On Mon, Jan 30, 2006 at 06:59:04AM +0800, John Darrington wrote: > It might be worth re-drafting that proposal, since we've moved on a > bit in the last 12 months anyway. I'm not sure what Jason's plans are > with lib/linreg --- I'd assumed that lib/linreg was supposed to be a > general purpose linear regression library which could be applied to > any software. But I note that there are some very PSPP specific > things there now (eg: #include "var.h").
I originally meant lib/linreg to be a general purpose linear regression library, ignorant of variable structures and other features specific to PSPP. But that approach made regression.q too complicated, so I made the linreg library aware of variable/value structure and xnmalloc. That makes coding other procedures that use linreg easier. The approach also makes linreg more dependent on PSPP, but only through the variable/value structures and xnmalloc. I intend to use linreg in other procedures, too. I'm not sure where it belongs in the new directory tree. Maybe there should be a 'statlib' library with subdirectories for statistical libraries; something eventually designed to make contributing easier for statistical programmers who do not know about PSPP internals. > > I take it your plan would be to have each directory compiled as a > static library, and linked into the binary at the end? And what about > header file search paths? Currently, *every* file in *every* > directory searches *every* directory for it's *.h #includes. IMHO > this is wrong, and would be cumbersome for such a large number of > directories. So this would need to be thought about. Yes, I thought each statistical library should have its own directory, then be linked statically to make it available to multiple procedures. And I agree with John's comment about the cumbersome searching through all the different directories. -Jason _______________________________________________ pspp-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/pspp-dev
