The integrand is real, except at where the denominator is zero. An antiderivative that is real for all such values of is
The rationalizing substitution can be used to obtain this antiderivative.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
After completing the square in the denominator and applying a partial fraction decomposition, the antiderivative is given in terms of logarithms. Alternatively, after completing the square in the denominator, an antiderivative of from Table 3.10.1 is invoked.
|
|
|
|
|
|
|
|
Consequently, the alternative antiderivative is
which is real except at those points where it is undefined.
Figure 6.6.3(a) is a graph of and the line (in green). Where , the graph is drawn in black; where , it is drawn in red.
>
|
f:=proc(s)
local C,CC;
C := (3/13)*sqrt(13)*abs(tan((1/2)*x)-2/3);
CC:=evalf(eval(C,x=s));
if CC<1 then CC else 10;end if;
end proc:
|
>
|
g:=proc(s)
local C,CC;
C := (3/13)*sqrt(13)*abs(tan((1/2)*x)-2/3);
CC:=evalf(eval(C,x=s));
if CC>1 then CC else -10;end if;
end proc:
|
>
|
p1:=plot(f,-9..9,0..3,discont=true,color=black):
p2:=plot(g,-9..9,0..3,discont=true,color=red):
p3:=plot(1,x=-9..9,color=green):
plots:-display(p1,p2,p3,scaling=constrained);
|
|
Figure 6.6.3(a) in black; in red
|
|
|