I found out that the output for Method A changed over time. I ran the bottleneck demo with an old jpsreport version (part of jpscore commit a2da3383d086d940e4e0b77f3fb1c18e21dc5dca) and the new one in the standalone repo. The output for the mean velocity differs and seems not valid for the newest jpsreport version.
My output data:
bottleneck_Output.zipbottleneck_Output_new.zip
New tests are needed, since this change was not noticed.
However, the mean velocity results calculated with the old jpsreport version should be revised again. I am not sure if it works as expected since negative and unrealistic high velocities are calculated.
I found out that the output for Method A changed over time. I ran the bottleneck demo with an old jpsreport version (part of jpscore commit a2da3383d086d940e4e0b77f3fb1c18e21dc5dca) and the new one in the standalone repo. The output for the mean velocity differs and seems not valid for the newest jpsreport version. My output data: bottleneck_Output.zip bottleneck_Output_new.zip
New tests are needed, since this change was not noticed.
However, the mean velocity results calculated with the old jpsreport version should be revised again. I am not sure if it works as expected since negative and unrealistic high velocities are calculated.