A parallel Barnes Hut package for Julia. Note that currently, the following examples do not use initial conditions which are particlarly accurate for learning about an actual galaxy, but the user can create custom initial particle distributions.
More tools, such as better initial distribution generators, dark matter particles, and plots of total energy and angular momentum over time would be useful additions in the future.
Run the following code in Julia to install the ParallelBarnesHut package:
import Pkg; Pkg.add(Pkg.PackageSpec(url = "https://github.com/alexhad6/ParallelBarnesHut.jl"))
using ParallelBarnesHutIt will take some time to install and precompile.
This documentation can also be accessed in Julia using the search feature, which is a question mark followed by the name of the object or function to learn about. See ParallelBarnesHut.jl for source code.
Julia struct that represents a particle in a Barnes Hut simulation. For a galaxy simulation, a particle is a star or a group of stars.
Particle.pos: position coordinates of a particle (in meters).Particle.vel: velocity vector of a particle (in meters per second).Particle.mass: mass of a particle (in kilograms).
Approximately solve N-body problem for an array of particles for a number of time steps into the future using the Barnes Hut method. Optionally specify Δt, the length of a time step in seconds, θ, the node size to distance threshold for the Barnes Hut algorithm, and the acceleration function. Return an array of particle arrays, where each particle array is the state of all particles at each time step.
Calculate the gravitation acceleration on a particle given the mass of another object and the distance vector from particle to object. The usual gravitational force is softened by a factor ϵ to prevent it from blowing up when the distance is very small.
Using the Plots package, display an array of particles in 3D with azimuthal camera angle θ and elevation angle ϕ.
Display four views of the particles in 3D at different camera angles.
Calculate and display in 3D the cubes represented by the Barnes Hut octree.
animate_frames(frames, frame_rate = 1, file_name = "animation"; θ = 30.0, ϕ = 30.0, quadview = false)
Plot particles for every frame_rate frames resulting from a Barnes Hut simulation, and generate a gif file of this animation with a custom name. Optionally, specify azimuthal camera angle θ and elevation angle ϕ, or set quadview to true in order to animate at four different angles.
Generate num_particles random particles using a distribution similar to a disk-shaped galaxy.
See ParallelBarnesHut_Examples.ipynb for a Jupyter notebook file which runs the following examples.
You can generate particles in any distribution you want using the Particle constructor as explained in the Documentation section above, or you can use the rand_particles function which comes with this package.
particles = rand_particles(1000)We can view the particles using the show_particles function.
show_particles(particles)
We can visualize cubical regions represented by nodes in the 3D Barnes Hut octree using the show_boxes function.
show_boxes(particles)
To run a Barnes Hut simulation and save frames of the states of all particles at each time step, we can use simulate!. The following code runs a simulation on our particles using the default values and going for 5000 time steps.
frames = simulate!(particles, 5000)The following code generates an animation, displaying every 5 frames (so 1000 frames total) and using the quadview display mode.
animate_frames(frames, 5, quadview = true)