Re: [ublas] Testing the new bindings

Subject: Re: [ublas] Testing the new bindings
From: Rutger ter Borg (rutger_at_[hidden])
Date: 2009-02-18 09:42:57

Jesse Perla wrote:
> Is there any chance that the argument ordering of gesv will
> change from "matrix, pivot vector, matrix" to "matrix, matrix, pivot
> vector" if you use auto-detection of the optimal pivot vector as an
> optional argument?

No, the ordering of arguments will remain fixed for gesv (and for all other
low-level driver/computational routines). For solve, on the other hand, it
will probably be different, more something like 

solve( A, B );
solve( A, B, pivots(i) );
solve( A, B, ... other optional args );

where the order of optional arguments does not matter.

Cheers, 

Rutger