I had not, thank you. :) It's also interesting how the behavior of that lab varies, between different J implementations.
I'll need some time to digest this. Thanks, -- Raul On Sat, Oct 5, 2013 at 10:04 AM, Aai <[email protected]> wrote: > Probably very obvious, but you never know: > > Did you try LAPACK from labs? > > While running I noticed: > > load 'math/lapack' > > load 'math/lapack/geev' > > examples with eigenvalues and vectors > > > > docs_jlapack_ '' NB. display list of lap files > ┌──────────┬──────────┬──────────┬──────────┬──────────┬ > │dlahqr.lap│zlaev2.lap│dlaed5.lap│ztgsja.lap│dgecon.lap│ > └──────────┴──────────┴──────────┴──────────┴──────────┴ > > > > > On 05-10-13 15:13, Raul Miller wrote: >> >> This is definitely a fun exercise, and I'm having fun playing with it >> (mostly looking at intermediate results and a few variations on the >> theme). >> >> That said, what I was really looking for was documentation or >> explorations on J's facilities (such as mt or lapack) for working with >> eigenvalues (and eigenvectors - in principle if you have the one the >> other is straightforward). >> >> Here, though, you did not use mt, nor did you use lapack. You started >> with a specific set of eigenvalues and demonstrated a context where >> you could plug them in. >> >> And I certainly appreciate the effort - and it does demonstrate some >> of the identities involving their use - but of course I am looking for >> more! >> >> Thanks, >> > > -- > Met vriendelijke groet, > @@i = Arie Groeneveld > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
