Bug introduced in 12.0.0

I'm having some problems with ListDensityPlot in Mathematica 12 on MacOS Mojave.

Here is the code:

test = Import["https://www.dropbox.com/s/0ypdo4pdh9q86j5/test.csv?dl=1"];
ListDensityPlot[test, PlotRange -> All, InterpolationOrder -> 0]

(data can be found here: https://www.dropbox.com/s/0ypdo4pdh9q86j5/test.csv?dl=0)

Output in MMA11.3

enter image description here

Output in MMA 12

enter image description here

In v12 there are a lot of unwanted triangles in the plot. Any idea of what the problem might be?

EDIT: I forgot to mention that the problem "disappears" with any value different than 0 assigned to InterpolationOrder.

EDIT 2: It actually seems to be a genuine bug, so I added the bugs tag and reported to mathematica support.

  • $\begingroup$ Can confirm on 64bit Windows 10 (Version 1809) box. $\endgroup$ Apr 18, 2019 at 14:32
  • $\begingroup$ InterpolationOrder -> 1 seems to give the same (correct?) output for 10.4, 11.3, and 12 (Windows 10). $\endgroup$
    – JimB
    Apr 18, 2019 at 15:36
  • $\begingroup$ @JimB yes, any interpolation value != 0 seems to work fine as I mentioned in the edit. $\endgroup$
    – Fraccalo
    Apr 18, 2019 at 15:39
  • $\begingroup$ Sorry. I went to have breakfast and didn't notice the update. I'll delete my comment shortly. Overlaying the (x,y) points shows the triangles occurring where the points are. Don't know if that adds anything or not. $\endgroup$
    – JimB
    Apr 18, 2019 at 15:45
  • 1
    $\begingroup$ @Kai pls report to Wolfram so that they might address the problem sooner $\endgroup$
    – Fraccalo
    Feb 15, 2020 at 15:50

1 Answer 1


A workaround that did the job sufficiently well for me was to overlay the plots with InterpolationOrder -> 1 and InterpolationOrder -> 0 using Show[]:

Show[ListDensityPlot[test, PlotRange -> All, InterpolationOrder -> 1],ListDensityPlot[test, PlotRange -> All, InterpolationOrder -> 0]]

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.