plot: bug or ?



Hi all,

I have :

define(bx(xn,t),sum(sum(bk[n,m]*bb(n,m,chnod(xn),t),n,1,np),m,1,NT))$

here xn is variable on radius and t - on latitude, chnod transforms from 
[-1,1] to custom interval,  bk[n,m] are components of eigen vector,
and bb(n,m,chnod(xn),t) are basis functions constructed from Chebyshev and 
Legendre polynomials.

I want to see the time evolution patern for eigen solution for a particular 
level of convection zone:

F(th,ti):=realpart(expand(bx(.92,cos(th))*exp(%i*ti)));
Note, that
(%i202) F(%pi/4,%pi/4);
Evaluation took 0.01 seconds (0.01 elapsed) using 336.305 KB.
(%o202) 		      [.1803733328091562]

However,

(%i214) contour_plot(F(x,y),[x,0,%pi],[y,0,3*%pi]);
		 2
Is  1.0 - 1.0 cos (x)  positive or negative?

;
Incorrect syntax: Premature termination of input at ;.

What can be a problem here?

best regards
V