As it is well known the integral of 1/(cos(theta)^2) is tan(theta)
But there seems to be a problem evaluating the definite integral:
var("theta", domain="real")
> (1/cos(theta)^2).integral(theta, -pi/4 , pi/4 , algorithm='sympy')#
> Works
> (1/cos(theta)^2).integral(theta, -pi/4 , pi/4 , al
= x^2*diff(y,x,2)-x*diff(y,x,1)+y==0; eq
> h = desolve(eq, dvar=y, ivar=x); h
Output:
...
(_K2*log(x) + _K1)*x
So it's definitely something related to way in which boundary conditions
are handled.
Regards
On Sunday, October 30, 2016 at 6:41:21 PM UTC+1, Nils Bruin wrote:
>
&
I hope that's the right place to post the bug.
When solving this linear second order ODE boundary value problem:
> x^2*diff(y,x,2)-x*diff(y,x,1)+y == 0
> y(1)=1
> y(e)=2*e
Sage says Maxima can't solve it but when calling Maxima through Sage's
interface I get the right solution.
Steps to repro
I hope that's the right place to post the bug.
When solving this linear second order ODE boundary value problem:
> x^2*diff(y,x,2)-x*diff(y,x,1)+y == 0
> y(1)=1
> y(e)=2*e
Sage says Maxima can't solve it but when calling Maxima through Sage's
interface I get the right solution.
Steps to repro
