segmented: Regression Models with Break-Points / Change-Points Estimation
(with Possibly Random Effects)
Fitting regression models where, in addition to possible linear terms, one or more covariates have segmented (i.e., broken-line
or piece-wise linear) or stepmented (i.e. piece-wise constant) effects.
Multiple breakpoints for the same variable are allowed. The estimation method is discussed in Muggeo (2003, <doi:10.1002/sim.1545>) and
illustrated in Muggeo (2008, <https://www.r-project.org/doc/Rnews/Rnews_2008-1.pdf>). An approach for hypothesis testing is presented
in Muggeo (2016, <doi:10.1080/00949655.2016.1149855>), and interval estimation for the breakpoint is discussed in Muggeo (2017, <doi:10.1111/anzs.12200>).
Segmented mixed models, i.e. random effects in the change point, are discussed in Muggeo (2014, <doi:10.1177/1471082X13504721>).
Estimation of piecewise-constant relationships and changepoints (mean-shift models) is
discussed in Fasola et al. (2018, <doi:10.1007/s00180-017-0740-4>).
Documentation:
Downloads:
Reverse dependencies:
| Reverse imports: |
AQEval, bioregion, ddiv, lactater, LadderFuelsR, mixtools, netSEM, PUPAK, PVplr, PWEXP, rcssci, respirometry, seqinr, spatialwarnings, StructuralDecompose, SWTools, takos, Trendy, weibulltools |
| Reverse suggests: |
fastcpd, nlraa, REddyProc |
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=segmented
to link to this page.
[aut, cre]