npwbs 0.4.0
- Added Mann–Whitney, Mood, Cramér–von Mises, and modified Baumgartner
detection alongside the existing Lepage method, using compact batch C++
scanners.
- Added whole-input random strict ordering for ties, controlled by
breakTies.
- Lepage
alpha = 0.05 now uses the final registered
book-chapter calibration and may produce small changepoint differences
from version 0.3.0.
- Lepage
alpha = 0.01 retains the established package
calibration because no corresponding book-chapter calibration was
generated.
- Both bundled Lepage threshold vectors are now stored only through
n = 10000.
npwbs 0.3.0
- Updated built-in Lepage thresholds to match the revised manuscript
calibration for the default settings
M = 10000 and
d = 2.
- Clarified that the packaged thresholds support
alpha = 0.05 and alpha = 0.01.
- For segment lengths above
n = 10000, the package now
uses the n = 10000 threshold with a warning instead of
silently extrapolating.
- Clarified the returned changepoint convention: a returned value
k denotes a split between observations k and
k + 1.