Add data clean-up postprocesses to Signal
Summary
Add data clean-up postprocesses to Signal
object.
Current behaviour/setbacks
We have post-processes that can function well as fundamental data clean-up processes. For example, these include smoothing/filtering (Savitzky-Golay), detrending, normalising data (scikit-learn
's Standard Scaler). Users can invoke these after obtaining post-processed data, but they have to do so manually.
Desired behaviour/advantages
As we anticipate that these clean-up processes are broadly desired before further analysis, the Signal
object can include these additional processes.
Implementation sketch
Within agora/io/signal.py
, the Signal
object contains an apply_prepost()
method (line 62). Currently, it only applies the picker
and merger
processes by default; additional post-processes (e.g. savgol
, detrend
, standardscaler
) can be added there.