[Magellan-users] A fully working version of magellan approaching

[Tryggvi Björgvinsson]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I just committed a rather large patch to magellan. I should have broken 
it down into more commits instead of one large commit (for better 
management), I'll do so in the future.

The commit included the following:

* Refactoring (restructuring the code for better performance and 
maintenance) of data.py, the file that reads different parameter files

* Implemented the magnetization file (it still isn't used in the model 
but it can read the values of a magnetization file and put them in the 
appropriate data structure)

* Implemented the configuration file so that users now only need to send 
in one file (the config file) which has links to other files which 
magellan uses. The configuration file reads which files to use through 
lines like:
assymetry = location of assymetry file
spreading rate = location of spreading rate file
jumps = location of jumps file
timescale = location of timescale file (if Cande & Kent [1995] is not used)
magnetization = location of magnetization file
data = location of data file containing track data

Defining files through flags (e.g. --assymetry assymetry_file) will 
overwrite the config file.

* Removed the idea of the parameter file (discussed in a previous TODO 
list) and implemented flags -d or --declination, -i or --inclination, -z 
or --thickness to control the parameters which were meant for the 
parameter file. These values can also be set in the configuration file, 
similar to the files (and overwritten with flags):
inclination = 75
declination = 18
thickness = 0.5

* Plot the pseudo faults and failed rifts in the bathymetry plot. Before 
magellan only marked where those were, now they get plotted (I chose 
yellow for pseudo faults and red for failed rifts).


Now the only thing that remains is to (all marked in the TODO file):

* Finish implementing the model (that's the major objective but we're 
well on our way)

* Implement the subroutine which computes the azimuth

* Create the graphs structure where users can define which graphs they 
want plotted (this requires some thinking and probably refactoring of 
plot.py). I have also been pondering the option of choosing "no graphs" 
which will make magellan write out information on standard output (into 
the terminal window where magellan is run). This might make it easier to 
see exact locations and read information (analyse by hand or send into 
another program). The no graphs option requires some heavy thinking 
about the structure of the output.


I believe that when we have implemented the remaining features above we 
could release an alpha version 1.0 of magellan. An alpha version is a 
version which might include bugs and be unstable. With an alpha version 
we could start using and testing magellan with real data and report all 
incidents where it doesn't work. I don't want to release "version 1.0" 
until it has been thoroughly tested and compared to the output of other 
similar programs like Magbath and MODMAG. After all we are aiming for 
software which has more functionality but it should compute correctly.

After the release of version 1.0 we could start optimizing magellan for 
performance, i.e. make it faster. We could then also start focusing on 
new functionality.

/Tryggvi


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFH4uUxTfUwC3N5Fj0RAojLAKC2wU/mbF5ePh8Tu6dwQ4CTzEZG7QCcCMTO
S00r4bZkHjC00bxc7NLlNCo=
=AhAp
-----END PGP SIGNATURE-----