
Re: An independent integration test suite
Posted:
Jul 16, 2013 4:05 AM


On Tuesday, July 16, 2013 4:00:13 AM UTC4, da...@axiomdeveloper.org wrote: > On Tuesday, July 16, 2013 2:19:26 AM UTC4, Albert Rich wrote: > > > On Monday, July 15, 2013 9:36:04 AM UTC10, da...@axiomdeveloper.org wrote: > > > > > > > > > > > > > In order to ensure that the answers of the integration differ by no > > > > > > > more than a constant I've been differencing the expected answer from > > > > > > > the Axiom answer and then taking the derivative. > > > > > > > > > > > > > > One curious pattern is that your answers differ from Axiom's answers > > > > > > > by nonzero constants. [...] > > > > > > > > > > > > The first sentence above correctly asserts that it is ok for antiderivatives to differ by a constant. Yet, the second sentence finds it surprising that they do differ. So what is the problem? > > > > > > > > > > > > Albert > > > > suppose > > t0:= expression > > r0:= expected result > > a0:= integrate(t0,x) > > m0:= a0  r0 > > d0:= differentiate(m0,x) > > > > m0 is the difference between Axiom's result and the expected result. > > d0 is the derivative of m0, usually with a value of 0. > > > > m0 often shows that Axiom's result and the expected result differ > > and the derivative result of 0 shows that this is just a constant. > > > > When I look at the reason for the constant difference it seems to be > > related to the trig identities we chose. What system did you use to > > create the expected results?
That last was just a dumb question... you used Rubi, no doubt. What I wanted to ask was what trig substitutions you use. Is there somewhere in the Rubi sources I should look?

