Documentation

[rabernat]

At this point, with a growing number of users, we need documentation for floater.

We should use sphinx and readthedocs. We can copy the template from xgcm:
https://github.com/xgcm/xgcm/tree/master/doc

[cmazoyer]

Dear all,
I'm trying to install floater and I encounter problems. Do you have some documentation for installation, and use that could help me for the first steps? (I'm using conda, with python 2.7)
Floater seems very nice. Creating a netcdf from mitgcm files is a very good idea.
Thanks a lot.

ps: sorry if it's not the right place for asking this, I search where to contact you, and ask questions, and I didn't find any place.

[rabernat]

Hi @cmazoyer. Thanks for your interest in this package! Unfortunately we don't have any documentation yet for this package. It is mostly being used internally by our group and is still under very heavy development / flux.

To install floater, you should be able to just do

python setup.py install

If you encountered an error, please post the details of your error as a new issue.

I want to make sure that you understand the scope of floater. It is only for working with the output of the MITgcm FLT package (Lagrangian particle tracking). It is not a general purpose tool for "creating a netcdf from mitgcm files." Are you working with the FLT package? Or are you looking for a general-purpose python package for reading MITgcm mds data? If so, I think you are looking for xmitgcm:
http://xmitgcm.readthedocs.io/en/latest/

[cmazoyer]

Hi Ryan,
Thanks a lot for your reply. I may not have expressed myself very well, but I am searching a tool for working with FLT package output. I don't use xmitgcm because I work with mitgcm netcdf outputs with Ferret (I ask mitgcm to write netcdf directly). I'm starting using FLT package, and as I like working with netcdf and Ferret, I wanted to try your package. So far, I trying to understand the FLT example in mitgcm, and my goal is to compare real drifter trajectories we used, with lagrangian buoys in my simulation.
I will write my questions and errors as a new issue.
Thanks again,
Camille