simulation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Running an example simulation\n",
    "It's all very simple, mostly because this is a simple simulation. \n",
    "First, let's import stuff we'll need later on:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "from pandemic_sim.simulation import Person, Simulation\n",
    "from pandemic_sim.geometries import RectangleGeometry\n",
    "from pandemic_sim.health_systems import SimpleHealthSystem\n",
    "from pandemic_sim.particle_engines import (DefaultParticleEngine,\n",
    "                                           VelocityVerletIntegrator)\n",
    "from pandemic_sim.transmission_models import DefaultTransmissionModel\n",
    "from pandemic_sim.disease_models import DefaultPersonalDiseaseModel\n",
    "from pandemic_sim.visualizations import (DefaultVisualization,\n",
    "                                         DefaultPersonsDrawer,\n",
    "                                         RectangleGeometryDrawer,\n",
    "                                         SimpleHealthSystemCurvesPlotter)\n",
    "from pandemic_sim.animators import CelluloidAnimator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `Person` is class representing a person with attributes such as position in the room, infection status and whether the person is dead or alive\n",
    "- `RectangleGeometry` represents a rectangular room and implements the force the walls exert on persons\n",
    "- `Simulation` is the class containing the actual simulation code\n",
    "- `SimpleHealthSystem` models the influence of the health system on the probability of a person dying during their infection. Setting a threshold and a death rate factor, you can emulate the limited ICU bed capacity.\n",
    "- `DefaultParticleEngine` is a class which performs the purely physics part of the simulation, meaning the movement of persons and the forces acting between them. It does that with the help of the `VelocityVerletIntegration` class, which numerically integrates equations of motion\n",
    "- `DefaultTransmissionModel` is a disease transmission model based on probabilities for exposing other people and for being susceptible\n",
    "- `DefaultPersonalDiseaseModel` is a model which, for a single person, determines how likely that person is to die or to be cured from the disease in given simulation time step\n",
    "- in `DefaultVisualization` is the... well... default visualization. It shows all persons as solid circles and additionally plots the timeline of infected and immune persons and fatalities\n",
    "- `RectangeGeometryDrawer`, `DefaultPersonsDrawer`, `SimpleHealthSystemCurvesPlotter`: classes in which the drawing / plotting of different things is implemented\n",
    "- `CelluloidAnimator` contains code to animate a visualization using the [`celluloid` package](https://github.com/jwkvam/celluloid)\n",
    "\n",
    "If you want to change any of these components, the object-oriented design should make it relatively easy to, e.g., implement a new visualization. The visualization and animation code is currently not documented, though.\n",
    "\n",
    "Preamble done, so let's start setting up our simulation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_persons = 200\n",
    "room = RectangleGeometry(25, 25)\n",
    "transmission_model = DefaultTransmissionModel(prob_dist=lambda d: d < 1)\n",
    "# persons start with random positions and velocities\n",
    "initial_positions = room.get_random_position_set(n_persons)\n",
    "persons = [Person(pos=pos,\n",
    "                  vel=np.random.uniform(low=(-2, -2), \n",
    "                                        high=(2, 2)),\n",
    "                  personal_disease_model=None, personal_transmission_model=None,\n",
    "                  infected=False, immune=False)\n",
    "           for pos in initial_positions]\n",
    "for p in persons:\n",
    "    p.personal_transmission_model = transmission_model.personal_tm_factory(\n",
    "        p, in_prob=1.0, out_prob=0.05)\n",
    "    p.personal_disease_model = DefaultPersonalDiseaseModel(p, death_probability=0.00015, time_to_heal=150)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So we're going to have 200 persons moving around in a room with width 25 and height 25 (arbitrary units). The transmission model has an argument `prob_dist` which is a base probability for transmission to occur and which depends on the distance between persons. `death_prob` is the probability for an infected person to die in one unit of simulation time. The person-specific transmission model takes a probability for being susceptible (`in_prob`) and a probability for exposing others (`out_prob`). The latter is rather low, simulating the effect of a simple tissue face mask. `time_to_heal` is the number of simulation time units it takes for an infected person to get cured (and thus become immune)\n",
    "Now we want some random persons to start out infected:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "chosen_ones = np.random.choice(np.arange(n_persons), n_persons // 50)\n",
    "for i in chosen_ones:\n",
    "    persons[i].infected = True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sorry, guys.  \n",
    "On to setting up the simulation object:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "health_system = SimpleHealthSystem(threshold=50, death_probability_factor=5.0)\n",
    "pe = DefaultParticleEngine(cutoff=0.75, geometry_gradient=room.gradient,\n",
    "                           integrator_params={'timestep': 0.1},\n",
    "                           integrator_class=VelocityVerletIntegrator,\n",
    "                           inter_particle_force_constant=20.0,\n",
    "                           geometry_force_constant=20.0)\n",
    "sim = Simulation(room, persons, health_system, transmission_model, particle_engine=pe)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The parameters with which the health system object is initialized mean that as soon as more than 150 persons are infected, the probability of a person dying in a time step is increased by a factor of three.  \n",
    "As for the simulaton object: lots of parameters here:\n",
    "- `timestep`: the time step for the integration scheme, which approximately solves the equations of motions for persons. The smaller that parameter is, the more accurate and detailed the simulated movement, but the longer the simulation takes to cover some predefined time span\n",
    "- `cutoff`: distance at which the repulsive force between two persons (which makes them bounce off each other) kicks in\n",
    "- `inter_particle_force_constant`: determines how hard two people bounce off each other. If this were set to zero, they could just pass through each other unhindered\n",
    "- `geometry_force_constant`: determines how hard persons bounce off walls\n",
    "\n",
    "Now we can run the simulation for some number `n_steps`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Simulating step 0/500...\n",
      "Simulating step 50/500...\n",
      "Simulating step 100/500...\n",
      "Simulating step 150/500...\n",
      "Simulating step 200/500...\n",
      "Simulating step 250/500...\n",
      "Simulating step 300/500...\n",
      "Simulating step 350/500...\n",
      "Simulating step 400/500...\n",
      "Simulating step 450/500...\n",
      "Done.\n"
     ]
    }
   ],
   "source": [
    "n_steps = 500\n",
    "simulation_results = sim.run(n_steps)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Done? Amazing. Let's use the above-discussed `DefaultVisualization` and the `CelluloidAnimator` to get a `.mp4` file with the animated simulation results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Animating step 0/500...\n",
      "Animating step 50/500...\n",
      "Animating step 100/500...\n",
      "Animating step 150/500...\n",
      "Animating step 200/500...\n",
      "Animating step 250/500...\n",
      "Animating step 300/500...\n",
      "Animating step 350/500...\n",
      "Animating step 400/500...\n",
      "Animating step 450/500...\n",
      "Done.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "radius = pe.cutoff / 2\n",
    "viz = DefaultVisualization(simulation_results, RectangleGeometryDrawer(room),\n",
    "                           DefaultPersonsDrawer(radius), SimpleHealthSystemCurvesPlotter(health_system))\n",
    "animator = CelluloidAnimator(viz, out=\"output.mp4\", frame_rate=20)\n",
    "animator.animate(n_steps, interval=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `radius` determines the radius of the circles representing the persons. You want to set this to `pe.cutoff / 2`, otherwise things will look weird (persons might not touch or overlap too much). The dashed red line in the \"# infected\" plot is the health system threshold.\n",
    "Animating takes quite a while, too; comparable to the time it takes to run the simulation itself. In the animation, blue dots represent healthy persons, orange dot infected persons, and gray dots dead persons.  \n",
    "Let's view the results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<video src=\"output.mp4\" controls  >\n",
       "      Your browser does not support the <code>video</code> element.\n",
       "    </video>"
      ],
      "text/plain": [
       "<IPython.core.display.Video object>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Video\n",
    "Video(\"output.mp4\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That's it&mdash;enjoy playing around with the parameters and the code and if you have suggestions on how to improve all this, feel free to open an issue and / or a pull request!"
   ]
  }
 ],
 "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": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}