compute_stats issue- actions with same start times, different end times
Found this issue relative to the adults2018 data corpus, path 0 (person 1, task 7), states 1 and 2 in the path. There should be an additional state between states 1 and 2 (first, place the red rectangle, then place the yellow rectangle). These two placements have the same start time, but different end times.
How does the compute_stats function currently handle the differentiation of states? Is this issue pervasive across all the data (actions with the same start time are collapsed into a single "step"?)