The idea for v0.04 is to use the differences of hydrostatic pressure observations as basis for outlier (AO) detection.
Based on geotech data analysis, 10 cmH2O and 30 cmH2O were taken as maximum values for differences for 5 min and 15 min time-intervals.
This information was further used for AO detection. The results can be found here.
Some highlights on the way it works:
Top is the original timeseries. Below is the plot of minimal absolute differences (right and left side of each point). The threshold is placed at 10 cmH2O since the intervals are 5 min. The four points are clear outliers and are represented with red dots in the top plot.
The idea for v0.04 is to use the differences of hydrostatic pressure observations as basis for outlier (AO) detection.
Based on geotech data analysis, 10 cmH2O and 30 cmH2O were taken as maximum values for differences for 5 min and 15 min time-intervals.
This information was further used for AO detection. The results can be found here.
Some highlights on the way it works:
Top is the original timeseries. Below is the plot of minimal absolute differences (right and left side of each point). The threshold is placed at 10 cmH2O since the intervals are 5 min. The four points are clear outliers and are represented with red dots in the top plot.