Hi Nathan,

This could be an issue with how you set up your covering_grid or smoothed_covering_grid.  Can you try out this script?
This depends on the script that you included in your post, so make sure it's in the same directory.  This will add up the number of cells in your grid object for which the dPhidx field is zero.


On Fri, Jun 29, 2012 at 3:19 PM, Nathan Butcher <butchernate@gmail.com> wrote:

I am having trouble using a derived field with a CoveringGrid. I am making fields for dPhi/dx, dPhi/dy, dPhi/dz using the PotentialField from ENZO. The derivative fields do not get passed into the CoveringGrid and get recalculated. This leads to the derivative being zero in areas that were originally in coarse cells.

Density and DivV fields were passed to the CoveringGrid properly. As a test I defined a field that was Density^2 and that was passed to the CoveringGrid properly, so I think it is a problem specific to derived fields based on the ENZO-specific PotentialField. I have included the source for my derivative functions.

Thank you,
Nathan Butcher

yt-users mailing list