Solving Equations

[TorbjoernAndrea]

The problem arises with the term cos(30 deg) in an ordinary equation. F.ex: 6^2=10^2+x^2-2*10*x*cos(30deg) solve.... The equation becomes red with no solution proposed. See attached file.

[RichardJ]

It's not a bug, just a limitation of the symbolic processor. The symbolic processor knows nothing about units, it just sees them as another variable. So writing cos(30*deg) is the same as writing cos(30*xxx). Your various uses of the numeric processor get round this, because the numeric processor does understand units. Richard

[TomGutman]

Just define deg as pi over 180. This is redundant for the numeric processor, but will properly define deg for the symbolic processor.

[RichardJ]

"Tom Gutman" wrote:

Just define deg as pi over 180. This is redundant for the numeric processor, but will properly define deg for the symbolic processor.

[TomGutman]

I've just looked at the sheet a bit more carefully. It demonstrates why it is important to include the exact error message when talking about an error. There is not actually an error. There is just a warning message, and a refusal to display the result that has been calculated, as it is too large. Add the simplify keyword and a symbolic result is shown (at least in MC14, not sure what your version is). And the later part of the sheet demonstrates that a symbolic value was found and that it can be evaluated numerically. Due to quota considerations I'm not posting a modified worksheet, it's easy enough to add the simplify keyword on your own.

[TomGutman]

For such a trivial problem this is developing a lot of ramifications. I just tried this in MC13. As written all the solves fail as MC13 requires a specification of the variable for which to solve (it doesn't guess). But with the addition of that specification, they all work. Turns out that in MC13 deg is defined to the symbolic processor. That makes the MC14 behaviour a bug, as the MC13 feature got lost in the shuffle.

[RichardJ]

"Tom Gutman" wrote:

Turns out that in MC13 deg is defined to the symbolic processor.

[TomGutman]

Ouch. MC13 symbolics does not know about the rad. It only thinks it does. It has a definition for it, but it is quite wrong. Try using it to calculate the sine of pi radians.

[RichardJ]

!! I just typed a random number in, and didn't notice the answer was wrong. Good thing I didn't (and still don't) use MC13 for much! Richard