This is pretty much the same problem as this question, however, that question has been dead in the water and I don't have the reputation to revive it in the comments, yet still find myself needing to resolve the problem.
The idea is that when using peakutils to detect peaks, it ignores peaks where it plateaus (even for only two consecutive points). Has anyone created a workaround for this?
I've considered just looping through the entire dataset and moving all instances beyond the first of matching numbers down a hair to make them uneven, but that's hacky and might only work for this specific case as I'm dealing with long floats that won't matter if there is a change as small as 0.00000001.
An example of where it'll fail is [1, 2, 3, 5, 5, 1, 2] where 5 will not be detected as it is equal to a point next to it.
It might be worth mentioning that I found a bitbucket reference to this problem, where it is noted as a feature worth including but with no intention to do so yet (as of 2015).
In working with the repo owner I discovered that the module was not up to date. I made the mistake as a beginner to python of assuming that imports were always the most recent version, a simple version update resolved this issue as the module has included this feature in version 1.1.0