The linked code drops the non-logarithmic column and renames the logarithmic ones. If someone is testing different distributions (logarithmic and non-logarithmic), the non-logarithmic distributions will be lost. I suggest that we short term skip dropping the non-logarithmic even though it creates some duplicates. Could be better to e.g. only keep the non-logarithmic which exists independent of chosen distribution, but instead have an option in the gui to use a logarithmic axis.