[dune-functions] [Dune-functions-commit] [Commit] dune-functions - c7b3989: Resize must now be done manually before interpolate()

[Carsten Gräser]

Hi,

Am 15.09.2015 um 10:30 schrieb Carsten Gräser:
> Am 15.09.2015 um 01:28 schrieb Carsten Gräser:
>> OK, here's a proposal to allow nested vectors while keeping
>> simple things simple:
>>
>> a)We derive MultiIndex<class size_type, size_type i> from e.g.
>>   array and implement the cast to size_type for i=1.
> This would be against our decision to not fix the MultiIndex
> implementation. Hence I modify my proposal: We can simply
> provide FlatMultiIndex<size_type> providing 'operator[]'
> and 'operator size_type'.
since we did explicitly not fix the type of the multiindex
using such a FlatMultiIndex is not an interface change.
Hence I implemented it right away. It's now used in the
Poisson example.

>> b)The result of basis.size() without prefix is no longer
>>   the same as the one with an empty prefix, but an object
>>   providing operator()(Prefix) and a cast to size_type.
>>   In the trivial case the result of this cast should be
>>   either the same as for an empty prefix or as dimension.
I also added a helper class SizeInfo, that allows what I
had in mind by calling sizeInfo(basis). The proposal is
to directly return this (or an implementation defined
type having the same interface) on basis.size().

With the current interface iteration all the backend
stuff was essentially moved to a single wrapper. The
latter provides a sensible default for istl and std
containers and is only needed for the non-trivial
case of non-flat multi-indices.

Next step is to use the machinery to implement
DiscreteFooGlobalBasisFunction with Foo='' instead
of 'Scalar'.

Best,
Carsten

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: OpenPGP digital signature
URL: <https://lists.dune-project.org/pipermail/dune-functions/attachments/20150915/9fb7fd1a/attachment.sig>