hi,

"Bald Eagle" <cre### [at] netscapenet> wrote:
> "jr" <cre### [at] gmailcom> wrote:
> > had a quick look, the page appears to give you the lot.
> Right, it looked solid enough to pursue, though it will require a lot of
> fiddling to adapt to SDL.
>
> > thinking you ought to have a look at the GSL (GNU Scientific Library); I just
> > found an index entry for singular value decomp.
>
> I got that set up, and now need to learn how to use such things.
> It appears that I have to write a c++ program and compile it, and then run it.

'C', not 'C++'.  ;-)  (in an ideal world, anyway)

I think cutting and pasting that c++ code into a single source and getting that
to compile, or at least spit out errors, would be a good exercise, then.

> Looks like this guy has a short program written, where we could just delete or
> comment out the Eigen stuff:
>
>
https://stackoverflow.com/questions/36959506/eigen-library-svd-is-slow-compared-to-gsl

I'll look at that later.  one thing at a time.

> If you can work something out that runs - tell me what format we need for the
> data input - CSV would be nice ;) and then we can see what this thing _does_.

first I suggest get that c++ to build, as exercise.

I did not read the article too closely but am under the impression that if one
re-assembles the functions in one source file, it'll build.  do you want to use
this code as a starting point?  if yes, then that will determine/influence the
data format.

> Can we run POV-Ray to #write trace() results to disk, and then call the c++ to
> write the SVD data to a file?  Can that c++ program then restart POV-Ray to use
> the SVD data? (have an if file exists block to switch between write data and
> read data)?

yes, in all likelihood _but_ how's your 'C', Bill?  or are you .. wedded to
'C++'?


regards, jr.

Post a reply to this message