Rf_error().The issue described athttps://github.com/RcppCore/Rcpp/issues/1287 has been fixed to avoidWARNINGs from CRAN checks on some platforms. Thank you to DirkEddelbuettel for providing the fix so quickly!
Fixed issues with the incorrect use of in some Rd files.
If the argumentk = 0 is supplied tokgaps() then an estimate of 1 is returned for the extremalindex for any input data. For this very special case the estimatedstandard error associated with this estimate is set to zero andconfidence intervals have a width of zero.
Corrected a typing error in the description ofuprobin the documentation forplot.choose_uk() andplot.choose_ud().
The unnecessary C++11 specification has been dropped to avoid aCRAN Package Check NOTE.
README.md: Used app.codecov.io as base for codecov link.
Create the help file for the package correctly, with aliasexdex-package.
The value returned bynobs.kgaps() was incorrect incases where there are censored K-gaps that are equal to zero. TheseK-gaps should not contribute to the number of observations. This hasbeen corrected.
In cases where the data used inkgaps are split intoseparate sequences, the threshold exceedance probability is estimatedusing all the data rather than locally within each sequence.
AlogLik method for objects inheriting from class"kgaps" has been added.
In the (unexported, internal) functionkgaps_conf_int() the limits of the confidence intervals forthe extremal index based on the K-gaps model are constrained manually to(0, 1) to avoid problems in calculating likelihood-based confidenceintervals in cases where the the log-likelihood is greater than theinterval cutoff when theta = 1.
In the documentation of the argumentk tokgaps() it is noted that in practicek shouldbe no smaller than 1.
The functionkgaps() also return standard errorsbased on the expected information.
In the package manual related functions have been arranged insections for easier reading.
Activated 3rd edition of thetestthatpackage
kgaps(),kgaps_imt() andchoose_uk() can now accept adata argumentthatNAs.cheeseboro is included, which is a matrixcontaining some missing values.kgaps(), the functionskgaps_imt() andchoose_uk() now have an extraargumentinc_cens, which allows contributions from censoredK-gaps to be included in the log-likelihood for the extremal index.inc_cens inkgaps()(and inkgaps_imt() andchoose_uk()) is nowinc_cens = TRUE."confint_gaps" returned fromconfint.kgaps().confint.spm() andconfint.kgaps() theinput confidencelevel is included in the outputobject.