2009/03/07 The value of beta_prime_input and beta in miller32.in look a little bit off to me. The miller.f90 routine on miller.in2 says that for the parameters in Table 1 of the Waltz-Miller99 paper, the pressure gradients and alpha are related by d(beta)/drho = - alpha * 0.111828047244350 which for this case with alpha=1.22 gives d(beta)/drho = 0.1364 but miller32.in (and gs2_template.in) sets it to beta_prime_input = -0.113 which is small by a factor of 0.828. I wonder if I might have missed a factor of r/a=0.83 in the definition of beta_prime_input. Perhaps miller32.in is correct, and there is an error in miller.f90. I checked certain aspects of miller.f90, and it looks correct, so most likely the errors are in the value of beta_prime_input in miller32.in. If I pursue this again with GS2, I should check that this small difference in beta_prime_input doesn't make much difference in the results or the comparison with Waltz's table I. If I change beta_prime_input in miller32.in, I should also change the electron beta paramter: beta_prime = dbeta_tot/drho = beta_tot*(a/L_T+a/L_n) = beta_tot*4 = beta*8 so the electron beta is beta = beta_prime/8