This module provides the initial power spectra
TInitialPowerLaw is parameterized as an expansion in ln k
ln P_s = ln A_s + (n_s -1)*ln(k/k_0_scalar) + n_{run}/2 * ln(k/k_0_scalar)^2 + n_{runrun}/6 * ln(k/k_0_scalar)^3
so if n_{run} = 0, n_{runrun}=0
P_s = A_s (k/k_0_scalar)^(n_s-1)
for the scalar spectrum, when n_s=an(in) is the in'th spectral index. k_0_scalar
is a pivot scale, fixed here to 0.05/Mpc (change it below as desired or via .ini file).
The tensor spectrum has three different supported parameterizations giving
ln P_t = ln A_t + n_t*ln(k/k_0_tensor) + n_{t,run}/2 * ln(k/k_0_tensor)^2
tensor_parameterization==tensor_param_indeptilt (=1) (default, same as CAMB pre-April 2014)
A_t = r A_s
tensor_parameterization==tensor_param_rpivot (=2)
A_t = r P_s(k_0_tensor)
tensor_parameterization==tensor_param_AT (=3)
A_t = tensor_amp
The absolute normalization of the Cls is unimportant here, but the relative ratio
of the tensor and scalar Cls generated with this module will be correct for general models