The changes here add another axis type that is similar to axis.logarithmic, only it treats negative values instead of positive. That is especially useful for plotting density data, as is done in the example in test/experimental/test_neglogaxis.py
The example shows that there is still work to do, because some other part of the axis partitioning gets the minimum/maximum of the axis wrong. I could use some input on where to search for the place where this goes wrong.
An alternative implementation of a neglog axis has been implemented upstream. Your test_neglogaxis.py example is contained in f171b59a3cc673b142ce3babdda0322528616855
The changes here add another axis type that is similar to axis.logarithmic, only it treats negative values instead of positive. That is especially useful for plotting density data, as is done in the example in test/experimental/test_neglogaxis.py
The example shows that there is still work to do, because some other part of the axis partitioning gets the minimum/maximum of the axis wrong. I could use some input on where to search for the place where this goes wrong.