what syntax should we support?

[zingale]

We want to make something inspired by gnuplot to make quick exploratory plots of our AMR data, building from yt.

a sample session:

plot 'plt00000' using "temp"
set logscale
set xrange [0:10]
replot
save "figure.png"

most of these have analogs to gnuplot, except save is much simpler than setting a terminal, output, and replotting as we need to do in gnuplot.

some other bits:

• what syntax do we use for making 1-d slices?
• for 3-d, is it assumed that we always want to plot the three slice planes? should we have a set origin (x,y,z)?

[adam-m-jcbs]

what syntax do we use for making 1-d slices?

Maybe profile? or some variation of that.

for 3-d, is it assumed that we always want to plot the three slice planes? should we have a set origin (x,y,z)?

I find that even with 3-d I tend to want just a single plane slice. Three slice could be an option, but my vote would be for a single slice for both 2- and 3-d. For the origin, I'd default to the simulation's coordinates but have the option to set the origin. For example, in full star simulations the origin is usually the corner of the box but the star's center is the center of the domain. In these cases it might be nice to make plots with a custom origin at the domain's center.