FlightNotebook.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import folium\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import geojson"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Example ADS-B data gathered from flightradar24\n",
    "flight = pd.read_csv('24299085.csv')\n",
    "\n",
    "print(list(flight))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We can use simple accessor methods to find out useful data\n",
    "flight['Altitude'].min(),flight['Altitude'].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We can do more complex operations like finding the difference between each successive pair of value\n",
    "flight['diff'] = flight['Altitude'].diff()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "flight['diff'].min(),flight['diff'].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We can also create functions to apply to our data\n",
    "def energy(speed,altitude,mass,gravity=9.81):\n",
    "    ke=0.5*mass*speed**2\n",
    "    pe=mass*gravity*altitude\n",
    "    return ke+pe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Using 5000kg as reasonably wrong answer for Jetstream 31\n",
    "print(energy(flight['Speed'],flight['Altitude'],5000))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We can also apply it in a vectorised fashion on our pandas dataframe which is more efficent if your data is large\n",
    "mass=5000\n",
    "gravity=9.81\n",
    "flight['ke'] = flight.apply(lambda r: 0.5*mass*r['Speed']**2,axis=1)\n",
    "flight['pe'] = flight.apply(lambda r: gravity*mass*r['Altitude'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "flight['Datetime'] = pd.to_datetime(flight['UTC'],utc=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, axes = plt.subplots(3,1,sharex='all')\n",
    "axes[0].plot(flight['Datetime'],flight['Altitude'])\n",
    "axes[0].set_ylabel('Altitude')\n",
    "axes[1].plot(flight['Datetime'],flight['Speed'])\n",
    "axes[1].set_ylabel('Speed')\n",
    "axes[2].plot(flight['Datetime'],flight['Direction'])\n",
    "axes[2].set_ylabel('Direction')\n",
    "axes[2].set_xlabel('UTC')\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, axes = plt.subplots(4,1,sharex='all')\n",
    "axes[0].plot(flight['Datetime'],flight['Altitude'])\n",
    "axes[0].set_ylabel('Alt')\n",
    "axes[1].plot(flight['Datetime'],flight['Speed'])\n",
    "axes[1].set_ylabel('Speed')\n",
    "axes[2].plot(flight['Datetime'],flight['ke'])\n",
    "axes[2].set_ylabel('KE')\n",
    "axes[3].plot(flight['Datetime'],flight['pe'])\n",
    "axes[3].set_ylabel('PE')\n",
    "axes[3].set_xlabel('UTC')\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "coords = [[float(b) for b in a.split(',')] for a in flight['Position']]\n",
    "#This is a nest list comphrehension to convert the paired coordinates from their \n",
    "# comma seperated string format to a pair of floats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "flipped_coords = [[a[1],a[0]] for a in coords]\n",
    "#A common problem in GIS is that different systems expect the coordinates in a different order.   \n",
    "#Here i use a list comphrehension to swap the order for every coordinate in my data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "flight_gj = geojson.FeatureCollection([\n",
    "    geojson.Feature(properties={\n",
    "    \"Name\": \"CFDD01P\",\n",
    "        \"Colour\":'green',\n",
    "        \"Layer\":\"Radar Data\"\n",
    "    },\n",
    "                        geometry={\"type\":\"LineString\", \"coordinates\":\n",
    "    flipped_coords} )\n",
    "]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "m = folium.Map()\n",
    "folium.GeoJson(flight_gj,name='CFDD01P').add_to(m)\n",
    "folium.LayerControl().add_to(m)\n",
    "m.fit_bounds(m.get_bounds())\n",
    "m"
   ]
  }
 ],
 "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": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}