that, if you collect many sample averages, then **the mean value of all such averages is a very good estimate of the true average of a coin**. Which we know it's 0.5. Go ahead, take a coin and run a lot of experiments (you do need many of them to converge to 0.5) and check yourself.\n",
    "\n",
    "These are the ideas that I am applying here!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Now, back to the analysis.** What I am going to do next is a lot of barplots. A barplot can show quickly how the distribution of values look like. I will do one bar plot for the first sample only, another for the first and second sample merged, and so on.\n",
    "\n",
    "There's a caveat though. Speed values are real number. Barplot is a categorical plot, hence if I do it directly on the dataframe then it's like that I will have a lot of \"categories\" with one value only. For example, a vehicle's detected speed can be 20.23 km/h and that's a very specific value. It's likely that there's only one occurrence of such a value.\n",
    "\n",
    "To solve this problem I will just truncate the speed values to their integer part. Brutal, but I think it will be fine for this use case."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1800 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This will contain 10 subplots\n",
    "figure = plt.figure(figsize=(20, 25))\n",
    "\n",
    "# Start from the first experiment\n",
    "current_df = get_experiment(df_tot, 1)\n",
    "speed_col = 'vehicle_speed'\n",
    "speedkm_col = 'vehicle_speed [km/h]'\n",
    "new_col = 'speed_int'\n",
    "convert_speed(current_df, label=speed_col)\n",
    "current_df = drop_speed_below(current_df, new_col)\n",
    "\n",
    "all_avg = [np.round(np.mean(current_df[new_col]), 2)]\n",
    "\n",
    "for i in range(1, 11):\n",
    "    if i > 1:\n",
    "        # If not first iteration\n",
    "        # then cumulate the next sample.\n",
    "        df_exp_i = get_experiment(df_tot, i)\n",
    "        convert_speed(df_exp_i, label=speed_col)\n",
    "        df_exp_i = drop_speed_below(df_exp_i, new_col)\n",
    "        all_avg.append(np.round(np.mean(df_exp_i[new_col]), 2))\n",
    "        current_df = pd.concat([current_df, df_exp_i])\n",
    "\n",
    "    # Plot the cumulated-so-far samples\n",
    "    current_ax = figure.add_subplot(4, 3, i)\n",
    "    toplot = pd.value_counts(current_df[new_col]).sort_index()\n",
    "    x = list(toplot.index)\n",
    "    y = list(toplot)\n",
    "    current_avg = np.round(np.mean(all_avg), 2)\n",
    "    if i > 1:\n",
    "        title = f'{i} samples'\n",
    "    else:\n",
    "        title = '1 sample'\n",
    "    title += f' (mean = {current_avg})'\n",
    "    current_ax.bar(x=x, height=y)\n",
    "    plt.title(title);\n",
    "    plt.xticks(list(range(0, np.max(x) + 1, int(np.max(x) / 10))));\n",
    "    plt.yticks(list(range(0, np.max(y) + 1, int(np.max(y) / 5))));\n",
    "    plt.axvline(x=current_avg, color='red');\n",
    "    plt.xlabel('speed [km/h]')\n",
    "    plt.ylabel('num. samples')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let me state the most important insight from the figure above. If I had a doubt about what distribution the vehicles' speed follows just by looking at the first subfigure, then by looking at the last one there's now doubt anymore: it's a gaussian with two peaks (bells).\n",
    "\n",
    "In fact, I find it very impressive that the shape of the distribution changes slightly from figure to figure, and **it slowly converges to the final shape**. Yes, distributions do converge.\n",
    "\n",
    "In simpler (and more intuitive) words, you may say that the number of outliers decreases as I keep cumulating more samples, and therefore when looking a the whole picture the shapes of the distribution becomes more and more clearly defined.\n",
    "\n",
    "I had already checked that the sample averages don't differ too much one from the others. These plots show the same, though from a different point of view: as I keep cumulating samples, the overall average doesn't change much. It changes slowly and slightly, like the entire distribution does.\n",
    "\n",
    "Thus, I can conclude that **the sample average distribution has converged to a value that is close to 38.3 km/h**. And therefore, that's a good estimate of the real vehicles' speed.\n",
    "\n",
    "That's just a number though. If I were asked \"how fast do vehicles drive in Monaco?\" I could answer \"They drive at 38.3 km/h, in average\". But with what certainty would I answer that?\n",
    "\n",
    "The graphs above clearly show that the sample average is not even aligned with the center of the most prominent peak; that's because the other peak steers the average value towards itself a bit. Thus, I would answer 38.3 km/h when in reality the most common value (the mode) is closer to 45 km/h. If you followed this article along, now you could type\n",
    "\n",
    "```\n",
    "st.mode(current_df['speed_int'])\n",
    "```\n",
    "\n",
    "and it would tell you the most frequently seen value of speed. It gave me 47 km/h, but remember that it may differ a bit from what you see, due to the \"realistic\" sampling procedure.\n",
    "\n",
    "The point I am making here is that responding to that question with just a value can be harmful: it doesn't really give good insights about the system under analysis.\n",
    "\n",
    "What I need is to come up with statistical evidence supporting the numerical values. Confidence intervals will do it nicely."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This type of computation in Python is very easy. I will start by computing the range of velocity values that has 95% probability to be the actual value of a sampled vehicle."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "95% confidence interval for speed values [km/h]: (37.60096829109406, 38.9857512331874)\n"
     ]
    }
   ],
   "source": [
    "obj = current_df['speed_int']\n",
    "conf = 0.95\n",
    "length = len(obj)\n",
    "scale = st.sem(obj)\n",
    "loc = np.mean(obj)\n",
    "intv = st.t.interval(conf, length, loc=loc, scale=scale)\n",
    "print(f'95% confidence interval for speed values [km/h]: {intv}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In other words, this means that there's a 95% probability that the actual average value of a vehicle is between 37.6 and 38.9 km/h.\n",
    "\n",
    "Is that not high enough as certainty? Then, let's take the 99% confidence interval:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "99% confidence interval for speed values [km/h]: (37.3831930728994, 39.20352645138206)\n"
     ]
    }
   ],
   "source": [
    "conf = 0.99\n",
    "intv = st.t.interval(conf, length, loc=loc, scale=scale)\n",
    "print(f'99% confidence interval for speed values [km/h]: {intv}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you could expect, having a higher confidence implies having a broader range. In this case I could say that the average speed value of a vehicle is between 37.38 and 39.2 km/h with 99% probability.\n",
    "\n",
    "Is that too broad of a range? Let's say you need a more narrow interval, then let's compute the 85% confidence interval."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "85% confidence interval for speed values [km/h]: (37.78489448014104, 38.80182504414043)\n"
     ]
    }
   ],
   "source": [
    "conf = 0.85\n",
    "intv = st.t.interval(conf, length, loc=loc, scale=scale)\n",
    "print(f'85% confidence interval for speed values [km/h]: {intv}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The interval is narrow indeeed. It tells us that in 85% of cases, a vehicle will have a speed between 37.78 and 38.8 km/h. That's a widht of just 1 km/h and, not surprisingly, it includes the sample average that I computed by cumulating all samples (that was 38.29 km/h in the last of the ten figures)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's a summary of the three confidence intervals I computed.\n",
    "\n",
    "| Confidence | Lower bound [km/h] | Upper bound [km/h]|\n",
    "| ---        | ---                | ---               |\n",
    "| 0.99       | 37.38              | 39.2              |\n",
    "| 0.95       | 37.6               | 38.98             |\n",
    "| 0.85       | 37.78              | 38.8              |\n",
    "\n",
    "There's no right or wrong value of confidence interval. It all depends on your application and what insights you are trying to generate to provide value as a Data Scientist. It's only based on this information that you can decide what confidence interval is most useful for you."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**What about the stop-and-go effect?**\n",
    "\n",
    "Right. I did remove brutally the speed values lower than 1 km/h before doing the inferential part of the analysis. What happens if I consider those values too?\n",
    "\n",
    "I expect the distribution will be so skewed towards the left part of axis (that is, the zero value) that it will not give any useful information other than... vehicles are really slow!\n",
    "\n",
    "Let's try. The code is the same as in the previous plot, I just need to remove the line that drops speed values below the threshold."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1800 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This will contain 10 subplots\n",
    "figure = plt.figure(figsize=(20, 25))\n",
    "\n",
    "# Start from the first experiment\n",
    "current_df = get_experiment(df_tot, 1)\n",
    "speed_col = 'vehicle_speed'\n",
    "speedkm_col = 'vehicle_speed [km/h]'\n",
    "new_col = 'speed_int'\n",
    "convert_speed(current_df, label=speed_col)\n",
    "\n",
    "all_avg = [np.round(np.mean(current_df[new_col]), 2)]\n",
    "\n",
    "for i in range(1, 11):\n",
    "    if i > 1:\n",
    "        # If not first iteration\n",
    "        # then cumulate the next sample.\n",
    "        df_exp_i = get_experiment(df_tot, i)\n",
    "        convert_speed(df_exp_i, label=speed_col)\n",
    "        all_avg.append(np.round(np.mean(df_exp_i[new_col]), 2))\n",
    "        current_df = pd.concat([current_df, df_exp_i])\n",
    "\n",
    "    # Plot the cumulated-so-far samples\n",
    "    current_ax = figure.add_subplot(4, 3, i)\n",
    "    toplot = pd.value_counts(current_df[new_col]).sort_index()\n",
    "    x = list(toplot.index)\n",
    "    y = list(toplot)\n",
    "    current_avg = np.round(np.mean(all_avg), 2)\n",
    "    if i > 1:\n",
    "        title = f'{i} samples'\n",
    "    else:\n",
    "        title = '1 sample'\n",
    "    title += f' (mean = {current_avg})'\n",
    "    current_ax.bar(x=x, height=y)\n",
    "    plt.title(title);\n",
    "    plt.xticks(list(range(0, np.max(x) + 1, int(np.max(x) / 10))));\n",
    "    plt.yticks(list(range(0, np.max(y) + 1, int(np.max(y) / 5))));\n",
    "    plt.axvline(x=current_avg, color='red');\n",
    "    plt.xlabel('speed [km/h]')\n",
    "    plt.ylabel('num. samples')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ups. Yeah, not very useful. The occurrence of very low values is so high that I almost can't see the other speed values on the plots! This is definitely not helpful here, but it does raise an additional question: Do vehicles spend a lot of time at so low speed values? That's a question for another article though."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**The real power of this analysis**\n",
    "\n",
    "And so I've got my nice confidence intervals that I can discuss with my team and take decisions based on this information. Confidence interval are indeed one of the most useful statistics to drive business decisions.\n",
    "\n",
    "The real reason why I wanted to go through all of the steps above is more subtle though.\n",
    "\n",
    "What I explained so far is a standard approach to data analysis that can be reproduced for more questions as well as for different datasets:\n",
    "- Get the dataset.\n",
    "- Sample it with a randomized procedure.\n",
    "- Analyze the distribution of the samples.\n",
    "- Make inferential reasoning (statistical) based on the samples distribution.\n",
    "\n",
    "This reasoning model has helped me countless times. Now you can make it yours, if you wish.\n",
    "\n",
    "Let's see it applied to a different data-driven question (but with some twists!)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Question: What is the maximum capacity of the network?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I want to know, with a certain statistical confidence, what is the maximum number of vehicles travelling the network at the same time. In absence of more information, I will assume that is also the maximum capacity of the city (even though it may not really be packed up yet)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before getting into any confidence interval calculation and hypothesis testing, let's take a look at a portion of the data that I have, as in a brief exploratory analysis.\n",
    "\n",
    "I will take the subset of the samples given by the first experiment only and look at the distribution of the network capacity there."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Samples in experiment 1: 3111\n",
      "Unique time steps in experiment 1: 2260\n"
     ]
    }
   ],
   "source": [
    "df1 = get_experiment(df_tot, 1)\n",
    "print(f'Samples in experiment 1: {df1.shape[0]}')\n",
    "df1_grouped = df1.groupby('timestep_time')\n",
    "print(f'Unique time steps in experiment 1: {len(df1_grouped.groups)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Mm. I need to stop right here. Can you see the problem?\n",
    "\n",
    "In the first experiment I collected 3111 samples. Among them, there are 2260 unique timesteps. Overall, it looks like the `TABLESAMPLE BERNOULLI` did a great job at diversifying!\n",
    "\n",
    "But in this specific case, this raise an issue. I was going to count the number of vehicles for each unique timesteps, but with the numbers so close (3111 samples and 2260 unique timesteps) it is likely that each sampled timestep only contains information about one vehicle.\n",
    "\n",
    "Let's take a look at the maximum size of the timestep groups."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Largest timestep group size: 6\n"
     ]
    }
   ],
   "source": [
    "max_l = max(len(v) for v in df1_grouped.groups.values())\n",
    "print(f'Largest timestep group size: {max_l}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This means that the largest number of samples I got for the same timestep is six. And that implies that if I count the number of vehicles at every timestep I will get at most six of them. But this clearly isn't the maximum network capacity, it's just a sampling issue!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I need a different approach. What if I think from a vehicle point of view, instead of a timestep point of view?\n",
    "\n",
    "I could find unique vehicle IDs, then for each of them find the first and last timestep in the dataset that vehicle appears in. This should give me the window of timesteps that vehicle was inside the network.\n",
    "Then I could overlap this window with the other vehicle's window, so to find the largest intersection of them all. That would give me the largest number of vehicles that _might_ be inside the network at the same time.\n",
    "\n",
    "Why _might_? Because I have no way to know whether a vehicle exited the network and then entered it again. I will have to make the assumption that this doesn't happen.\n",
    "\n",
    "All this sounds a bit too much for a \"brief\" exploratory look, not to mention that I am not sure what type of statistical confidence I could infer from it. Anyway, let me try."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unique vehicles in experiment 1: 2872\n",
      "Max vehicle time experiment 1: 20375\n",
      "Number of vehicles detected in more than 1 timestep (experiment 1): 229\n"
     ]
    }
   ],
   "source": [
    "def find_min_max_time_veh(vehicle_id: str,\n",
    "                          df: pd.DataFrame) -> tuple:\n",
    "    \"\"\"Finds the min and max timestep for the given vehicle_id\"\"\"\n",
    "    afilter = df['vehicle_id'] == vehicle_id\n",
    "    filtered = df.loc[afilter]\n",
    "    min_ = min(filtered['timestep_time'])\n",
    "    max_ = max(filtered['timestep_time'])\n",
    "    return (min_, max_)\n",
    "\n",
    "\n",
    "def veh_time_inside(min_: float, max_: float) -> int:\n",
    "    \"\"\"Number of steps a vehicle was inside the network.\n",
    "    *Note* It assumes the vehicle didn't leave and come back.\n",
    "    \"\"\"\n",
    "    return int(max_ - min_)\n",
    "\n",
    "df1_grouped_veh = df1.groupby('vehicle_id')\n",
    "print(f'Unique vehicles in experiment 1: {len(df1_grouped_veh.groups)}')\n",
    "\n",
    "all_times = df1_grouped_veh.apply(lambda x: find_min_max_time_veh(x.iloc[0]['vehicle_id'], df1))\n",
    "new_df = pd.DataFrame({\n",
    "    'vehicle_id': all_times.index,\n",
    "    'min_': [t[0] for t in all_times],\n",
    "    'max_': [t[1] for t in all_times]\n",
    "})\n",
    "new_df['window'] = new_df.apply(lambda x: veh_time_inside(x['min_'], x['max_']), axis=1)\n",
    "max_time_inside = max(new_df['window'])\n",
    "print(f'Max vehicle time experiment 1: {max_time_inside}')\n",
    "num_veh_traced = sum(new_df['window'] > 0)\n",
    "print(f'Number of vehicles detected in more than 1 timestep (experiment 1): {num_veh_traced}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's what I have got:\n",
    "\n",
    "- In experiment 1, 229 vehicles were sampled more than once.\n",
    "- In experiment 1, the largest number of steps a vehicle is inside the network is 20375. That's seconds, so more than 5 hours. I bet is a bus.\n",
    "\n",
    "Let's think about the extreme case. Even if the time windows for those 229 vehicles overlapped, I could only conclude that the network handles 229 vehicles at the same time. That's obviously too low.\n",
    "\n",
    "And even if those time window were exactly the same for 200 vehicles more for every other experiment, that would mean there are around 200 vehicles at the same in the network but... with what statistical confidence could I say it? It sounds like plain coincidence.\n",
    "\n",
    "Again, it feels as if I don't have enough data to answer this question."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I will then go back to the source of data. Overall, the problem I am having so far is that the collected data seems to be taken from a GPS system that can cover the whole city but can only look at some random vehicles in every time steps.\n",
    "\n",
    "In other words, it cannot follow a vehicle.\n",
    "\n",
    "Can I get data that instead is more similar to a in-vehicle GPS system? For sure, but then I have to be careful because doing it for _all_ vehicles would result in too much data to transmit, load in memory and analyze.\n",
    "\n",
    "I am going to try to select samples from the database as if they were taken by a satellite. The satellite can take \"screenshots\" of the systems at given time instants. No system is perfect though, so the satellite can't really know _everything_. It can count though, so let's have it count the number of vehicles in those time instants.\n",
    "\n",
    "Like before, I will run multiple experiments (data extractions, really). The code is similar to the one used before, so let me skip the details this time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def satellite_experiment():\n",
    "    query = \"\"\"\n",
    "        SELECT timestep_time\n",
    "               , count(vehicle_id) as num_veh\n",
    "        FROM most_0400_1400_1_5\n",
    "        WHERE timestep_time in (\n",
    "            -- Randomly choose at what timesteps the system\n",
    "            -- can be \"photographed\".\n",
    "            SELECT distinct timestep_time\n",
    "            FROM most_0400_1400_1_5\n",
    "            TABLESAMPLE BERNOULLI (0.01)\n",
    "        )\n",
    "        GROUP BY timestep_time\n",
    "        ;\n",
    "    \"\"\"\n",
    "    with db_conn() as conn:\n",
    "        df_sat = pd.read_sql(query, conn)\n",
    "    return df_sat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 217 ms, sys: 51.7 ms, total: 269 ms\n",
      "Wall time: 6min 23s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "n_sat_experiments = 10\n",
    "sat_workers = 10\n",
    "with concurrent.futures.ThreadPoolExecutor(max_workers=sat_workers) as executor:\n",
    "    futures = [executor.submit(satellite_experiment)\n",
    "               for _ in range(n_sat_experiments)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "filep = '/home/pietro/statinf_satexp.csv'\n",
    "dframes = [f.result() for f in futures]\n",
    "for i, df in enumerate(dframes):\n",
    "    df['experiment'] = i + 1\n",
    "df_sat = pd.concat(dframes)\n",
    "df_sat.to_csv(filep, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>timestep_time</th>\n",
       "      <th>num_veh</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17755.0</td>\n",
       "      <td>34</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>17775.0</td>\n",
       "      <td>35</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18230.0</td>\n",
       "      <td>358</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>18280.0</td>\n",
       "      <td>423</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>18290.0</td>\n",
       "      <td>433</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   timestep_time  num_veh  experiment\n",
       "0        17755.0       34           1\n",
       "1        17775.0       35           1\n",
       "2        18230.0      358           1\n",
       "3        18280.0      423           1\n",
       "4        18290.0      433           1"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_sat.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(42278, 3)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_sat.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time instants sampled in experiment 1: 3696\n",
      "Average capacity in experiment 1: 12432.83 vehicles\n",
      "Max capacity in experiment 1: 18644 vehicles\n",
      "Capacity standard deviation in experiment 1: 5088.86 vehicles\n"
     ]
    }
   ],
   "source": [
    "df1_sat = get_experiment(df_sat, 1)\n",
    "num_steps = df1_sat.shape[0]\n",
    "avg_capacity = np.round(np.mean(df1_sat['num_veh']), 2)\n",
    "max_capacity = np.max(df1_sat['num_veh'])\n",
    "std_capacity = np.round(np.std(df1_sat['num_veh']), 2)\n",
    "print(f'Time instants sampled in experiment 1: {num_steps}')\n",
    "print(f'Average capacity in experiment 1: {avg_capacity} vehicles')\n",
    "print(f'Max capacity in experiment 1: {max_capacity} vehicles')\n",
    "print(f'Capacity standard deviation in experiment 1: {std_capacity} vehicles')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "OK, it seems like I've got some real data this time! In one time instant there were 18644 vehicles. Let's say in average a vehicle has the dimension of a 2 meters x 4 meters rectangle. That makes 149,152 squared meters occupied by vehicles, at that time instant. Which is around 1/5th of the area covered by the simulation. It sounds like a reasonable number, considering that most of the area is not available to vehicles (there are mostly buildings, and hopefully parks or other non-transit zones).\n",
    "\n",
    "This was just a sanity check, very rough but useful in my opinion. The numbers make sense.\n",
    "\n",
    "I will now apply to this question the same analysis strategy that I used before. Let's see where it gets me.\n",
    "\n",
    "First of all, I will take a look at the distribution for this sample only."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(20, 5))\n",
    "ax0 = axes[0]\n",
    "ax1 = axes[1]\n",
    "veh_col = 'num_veh'\n",
    "time_col = 'timestep_time'\n",
    "time_ax_label = 'time [H:M:S]'\n",
    "fig1 = 'Scatter plot of the number of vehicles'\n",
    "fig2 = 'Boxplot of the number of vehicles'\n",
    "\n",
    "xticks = [20e3, 25e3, 30e3, 35e3, 40e3, 45e3]\n",
    "df1_sat.plot.scatter(y=veh_col,\n",
    "                     x=time_col,\n",
    "                     xticks=xticks,\n",
    "                     ax=ax0);\n",
    "ax0.set_xlabel(time_ax_label);\n",
    "ax0.set_xticklabels([step_2_time(tick) for tick in xticks]);\n",
    "ax0.set_title(fig1);\n",
    "\n",
    "df1_sat.boxplot(veh_col, ax=ax1);\n",
    "ax1.set_title(fig2);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Interestingly, the number of vehicles when plotted against the sampled time instants looks like a normal function. But **remember this**: there are actually a lot of \"holes\" in that function, because I sampled the time instants randomly. Hence, I don't really know if for a time instant that wasn't sampled the number of vehicles is suddenly zero. But I agree with you that this is very unlikely.\n",
    "\n",
    "The box plot on the right is the distribution plot, just in a different form. It looks much nicer than the data I analyzed before. There's a well defined median, clear quartiles and so on.\n",
    "\n",
    "Now I would like to plot the distribution for this sample (experiment 1) in the classical form (that is a bar plot), that shows the number of times a certain value has occurred. However, this is a bit more tricky than before.\n",
    "\n",
    "The problem is that the value I want to see is a discrete counter of the vehicles. Thus, each such number is relatively unlikely to occurr a lot of times. For instance, if at a given time step there are 10,001 vehicles, that's a very specific number. I will surely have 10,000 in a different time step, maybe 10,002 as well, but maybe never again the same exact values.\n",
    "\n",
    "This means that a bar plot could just a lot of bars with low height, since the heights are the number of occurrences of each value. Let's take a look nonetheless."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "col = 'num_veh'\n",
    "toplot = pd.value_counts(df1_sat[col]).sort_index()\n",
    "x = list(toplot.index)\n",
    "y = list(toplot)\n",
    "avg = np.round(np.mean(df1_sat[col]), 2)\n",
    "title = 'Distribution of the number of vehicles (experiment 1)'\n",
    "plt.bar(x=x, height=y)\n",
    "plt.title(title);\n",
    "plt.axvline(x=avg, color='red');\n",
    "plt.xlabel('number of vehicles');\n",
    "plt.ylabel('num. samples');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Not very useful, as expected. To say the truth, in terms of aggregated statistics the box plot above says it all. Nonetheless, I want to take a look at a bar plot for this distribution.\n",
    "\n",
    "The common solution to this problem is to use histogram. A histogram is a plot with bins, and the width of every bin is decided based on the distribution. Matplotlib handles it off-the-shelf, but basically is like fixing a bucket width (say, 100) and assign all values into a bucket (say, all values between 0 and 100 go in bucket one, all those between 101 and 200 go in bucket two, and so on).\n",
    "\n",
    "For the sake of the argument, I will plot the histogram computed by matplotlib alongside a bar plot of manually computed buckets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "new_col = 'bucket'\n",
    "bucket = 1000\n",
    "df1_sat[new_col] = df1_sat['num_veh'].apply(lambda x: int(x / bucket))\n",
    "\n",
    "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(15,5))\n",
    "ax0 = axes[0]\n",
    "ax1 = axes[1]\n",
    "\n",
    "ax0.hist(df1_sat['num_veh'], density=False)\n",
    "ax0.set_title('Histogram (number of vehicles)')\n",
    "ax0.set_xlabel('num. vehicles')\n",
    "ax0.set_ylabel('num. samples')\n",
    "\n",
    "toplot = pd.value_counts(df1_sat[new_col]).sort_index()\n",
    "x = list(toplot.index)\n",
    "y = list(toplot)\n",
    "title = 'Bucket distribution (manually computed)'\n",
    "ax1.bar(x=x, height=y);\n",
    "ax1.set_title(title);\n",
    "ax1.set_xlabel(f'num. vehicles  x{bucket}');\n",
    "ax1.set_ylabel('num. samples');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The good news is that the figures now make sense (and they are similar, so I wasn't totally off road!). The bad news is that the distribution looks uniform-ish. There's a clear increase in frequency: larger values (towards the right-end of the figure) have larger frequencies (higher peaks), but other than that there's no clear pattern.\n",
    "\n",
    "In my opinion, the increase of the number of samples with higher number of vehicles can be explained intuitively: when there are a lot of vehicles they travel more slowly and therefore leave the city at slower rate. Thus, the actual number of vehicles stays similar for more time instants, hence it occurs more times.\n",
    "\n",
    "Apart from this speculation of mine, I have another problem with uniform distributions: I don't find their confidence intervals very useful.\n",
    "\n",
    "To be clear, t-intervals like the ones I've computed before with `st.t.interval(...)` are quite robust to non-gaussian distributions. Even then, a uniform distribution is a weird beast: all values are equally likely to occur! As such, a confidence interval will be quite the same (in terms of width) for all values, not just the mean.\n",
    "\n",
    "If I were in a situation like this, I would bring up to my team's attention the chart above rather than a confidence interval or other inferential reasoning.\n",
    "\n",
    "Now, with the same strategy that I already used twice, I am going to cumulate all samples from the satellite in a growing distribution and look at it _as it cumulates_."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1800 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure = plt.figure(figsize=(20, 25))\n",
    "\n",
    "# Start from the first experiment\n",
    "current_df = get_experiment(df_sat, 1)\n",
    "col = 'num_veh'\n",
    "\n",
    "all_avg = [np.round(np.mean(current_df[col]), 2)]\n",
    "\n",
    "for i in range(1, 11):\n",
    "    if i > 1:\n",
    "        # If not first iteration\n",
    "        # then cumulate the next sample.\n",
    "        df_exp_i = get_experiment(df_sat, i)\n",
    "        all_avg.append(np.round(np.mean(df_exp_i[col]), 2))\n",
    "        current_df = pd.concat([current_df, df_exp_i])\n",
    "\n",
    "    # Plot the cumulated-so-far samples\n",
    "    current_ax = figure.add_subplot(4, 3, i)\n",
    "    avg = np.round(np.mean(all_avg), 2)\n",
    "    if i > 1:\n",
    "        title = f'{i} samples'\n",
    "    else:\n",
    "        title = '1 sample'\n",
    "    title += f' (mean = {avg})'\n",
    "    current_ax.hist(current_df[col], density=False)\n",
    "    plt.title(title);\n",
    "    plt.axvline(x=avg, color='red');\n",
    "    plt.xlabel('num. vehicles')\n",
    "    plt.ylabel('num. samples')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Oh well, the distribution really converges to a almost-uniform distribution. I hadn't looked at these 10 plots when I made my conclusions in the previous paragraph!\n",
    "\n",
    "What could be the reason for it?\n",
    "\n",
    "I will take a guess, and say that over many cumulated samples the odds of congested traffic (like during rush-hours) vs free-flow (like at 5am) even out and produce a uniform-looking behavior. But I am convinced that if somebody looked at the data more closely, partitioning it based on hours of the day, then the congested trend would clearly appear.\n",
    "\n",
    "With all that being said, the result still stands: over ten experiments with a few thousand samples each, the most frequent values of number of vehicles are the high values, well above 17,500 vehicles."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Open questions\n",
    "\n",
    "That was ... a lot!\n",
    "\n",
    "A lot of figures, a lot of analysis and thoughts. And I was able to touch just the surface of the data, making two neat conclusions:\n",
    "\n",
    "  1. Vehicles in Monaco travel at a speed between 37.6 km/h and 38.98 km/h with 95% probability.\n",
    "  2. The highest capacity experienced by the city is well above 17,500 vehicles, though over one day the instantaneous capacity looks like a uniformly distributed random number.\n",
    "\n",
    "There's so much more to explore, and so many unanswered questions. Here are a few, off the top of my mind:\n",
    "\n",
    " - **From where do most vehicles enter the network?** I think it would be very useful to generate a 2D map of the city that shows the hotspots where a lot of objects enter the city.\n",
    "\n",
    " - **What are the most congested spots?** What about a 2D map that shows the most congested spot, where traffic is the worst.\n",
    "\n",
    " - **What are the lengths of the 10 longest queues of vehicles? Where are they?**\n",
    "\n",
    "If you are interested in exploring these questions, remember that the dataset is free and available to everyone, you included. If you start playing with the data, fork this repository and when you're finished submit a PR so we can publish your article!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}