This section describes how to install the pipeline in development mode directly from the source code repositories.
This method requires you have an IDL license. If you do not have an IDL license, please refer to this page for installing compiled executables instead.
The GPI Pipeline souce code is now developed using GitHub at http://github.com/geminiplanetimager/gpi_pipeline
If you are not familiar with version control or Github, there are many tutorials available online, for instance this one.
Once you have that account, in a directory of your choosing (preferably somewhere inside your $IDL_PATH) execute the commands:
>>> git clone https://github.com/geminiplanetimager/gpi_pipeline.git >>> git clone https://github.com/geminiplanetimager/gpi_pipeline_external.git
The above commands will download the GPI data pipeline and associated tools in gpi_pipeline, plus a directory gpi_pipeline_external containing various external dependencies of the code, for instance the Goddard IDL library, the Coyote library, etc. You may already have copies of many of these routines in your own IDL library, in which case you can delete the excess copies from this directory if you so desire. Two folders (gpi_pipeline and gpi_pipeline_external) will be created in the directory from which you performed the command.
All code has been tested using the versions of external program in the repository. Error-free operation is not guaranteed for other versions of these libraries. The usual caveats about name collisions between different versions of IDL routines apply. If you have old versions of e.g. the Goddard IDL library functions in your $IDL_PATH, you may encounter difficulties. We suggest placing the data pipeline code first in your $IDL_PATH.
Mac OS and Linux
On Mac OS and Linux, you will likely want to add the gpi_pipeline/scripts subdirectory to your shell’s $PATH.
For users having IDL 8.2+, the str_sep.pro program is now an obsolete command. Although no pipeline source code calls this function, it is still used in other external dependencies. For the time being, users should add the idl/lib/obsolete folder to their $IDL_PATH to remedy this issue.
Proceed now to Configuring the Pipeline.