Jesse Hopkins

Think I got it figured out. Was missing MSVCP100.dll. I installed Microsoft Visual C++ 2010 SP1 Redistributable from here: http://www.microsoft.com/download/en/confirmation.aspx?id=8328

I found an interesting thread on the topic of static vs dynamic linking of these libraries here: http://social.msdn.microsoft.com/Forums/en-US/Vsexpressvc/thread/3a007184-80e9-4e25-b5ad-ff31b028c051

Everything went pretty smooth, got some good looking documentation. I did however, manage to get a Grrr! message :). Here's the output:

MTOCPP:C:/HopkinsSynergySandbox/CSI/CSI_CORE/0.0/jhopkin/CsiPlottingTool/@AmsPlotManager/AmsPlotManager.m: PARSE ERROR in line 27
Message: Grrrr!!!!
Next 20 characters to parse:
% Matlab class for c

Patrick and Ian, take a look at CompressLib on the file exchange. Using this you can serialize almost any matlab structure to an array of bytes, and deserialize it back to a matlab structure on the other side. You can do classes as well, although that takes some work in the class definition itself.

Thanks for the suggestion Jan, I'll take a look into that.

Thanks for the suggestion. I have tried it with my version (which chunks the data in 5 MiB blocks to prevent running out of heap space), but the speed-up is not measurable. Compressing ~120 MiB takes 0.035 seconds for the custom typecast and 0.161 for the built-in function. In contrast, the java GZIP function takes 7 seconds (and uses only 1 thread).

However, sharing pointers is of course a much nicer solution.

Using James Tursa's TYPECASTX will increase the speed, because it creates shared data copies instead of deep copies: http://www.mathworks.com/matlabcentral/fileexchange/17476-typecast-and-typecastx-c-mex-functions