13 Oct 2008 tampe   » (Journeyer)

And I tried so hard to be academic

Two more test cases added. Handling of sparse tensors and handling of Blocked tensors. Operations on block matrices can be faster for matrix local operations because you use the cache more efficiently. and sparse matrices is a must to have when doing CFD where you cannot store a full matrix. Doing operations on these creatures such as doing different marginals and permutations carries over just fine. There is some refinement in the logic though. my g.. operator has to be modified for the sparse case to store internally an index tree. One sparse representation can be based on lists of pairs of an index and a value. when you encounter a pair you look into your tree to try to find the correct generator if not there you need to recreate it. Else the logic is the same. For blocked matrices you need to consider blocked symbols coll.X.Y which will use a generator of generators generating generators. Permuting blocked sparse matrices are now a dead simple problem and a large set of basic matrix manipulations becomes trivial using this. Also add the .f modifier, eg sum.X.Y.f will mean that the final value of (sum.X.Y ..) is returned at the calculation. It is not available of cause, but again delayed variables come to rescue and everything should work just fine. Wow maybe academic play just turned very practical.

Good Night

Latest blog entries     Older blog entries

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!