The first is definitely possible. Just add data_source=extractCube to your call to add_projection or add_slice or whatever. The second probably won't work because the projection or slice routine will look for other things like cell dxs that won't be present from just a similarly sized array. If you want to do something like that, you're probably better off making a new field and then projecting that with the above data_source keyword argument in there.
Is it possible to make a slice or projection of a extracted, fixed resolution region? i.e. the resultant object of:
extractCube = pf.h.covering_grid(extract_level,
# How many dimensions along each axis
# And any fields to preload (this is optional!)
num_ghost_zones = 3)
Or of an array later made to the same size? e.g.
cloud = na.zeros(extractCube["x-velocity"].shape, dtype='float64')
yt-users mailing list