simplifying integral definition

Hi,
following the issue [1], the dimension in the integral name (e.g. 'gauss_o2_d2') is now useless. Moreover, we do not have any other kind of quadrature than Gaussian(-like). So what about leaving just the order, that is, instead of
integrals = {'i1' : ('v', 'gauss_o2_d3')}
we would have
integrals = {'i1' : ('v', 2)}
What do you think? r.

Hi,
On 07/12/2012 05:35 PM, Robert Cimrman wrote:
Hi,
following the issue [1], the dimension in the integral name (e.g. 'gauss_o2_d2') is now useless. Moreover, we do not have any other kind of quadrature than Gaussian(-like). So what about leaving just the order, that is, instead of
integrals = {'i1' : ('v', 'gauss_o2_d3')}
we would have
integrals = {'i1' : ('v', 2)}
Good idea, it is simple and it corresponds to the integral order definition via terms.
vl
What do you think? r.

On 07/12/2012 07:41 PM, Vladimír Lukeš wrote:
Hi,
On 07/12/2012 05:35 PM, Robert Cimrman wrote:
Hi,
following the issue [1], the dimension in the integral name (e.g. 'gauss_o2_d2') is now useless. Moreover, we do not have any other kind of quadrature than Gaussian(-like). So what about leaving just the order, that is, instead of
integrals = {'i1' : ('v', 'gauss_o2_d3')}
we would have
integrals = {'i1' : ('v', 2)}
Good idea, it is simple and it corresponds to the integral order definition via terms.
Done!
r.
vl
What do you think? r.
participants (2)
-
Robert Cimrman
-
Vladimír Lukeš