In article <874n0xvd85....@dpt-info.u-strasbg.fr>, Alain Ketterlin <al...@dpt-info.u-strasbg.fr> wrote: >alb...@spenarnc.xs4all.nl (Albert van der Horst) writes: > >[...] >> Now on some matrices the assert triggers, meaning that nom is zero. >> How can that ever happen? mon start out as 1. and gets multiplied > >[several times] > >> with a number that is asserted to be not zero. > >Finite precision. Try: 1.*1e-162*1e-162. Equals zero. > >-- Alain.
Thanks you Alan and all others to point this out. That was indeed the problem. Somehow I expected that floating underflow would raise an exception, so I had a blind spot there. Groetjes Albert -- Albert van der Horst, UTRECHT,THE NETHERLANDS Economic growth -- being exponential -- ultimately falters. albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst -- https://mail.python.org/mailman/listinfo/python-list