Right, I think I can guess from your file naming where the files come
from ;) I've had a look at the code in the *_jw_mf.py files, and I
can see that this code is redundant The changes are simply a massive
renaming of 1 single variable across the entire code base. I'm
looking at the code in:
https://www.ncbr.chemi.muni.cz/~kada/relax-1.2.10_cst.tar
There is a far easier and quicker way to do this for the 1.3 line, and
that is in the mf_csa.py file (when I get to this one, I'll look and
see if this needs to be a separate file (which I don't like)).
Because these are all functions accepting the 'data' data storage
object, you should create 4 of these data objects instead and run the
functions on each one by one. These 4 structures should be named
something like 'data_csa1', 'data_csa2', 'data_csacross', 'data_dipY'.
I need to study your code more to understand everything that is
happening. These then all should use the identical jw_mf.py
functions, as these do not in anyway change.
For the 'cst_1.2.10' branch, I'll use 'svn cp' to create your files,
and then put in all your changes into the 'maths_fns' directory (the
other parts from Petr will go in later). Please watch the commit
messages going to the relax-commits mailing list. For the 'cst'
branch, this duplication is bad programing style and the 4 different
structures will need to be created. Don't worry about your
programming abilities though as I will help you here if you need it.
Regards,
Edward
On Thu, Jan 22, 2009 at 1:42 PM, Edward d'Auvergne <edward@xxxxxxxxxxxxx>
wrote:
On Thu, Jan 22, 2009 at 12:41 PM, Pavel Kaderavek <kada@xxxxxxxxxxxxx>
wrote:
I have forgotten to answer one question. Everything added in maths_fns
directory come from an another file ... so that I wanted to mimic the old
code as much as possible.
Ah, that's useful to know. Once I find a new file, I'll ask where it
comes from. Unless you already know the origin of everything and can
send that information in a post to relax-devel.
Cheers,
Edward
Re: cst branch development