diff --suppress-common-lines -d --speed-large-files --suppress-blank-empty /usr/share/octave/packages/3.2/statistics-1.0.10/regress.m regress.m
1a2 > ## Nir Krakauer, 2011: Modified to avoid computing X * pinv(X), which will > lead to running out of memory when the number of data points is around the > square root of the memory size 127c128 < H = X * pinv_X; --- > # H = X * pinv_X; #commented out -- Nir 129c130,131 < r = (eye (n) - H) * y; --- > # r = (eye (n) - H) * y; #commented out -- Nir > r = y - X * b; #added -- Nir 146c148,149 < h = diag (H); --- > # h = diag (H); #commented out -- Nir > h = sum(X.*pinv_X', 2); #added -- Nir (same as diag(X*pinv_X), without > doing the matrix multiply) On Sun, Feb 13, 2011 at 2:46 PM, Søren Hauberg <so...@hauberg.org> wrote: > > Thanks for looking into this. Could you send a diff with your changes as > that makes it easier to inspect the changes you suggest? > > Thanks > Søren > > ------------------------------------------------------------------------------ The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb _______________________________________________ Octave-dev mailing list Octave-dev@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/octave-dev
