- New method
shapeConstr
to build constraint matrices for shape-constrained splines. Currently works with classesns
,bs
(from packagesplines
),ps
andonebasis
(fromdlnm
). Also includes a default method for more general bases.
- Changed the default QP solver to
quadprog
after some expriments. - Added the argument
complete
in inferential functions, to allow keeping or discarding aliased coefficients. Same interpretation as invcov.lm
. vcov
now allows returning the usual variance-covariance returned byvcov.glm
whenconstrained = FALSE
.simu_coef
renamed assimulCoef
and now includes an argument to set the seed.check_cmat
has been renamedcheckCmat
.- Now the
Cmat
,lb
andub
used are not returned in thecontrol
object from the result ofglm
withcirls.fit
. - Now
lb
andub
can be passed by term.
- Fixed issue with R matrix when there was less observations than variables.
- Fixed error from
solve.QP
when there are large values in the response. - Now
checkCmat
also checks if there are "zero" constraints
- Added
check_cmat
andcoef_simu
to the list of exported functions as they can be useful for specific use cases. - Added full documentation.
- In
check_cmat
, removed the call tolimSolve::nnls()
to be replaced byconeproj::coneB
(also a NNLS solver) to reduce the number of dependencies.
- Initialization of a short documentation for several functions.
cirls.control
now checks for constraint matrix irreducibility.cirls.control
is now exported.- changed the function to determine redundant constraints in
check_cmat
- No warning message on row rank of
Cmat
anymore Cmat
is now checked only when there are more than one row
- Method vcov for cirls to compute corrected covariance matrices
- Method confint for cirls to compute feasible confidence intervals
- Added several QP solvers: quadprog (the original one), osqp and coneproj.
- A warning is now displayed when Cmat is not of full row rank
- vcov and confint return NA matrices if Cmat is not of full row rank
- Changed residual df computation to account for active constraints
- Replaced bvec by lb (lower bound) and ub (upper bound). Allows equality constraints.
- Added cirls class to glm output
- cirls.fit has the same behaviour as glm.fit when model is singular: fill with NAs