[yt-users] create_profile of linear fields

[Douglas Harvey Rudd]

Hi all,

Is there a way to specify the log/non-log scaling of a y-axis variable in create_profile or ProfilePlot.from_profiles?  I have some profiles that start flat at an initial time and so the velocity_magnitude is everywhere 0.  create_profile succeeds, but ProfilePlot.from_profiles generates a Matplotlib error before I have the ability to specify a scaling:

    sph = ds.sphere(ds.domain_center, ds.domain_width[0]/2)
    profile = yt.create_profile(sph, "radius",
                                ["density", "temperature", "velocity_magnitude", "dx"],
                                extrema={'radius':(ds.domain_width[0]/ds.num_grid,ds.domain_width[0]/2.)},
                                n_bins=int(ds.domain_width[0]*2.**(ds.max_level-1)),
                                logs={'radius':False},
                                units={'radius':'pc'})
    plot = yt.ProfilePlot.from_profiles(profile)

Relevant part of the dump:
/software/python-2.7-2014q3-el6-x86_64/lib/python2.7/site-packages/matplotlib-1.4.0-py2.7-linux-x86_64.egg/matplotlib/ticker.pyc in view_limits(self, vmin, vmax)
   1523         if minpos <= 0 or not np.isfinite(minpos):
   1524             raise ValueError(
-> 1525                 "Data has no positive values, and therefore can not be "
   1526                 "log-scaled.")
   1527


Douglas Rudd
Scientific Computing Consultant
Research Computing Center
drudd at uchicago.edu