Prime 6 new symbolic engine's bug

[ttokoro]

Prime 6 new symbolic engine has a bug for RMS and Average values evaluation. Without using step function, how to do to get the symbolic results?

[Werner_E]

I agree that what we see here is not "only" a disability, incapability, lack of ability but a series of severe bugs - not sure if all of them are related to Primes/friCAS inability to handle the integral of the Heaviside step function correctly.

The input is a definite integral and Prime returns an indefinite integral with the constant pi as "variable" of integration?

"2+6" is not simplified and the second integral (again indefinite) is missing the variable of integration completely!!

Its also noticeable that "simplify" or "simplify, max" have no effect on the outcome.

---

@ttokoro wrote:

Prime 6 new symbolic engine has a bug for RMS and Average values evaluation. Without using step function, how to do to get the symbolic results?

---

Thats simple: As long as its available use the old symbolics (muPad) in Prime and the next release of Prime will show just how PTC actually is at improving their new symbolic engine. But I guess that it sure would be a good idea to think about moving on. Maybe MatLab (with its much newer and capable muPad integrated)  is suitable for you. For harder symbolic tasks you may additionally use Maple. And yes, you will lose a lot by doing so as none of the aforementioned programs would provide the whiteboard interface of Mathcad or the natural way Mathcad handles units. But on the other side you may consider, that Mathcad(15) is a dead end and it really does not look like Prime would ever live up to it. The last "development" step with the integration of the new symbolic shows that PTC does not hesitate making steps backwards, too.

BTW, what was the reaction of PTC when you reported them the bugs in the new symbolics you have found so far?

For completeness sake for all who can't open a P6 file I add a screenshot which also shows the definition of your function f(t) and the correct results we get when we switch back to the legacy muPad:

[tslewis]

The webinar last week on prime 6 talks about new symbolic engine and provides future details of the engine and how to report bugs

[Werner_E]

---

@tslewis wrote:
The webinar last week on prime 6 talks about new symbolic engine and provides future details of the engine and how to report bugs

---

Personally I dislike that kind of "talks", videos, etc. and prefer clearly written information.

So the question is - where can we read about those details of friCAS and how bugs are quick and easy to report by anyone? Can you give details about the official sources where we can read it up?

[-MFra-]

So what WIKI says about FriCAS is not true ....

(https://en.wikipedia.org/wiki/List_of_computer_algebra_systems)

[Werner_E]

---

@-MFra- wrote:

So what WIKI says about FriCAS is not true ....

(https://en.wikipedia.org/wiki/List_of_computer_algebra_systems)

---

We don't know if the problem here is FriCAS itself or if its the way its migrated in Prime. I neverused FriCAS and have no installation of it running here, so I can't verify. As far as I am aware FriCAS is best used with a Linux derivate as OS. The windows ports seem to be rather old.

And when Wiki says that FriCAS is "strong at symbolic integration", it should state in comparison with which other programs. muPad never was famous for being aspecially strong at integrals. I remember that MC14 had a lot of severe bugs in that respect and not all of them were fixed with MC15.

While FriCAS in Prime so far could not convince and is a bit disappointing, I guess that there are some integrals which muPad handled wrong and FriCAS calculates correct.