shewhart: when argument limits=NA (i.e., when the control limits are not precomputed), the control limits are computed using a permutation approach also for the rank based statistics; hence, they guarantee the desired FAP also when the in-control distribution is not absolutely continuous.
shewhart: added control charts based on the Lepage and Cucconi control statistics.
shewhart: default for argument aggregation is now ‘mean’ (and not ‘median’ as before). This was introduced for ‘textbook obedience’.
mshewhart: added the possibility to take into account the subgroup structure of the data when computing the scatter matrix used to (inner) standardise the spatial signs and ranks and the multivariate signed ranks.
Minor changes to the Rcpp sources to cope with Rcpp’s STRICT_R_HEADERS and USE_FC_LEN_T/FCONE declarations
Added CITATION
rsp
uses the Rcpp interface as the rest of
the package (and not the .C interface as before).
Correction of some small problems resulting in some warnings
Eliminated the dependency from the deprecated (since C++14)
std::shuffle