At the moment, the degradation_timeseries_plot shows the timeseries shifted pretty far to the right, relative to the raw data timeseries. This is because the yoy_results output from the TA object is right-labeled (annual YoY slopes are 1 year in length and labeled on the right side), and the pd.rolling function to get the rolling median for the timeseries is also right labeled. So if you have the default 365 day rolling median, your timeseries starts 2 years delayed from the raw data.
Being able to select options for 'right', 'center' and 'left' labeling would allow an output closer to what I would expect, which would have the rates line up more closely with the raw data timeseries.
At the moment, the degradation_timeseries_plot shows the timeseries shifted pretty far to the right, relative to the raw data timeseries. This is because the
yoy_resultsoutput from the TA object is right-labeled (annual YoY slopes are 1 year in length and labeled on the right side), and the pd.rolling function to get the rolling median for the timeseries is also right labeled. So if you have the default 365 day rolling median, your timeseries starts 2 years delayed from the raw data.Being able to select options for 'right', 'center' and 'left' labeling would allow an output closer to what I would expect, which would have the rates line up more closely with the raw data timeseries.