- The package is considered stable enough to be used in production
- New maintainer
- Bugfix: additional outputs are indeed passed on to a custom display function (Issue #13 raised by @Dmitrilbr, Bugfix in PR #14)
- Adaptation to the upcoming Matrix 1.5
- Remove the giveCsparse warning (appeared with Matrix 1.3 shipped with R 4.0) (issue #7)
- Fix the discretization problem with qualitative variables with one modality (issue #8), thanks to @ThomasDeroyon
Change maintainer e-mail address.
Debug against R-devel: as of R 4.0.0, class(.) on a matrix object has length > 1.
-
Breaking: Heavy remanufacturing of
define_variance_wrapper
- New:
technical_data
argument offers a more consistent way to include technical data within the enclosing environment of the wrapper.objects_to_include
is kept for non-data objects (such as additional statistic wrappers) or advanced customization. - New:
technical_param
argument offers a more convenient way to specify default values for parameters used by the variance function. - New:
reference_weight
replacesdefault$weight
. This means that the reference weight used for point estimation and linearization is set while defining the variance wrapper and not at run-time. - Deprecated:
stat
, which was a remain of an early implementation of linearization functions, is not a parameter of the variance wrappers anymore. Its purpose (to apply a given variance wrapper to several variables without having to type the name of the linearization wrapper) is now covered by the standard evaluation capabilities of statistic wrappers (see below). - Deprecated:
default
is replaced bydefault_id
, asdefault$weight
anddefault$stat
are no longer needed. As fordefault$alpha
, its value is set to 0.05 and cannot be changed anymore while defining the variance wrapper (as this can easily be done afterwards usingformals<-
). - Deprecated:
objects_to_include_from
- New:
-
Breaking: Rebranding and heavy remanufacturing of
define_statistic_wrapper
(previously known asdefine_linearization_wrapper
), added support for standard evaluation (seedefine_variance_wrapper
examples). -
New: the
qvar
function allows for a straigthforward variance estimation in common cases (stratified simple random sampling with non-response through reweighting and calibration) and performs both technical and methodological checks. -
Some normalization in function names:
add0
becomesadd_zero
,sumby
becomessum_by
,rescal
becomesres_cal
-
Example data: calibration variables in ict_sample instead of ict_survey, new LFS example data
-
Significant increase of unit tests
- Hotfix: Add calibrated weights to
define_variance_wrapper
example.
- Simulated data added
- Significant increase of unit tests
- Documentation completed
- Simplification of the structure of the main object processed by the variance wrapper
- Removal of unnecessary arguments in linearization wrappers
- Removal of the linerization wrappers for the Laeken indicators based on the vardpoor package (better integration in a future release)
- Preparation for a first CRAN release
- Now linearization with all data parameters set to NULL are discarded from the estimation.
- Bug fix: evaluation of variables can occur either in the data argument or in the evaluation environment (envir argument)
- Several attempts to output more metadata from linearization functions.
- At the end : ratio() gains two metadata slots, est_num and est_denom
- Minor bug fixes
- Beginning of the documentation
- Renaming of numerous functions and arguments
- Change the precalc structure in varDT
- Normalize the treatment of weights
- New linearization wrappers: gini() and arpr()
- No more dependency to package pryr
- Add the generalized inverse in
varDT
- Other bug fixes