Michael.Soegtrop

Since this is not an issue with marginal precision I would say that this is simply a bug which should be fixed. I don't see why such a computation should fail with floats.

Wrong result of integral with ",float" - analytic result is correct

I find it increasingly hard to find the official source repo, download page and bug reporting site for wxMaxima. A few confusing points: The wxbug_report() function mentions https://github.com/wxMaxima-developers/wxmaxima/issues?direction=desc&sort=created&state=open The discussion here on sourceforge seems to be fairly active and also used by wxMaxima key developers. The last updated date on https://sourceforge.net/projects/wxmaxima/ is September 2017. The download page for wxMaxima seems to be...

VOTE: maxima

Documentation of scalarmatrixp in Dot

If you have something I can test, please let me know.

In maxima-discuss (subject: Should minimize_lp (simplex LP algorithm) with rational coefficients be exact / optimal) Leo Butler suggested to check case by case if the numbers compared with epsilon_lp are rational numbers or floating point numbers. He also suggested to issue a warning instead of avoiding divisions by small floating point numbers. There are definitely cases where this would lead to better results, but I guess this test is in there because otherwise other cases would lead to issues....

Documentation: mention parameter epsilon_lp more explicitly in documentation for minimize_lp