Hi Andrew,
Related to this, I'd also like to query the group for thoughts the best approach to a specific kind of analysis I have in mind with this.
I'd like to know what the azimuthally averaged radiation flux is at some fixed distance r from the star plotted as a function of polar angle.
My plan is to create a radiation flux derived quantity (which is why I need a derivative), then create a flux dot rhat quantity and then extract to a fixed interpolated grid and then proceed with the extracting, averaging, plotting.
I think that definitely generating the radiation flux in the manner you describe will work. I'm not sure that you need to extract to a fixed interpolated grid -- I think you should be able to generate the radiation flux inside a given cell, if you presuppose that the ghost zones are being generated correctly. You can manually verify that the ghost zones are looking like you expect them to if you look at the output of grid.retrieve_ghost_zones( ... ). This should perform a cascading interpolation that fills in the boundary values. More complicated calculations of the flux, through isocontours or the like, would be a bit harder but doable using the clump object. But I think for what you're looking at, this should do it.
Does that sound like the correct approach or is there some yt functionality that I don't yet know about that might be more elegant? - ie) some way to extract the field on a spherical count our and return that as a list of values and positions.
I believe that you should be able to use something like the field DiskAngle. If you look at the source, you can see that it grabs the field_parameter s height_vector and center, which you can set on an object with set_field_parameter. This then calculates a declination from that -- I think you may be able to use this as input to either a 1D or 2D profile, to get the average value of any arbitrary field, in this case your newly defined flux field versus the declination from the axis. Let me know if I'm crazy, or if this'll do what you're looking for! Best, Matt