abensonca
commented
2 years ago This is implemented by 3c52759cff5d339bb9d9897f39d2e2a72b97566f - it takes a slightly different approach by interfacing with the starFormationHistoryAdaptive class (or any other starFormationHistoryClass in principle) and then integrating over the relevant SSPs.
Original report by Andrew Benson (Bitbucket: abensonca, GitHub: abensonca).
Compute full SEDs by accumulating coefficients of SSP templates.
Computationally challenging:
In terms of memory can at least truncate the set of coefficients at each output so that they excluded ages greater than present age of universe (or, better, age of earliest node in branch of tree)
In terms of CPU, could be done quickly using MSBDF integrator - this gives a polynomial solution for stellar mass and metallicity through timestep. Possibly get an analytic solution to integral for each coefficient.