Getting Started

Welcome to the gala documentation!

For practical reasons, this documentation generally assumes that you are familiar with the Python programming language, including numerical and computational libraries like numpy, scipy, and matplotlib. If you need a refresher on Python programming, we recommend starting with the official Python tutorial, but many other good resources are available on the internet, such as tutorials and lectures specifically designed for using Python for scientific applications.

On this introductory page, we will demonstrate a few common use cases for gala and give an overview of the package functionality. For the examples below, we will assume that the following imports have already been executed because these packages will be generally required:

>>> import astropy.units as u
>>> import numpy as np

Computing your first stellar orbit

One of the most common use cases for gala is to compute an orbit for a star within a mass model for the Milky Way. To do this, we need to specify two things: (1) the model of the Milky Way that we would like to use to represent the mass distribution, and (2) the initial conditions of the star’s orbit.

Mass models in gala are specified using Python classes that represent standard gravitational potential models. For example, most of the standard, parametrized gravitational potential models introduced in Binney and Tremaine [2008] are available as classes in the gala.potential module. The standard Milky Way model recommended for use in gala is the MilkyWayPotential, which is a pre-defined, multi-component mass model with parameters set to fiducial values that match the rotation curve of the Galactic disk and the mass profile of the dark matter halo. We can create an instance of this model with the fiducial parameters by instantiating the MilkyWayPotential class without any input:

>>> import gala.potential as gp
>>> mw = gp.MilkyWayPotential()
>>> mw
<CompositePotential disk,bulge,nucleus,halo>

This model, by default, contains four distinct potential components as listed in the output above: disk, bulge, nucleus, and halo components. You can configure any of the parameters of these components, or create your own “composite” potential model using other potential models defined in gala.potential, but for now we will use the fiducial model as we defined it, the variable mw.

All of the gala.potential class instances have a set of standard methods that enable fast calculations of computed or derived quantities. For example, we could compute the potential energy or the acceleration at a Cartesian position near the Sun:

>>> xyz = [-8., 0, 0] * u.kpc
<Quantity [-0.16440296] kpc2 / Myr2>
>>> mw.acceleration(xyz)  
<Quantity [[ 0.00702262],
           [-0.        ],
           [-0.        ]] kpc / Myr2>

The values that are returned by most methods in gala are provided as Astropy Quantity objects, which represent numerical data with associated physical units. Quantity objects can be re-represented in any equivalent units, so, for example, we could display the energy or acceleration in other units:

>>> E =
<Quantity [-157181.98979398] km2 / s2>
>>> acc = mw.acceleration(xyz)
>>> / u.Myr)  
<Quantity [[ 6.86666358],
           [-0.        ],
           [-0.        ]] km / (Myr s)>

Now that we have a potential model, if we want to compute an orbit, we need to specify a set of initial conditions to initialize the numerical orbit integration. In gala, initial conditions and other positions in phase-space (locations in position and velocity space) are defined using the PhaseSpacePosition class. This class allows a number of possible inputs, but one of the most common inputs are Cartesian position and velocity vectors. As an example orbit, we will use a position and velocity that is close to the Sun’s Galactocentric position and velocity:

>>> import gala.dynamics as gd
>>> w0 = gd.PhaseSpacePosition(pos=[-8.1, 0, 0.02] * u.kpc,
...                            vel=[13, 245, 8.] *

By convention, I typically use the variable w to represent phase-space positions, so here w0 is meant to imply “initial conditions.” Note that, when passing in Cartesian position and velocity values, we typically have to pass them in as Quantity objects (i.e., with units). This is required whenever the potential class you are using has a unit system, which you can check by calling the units attribute of your potential object:

>>> mw.units
<UnitSystem (kpc, Myr, solMass, rad)>

Here, our Milky Way potential model has a unit system with dimensional units. Note that we could have used any length unit for the position and any velocity unit for the velocity, because gala handles the unit conversions internally.

Now with a potential model defined and a set of initial conditions, we are set to compute an orbit! To do this, we use the numerical integration system defined in gala.integrate, but do so using the convenience interface available on any Potential object through the integrate_orbit() method:

>>> orbit = mw.integrate_orbit(w0, dt=1*u.Myr, t1=0, t2=2*u.Gyr)

By default, this method uses Leapfrog integration , which is a fast, symplectic integration scheme. The returned object is an instance of the Orbit class, which is similar to the PhaseSpacePosition but represents a collection of phase-space positions at times:

>>> orbit
<Orbit cartesian, dim=3, shape=(2000,)>

Orbit objects have many of their own useful methods for performing common tasks, like plotting an orbit:

>>> orbit.plot(['x', 'y'])  

(Source code, png, pdf)


Orbit objects by default assume and use Cartesian coordinate representations, but these can also be transformed into other representations, like Cylindrical coordinates. For example, we could re-represent the orbit in cylindrical coordinates and then plot the orbit in the “meridional plane”:

>>> fig = orbit.cylindrical.plot(['rho', 'z'])  

(Source code, png, pdf)


Or estimate the pericenter, apocenter, and eccentricity of the orbit:

>>> orbit.pericenter()  
<Quantity 8.00498069 kpc>
>>> orbit.apocenter()  
<Quantity 9.30721946 kpc>
>>> orbit.eccentricity()  
<Quantity 0.07522087>

gala.potential Potential objects and Orbit objects have many more possibilities, so please do check out the narrative documentation for gala.potential and gala.dynamics if you would like to learn more!

What else can gala do?

This page is meant to demonstrate a few initial things you may want to do with gala. There is much more functionality that you can discover either through the tutorials or by perusing the user guide. Some other commonly-used functionality includes:

Where to go from here

The two places to learn more are the tutorials and the user guide:

  • The Tutorials are narrative demonstrations of functionality that walk through simplified, real-world use cases for the tools available in gala.

  • The User Guide contains more exhaustive descriptions of all of the functions and classes available in gala, and should be treated more like reference material.



J. Binney and S. Tremaine. Galactic Dynamics: Second Edition. Princeton Series in Astrophysics. Princeton University Press, 2008. ISBN 9781400828722. URL: