Installing brutus¶
The most recent release of brutus is available for download from its Github repository.
Note
You can also get the current (unreleased and possibly buggy) code from here.
Installing brutus merely requires to let your Python installation know about its existence. Specifically:
- Unzip the compressed files (downloaded from the link above),
- place them anywhere you like, and
- add
brutus/src/
to your Python path.
In my case (on MAC OSX), my .bash_profile
would look like :
export PYTHONPATH=$PYTHONPATH:/Users/fvogt/Tools/Python/fpav_pylib/brutus/src/
Github users can fork the brutus repository if they want to get access to the latest updates not yet formally released. Push requests for bug fixes and new features are welcome and will be examined in detail.
Requirements¶
The basic packages below are required for brutus to work properly:
- numpy (1.9.0 or or above)
- scipy (0.14.0 or above)
- matplotlib (1.4.2 or above)
- aplpy
- montage (download | install instructions)
- astropy
- photutils
- statsmodel
The packages below might be required, depending on what step is being executed. Install what you need:
A note on the brutus plots¶
brutus relies on matplotlib (and aplpy which itself relies on matplotlib) to create plots.
By default, brutus will try to use a full \(\LaTeX\) installation present on the system
(a.k.a. via the rcparams
setting usetex: True
) to create good looking diagrams.
If this is a problem (e.g. if \(\LaTeX\) is not present on the machine that runs brutus),
you can:
- install TexLive [recommended], to get the brutus plots to look they way they are meant to look, or
- set
mpl.rc('text', **{'usetex':False})
inbrutus_plots.py
to use the limited \(\LaTeX\) capabilities that ship with matplotlib.
Nate that if you decide to go with the second option, you might still encounter issues here and there.
Testing the installation¶
Todo
Eventually, I should create an elegant little thing to test that brutus is working fine.
For now, just try to import brutus
.
Troubleshooting¶
If something goes bad, here is what you should do (mind the order !):
Are all the required packages up-to-date ? Check the Requirements.
Still no luck ? Check the FAQ.
Still no luck ? Check if the issue has been reported before.
Still no luck ? Please submit a new issue on the Github repository of the project.
Provide as much detail as possible (error message, minimal example able to reproduce the error, operating system, Python version, multiprocessing state, etc ...).
Note
Submitting a Github issue is the best way for you to get help rapidly, for everyone to keep track of the problems that need solving, and for future users to see what changes have been made over time (and look for existing solutions to their problem which may be the same as yours). Submitting a new issue on Github is rapid and easy, but if you are really against doing it (why would you ?), you can always email frederic.vogt@alumni.anu.edu.au for help.