There is no more *.grid* file. That was for Enzo 1.0 if I am not wrong. You probably meant the *.cpu* files, right ?
That's correct.
To go further, is there a way I can:
- restart Enzo with WritePotential = 1 using a file that had WritePotential = 0 ? - restart Enzo and use Tracer Particles with a file that did not have any of them ?
Both are possible, but you have to rewrite the datasets as well a the supporting meta data (boundary and hierarchy). I don't know the details about writing tracer particles, but very recently I wrote a script to add the potential field to a dataset. Unfortunately, the script was for my version of Enzo, which doesn't use packed AMR and has extra fields for my MHD implementation. I've attached a draft of a code that will do the same thing for the more recent Enzo 1.5 with the PackedAMR version, though I don't have time right now to do much more. Basically, you need to have a GravPotential field in the hdf5 files, it needs to have a label in the parameter file, it needs to have it's field_type listed and be counted in NumberOfBaryonFields in the hierarchy file and the boundary file, and (if you're not using OutOfCore Boundary) needs to have space in the hdf5 boundary file. There are a lot of things to get right. Alternatively, you could start a simulation with the desired size and processor layout, and use h5py to write the density, energy, and velocity fields into that. That will ensure the hierarchy and boundary are constructed properly. Easiest still would be to just start the initial relax sims with the gravitational potential written, that will have everything set up right. If anyone out there has an easier solution, that would be sweet. d.
Thanks again for your help,
JC
-- Sent from my Stone Tablet and carried by my Pterodactyl.