Maxima gives wrong antiderivative for trig with possible zero coefficient

[kcrisman]

Is there an existing issue for this?

• I have searched the existing issues for a bug report that matches the one I want to file, without success.

Did you read the documentation and troubleshoot guide?

• I have read the documentation and troubleshoot guide

Environment

All

Steps To Reproduce

No response

Expected Behavior

As reported here and here:

var('n t')
assume(n, 'integer')
integrate(cos(2 * pi * n * t), t, 0, 1)

should be one if n=0.

Actual Behavior

Instead, we get zero, because Maxima uses a generic antiderivative:

integrate(cos(2 * %pi * n * t),t);
sin(2*%pi*n*t)/(2*%pi*n)

Additional Information

As seen in the cell server, this has nothing to do with declaration, it's a pure Maxima issue, reported upstream at Bug 4125.

Note that Sympy does this correctly:

var('n t')
integrate(cos(2 * pi * n * t), t, algorithm='sympy')

cases(((pi*n != 0, 1/2*sin(2*pi*n*t)/(pi*n)), (1, t)))

but for the definite integral Sage can't interpret the case structure, I guess, which is a separate issue.

[kcrisman]

PS: the process to create this bug was kind of onerous for those of us used to the Trac form. Is there a way to not have to fill in all these form boxes and instead just report, you know, the bug? I agree it's helpful for new folk, but it's somewhat time-consuming.

[kcrisman]

Update: Maxima devs don't consider the antiderivative a bug (removable discontinuity) but the integer assumption issue stands, I would think.

[kcrisman]

BTW, is there an equivalent to the Trac upstream response reports? I couldn't find labels for that.