The
radius is
R=sup{r≥0:∑anzn converges whenever ∣z∣<r}; equivalently
R−1=limsup∣an∣1/n. The same
definition applies over
R and
C.
Here a2n=(−1)n and a2n+1=0, so ∑anzn=∑(−z2)n=(1+z2)−1=((z−i)(z+i))−1 for ∣z∣<1.
Along real
z=x↑1, the
modulus is
1/(1+x2) and stays bounded, but along
z=iy with
y↑1 it is
1/(1−y2)→∞. Since
limsup∣an∣1/n=1,
R=1.
No. For example
∑n≥0(−1)nxn=(1+x)−1 has
radius 1 and is bounded by
1 on
(0,1).