robust reconstruction of grain size distributions from the nucleation moments

[danieljprice]

For post-processing of phantom snapshots, we need a robust procedure for reconstructing a grain size distribution from the dust nucleation moments (K_0, K_1, K_2, K_3 as per Siess et al. 2022). We have some code for this but the procedure is currently not robust and needs to work reliably under all circumstances.

[danieljprice]

see reconstruct-moments branch