help-glpk
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Help-glpk] Comments on version 3.1 API


From: Erik de Castro Lopo
Subject: [Help-glpk] Comments on version 3.1 API
Date: Sat, 1 Jun 2002 14:45:41 +1000

Andrew,

I am in the process of converting my FIR design program from version
3.0.6 to 3.1 and found something that distrubs me somewhat. 

Firstly, there is no direct replacement for the function glp_new_aij().
I will be the first to admit that this is a very inefficient way of 
filling in the required data but it is relatively foolproof. In addition, 
since filling in the data takes less than a second and finding the solution
takes many minutes, I do not see this as a problem. To fix this minor 
oversight, it would be nice to have a function 

    void lpx_set_mat_row_col (LPX *lp, int row, int col, double val);

to replace glp_new_aij(). Currently I am forced to do this:

    lpx_set_mat_row (problem, row, 1, (&col) - 1, (&aij) - 1) ;

Anyway, that brings me to my main point, the specification of the functions
lpx_set_mat_row() and lpx_set_mat_col() which from the documentation seem to
access arrays from index 1 through N rather than the normal C language 0
through N-1. 

Since anyone who is using your library is probably writing in and familiar
with C, they would be expecting arrays with a zero base index. Every 
programmer who assumes this is the case with GLPK will have a bug they need 
to fix before they can get any useful results out of the library. 

As a library author myself I predict that you will be getting a lot of email 
about this issue. Is it too late to reconsider this decision?

Regards,
Erik
-- 
+-----------------------------------------------------------+
  Erik de Castro Lopo  address@hidden (Yes it's valid)
+-----------------------------------------------------------+
"Every time microshaft's stock price drops again, I rejoice. I want to see that 
bunch of criminals brought to their knees. Preferably at the chopping block."
-- rixt in http://linuxtoday.com/stories/20659_flat.html



reply via email to

[Prev in Thread] Current Thread [Next in Thread]