I've investigated bug #45925 "Incomplete Gamma Functions flipped?" and I believe there is in fact no problem with gsl. The bug author compares the (upper) incomplete gamma function with the function P(a,x) and correctly notes that the values disagree. However the comparison should actually be done with the function Q(a,x) rather than P(a,x), and here the values agree. The definitions for the upper incomplete gamma function and Q(a,x) are contained in the gsl reference document under the section on "Incomplete Gamma Functions". I recommend that no code changes be made and that this bug be closed.
I've never worked on gsl before and am not sure what I should do to get this bug assigned to me or closed out. Could someone let me know what I need to do? -Bill Maier
