Sorry to beat this into the ground, but I think the actual integral is ...
(1/T) * -cos(wt)/w evaluated from 180 to 45 or pi to pi/4.
From my understanding the relationships are as follows ...
T =1/f, w=2*pi*f, therefore w=2*pi/T
this should cancel out the T and leave the result as ... -cos(wt)/(2*pi) which is roughly half of the solution they provide.
Seems to me that the average value looks more appropriate at approximately 0.6, but would like to know for future reference whether modifying the sine wave does away with the basic frequency, angular frequency, and period rules. I do agree that the pi is the period but they should state somewhere that w (omega) is 1 rad/s.