I am computing a switch to choose a criteria depending on a value. In doing it, I need to compute a value like below

(-real number)**(- or + real number)..

so I used the usual syntax "**" to consider the power as the pow in c++.

But it gave me a NaN value. Please let me know how to treat such a power in fortran.

Thanks in advance.

The problem is occurred when I tried to compute following

-0.24987**0.3846 --> it's okay but when I tried to use

r**GammaK --> it cause NaN

Of course, r is a value containing -0.24978 and Gammak has something.

1) -0.24987**0.3846 is evaluated as -(0.24987**0.3846) because of the operator precedence in Fortran

2) r**GammaK is not a real number for r<0 and GammaK real. Check your formulas!

Thanks. You're right. It's a complex number.