make 32bit float (float) versions of all matrices

[GoogleCodeExporter]

I have an application that runs on an embedded device, for performance reasons 
I would like to use 32bit float instead of double.

I also have to use several other libraries that only accept 32bit float. 
Iterating over a matrix data array and casting each 64bit float to 32bit float 
is hurting performance.

Is it possible to create an 32bit float matrix & operations?

Original issue reported on code.google.com by DeRijcke.Erik@gmail.com on 13 Jun 2014 at 1:48

[GoogleCodeExporter]

I've toyed with the idea before.  For another project (GeoRegression) I 
developed it such that it was easy to autogenerate 32bit float version from the 
64bit code.  However, EJML wasn't written that way from the start so it would 
take a bit more effort.

Maybe if there is enough interest I'll give it a shot.

Original comment by peter.ab...@gmail.com on 15 Jun 2014 at 3:48