SuppressDominantCells()
is now considered a common function for both the nk-dominance rule and the p-percent rule.- The
pPercent
parameter is now exposed in theSuppressDominantCells()
documentation.
- The
- The
n
parameter inSuppressDominantCells()
now defaults to1:length(k)
.- To simplify common use.
- A problem in experimental interval suppression is now fixed.
- It was a bug occurring in cases where a response value of zero was secondary suppressed.
- Improvements due to updates in the SSBtools package (version 1.5.2).
- Fix for a rare problem in
GaussSuppression()
,- Could happen with parallel eliminations combined with integer overflow. Then warning message: longer object length is not a multiple of shorter object length
- Minor change to the singleton method
"anySum"
inGaussSuppression()
to align with best theory.- In practice, this rarely makes a difference.
- The previous behavior can be ensured by setting
singletonMethod
to either"anySumOld"
or"anySumNOTprimaryOld"
.
- Fix for a rare problem in
- Experimental functionality to meet interval width requirements has been incorporated
- If at least one of the two parameters below is specified, in addition to the
lpPackage
parameter, further suppression will be performed to satisfy the interval width requirements.rangePercent
: Required interval width expressed as a percentagerangeMin
: Minimum required width of the interval
- If at least one of the two parameters below is specified, in addition to the
- Improved singleton methods for magnitude tables
- A new default setting,
singletonMethod = "numttHTT"
, has been introduced in the wrappersSuppressDominantCells()
andSuppressFewContribitors()
. This setting represents the method that offers the highest level of protection. However, it should be noted that with this setting, the computational load of the suppression algorithm may double, which could potentially lead to a doubling of the execution time as well. During these computations, ":::" will be displayed instead of "....".- To prevent this doubling, set
singletonMethod = "numttHtT"
. - The behavior of version 0.7.0 can be restored by setting
singletonMethod = "numttH"
. - A simpler and faster method is achievable with
singletonMethod = "numttT"
. - As in previous versions, singleton handling can be disabled by setting
singletonMethod = "none"
.
- To prevent this doubling, set
- Additional information can be found by
?SSBtools::NumSingleton
. - Some explanation is provided at the bottom of the magnitude vignette.
- This will later be documented in a more comprehensive manner (paper).
- A new default setting,
- Improved functionality of the
SuppressDominantCells()
andSuppressFewContributors()
wrappers.- Improved support for handling multiple numerical variables, introducing new parameters:
dominanceVar
andcandidatesVar
. - The
removeCodes
parameter is now also available in theDominanceRule()
andSuppressDominantCells()
functions. - Support for multiple
contributorVar
(charVar
) in theSuppressFewContributors()
andNContributorsRule()
functions. - Now,
SuppressDominantCells()
includes special functionality to prevent zero cells, which have been suppressed, from being revealable in cases where negative values cannot occur. See the parametersingletonZeros
. - The update described below enables the specification of the
pPercent
parameter directly throughSuppressDominantCells()
.
- Improved support for handling multiple numerical variables, introducing new parameters:
- The p% rule for magnitude tables has been implemented through the introduction of a
new primary suppression function,
PPercentRule()
.- Technically,
PPercentRule()
andDominanceRule()
now serve as wrappers for the newly introduced. general functionMagnitudeRule()
.
- Technically,
AdditionalSuppression()
generalized to take a wrappers as input.- New special functions for the avoidance of suppression
PrimaryRemoveWg()
,CandidatesNumWg()
andForcedWg()
- Experimental functionality for interval calculations has been included
- Intervals can now be calculated using the new function
ComputeIntervals()
. - When the
lpPackage
parameter is specified inGaussSuppressionFromData()
or in any of its wrappers, intervals for primary suppressed cells will be computed and included in the output.
- Intervals can now be calculated using the new function
- More vignettes are included.
- Better singleton handling for magnitude tables when using
SuppressDominantCells()
andSuppressFewContributors()
.- Due to improvements in the SSBtools package (version 1.4.6).
- Due to new default
extraAggregate = TRUE
in the specs,dominanceSpec
andfewContributorsSpec
.
- More default values are explicitly included in the in specs so that they are easier seen.
- See
PackageSpecs()
.
- See
- A vignette entitled "Defining Tables for GaussSuppression" is now included.
- Now, easy-to-use wrapper functions are included.
SuppressSmallCounts()
,SuppressDominantCells()
, andSuppressFewContributors()
, along withSuppressKDisclosure()
(which was available in the previous version).
- Built-in specs that contribute to a simpler interface have been adopted.
- See
PackageSpecs()
.
- See
- Sampling weights are now possible in the dominance rule.
- See
DominanceRule()
.
- See
- More advanced singleton handling that makes use of new functionality in the SSBtools package (version 1.4.4).
- See
SSBtools::GaussSuppression()
. - See
SingletonUniqueContributor()
.
- See
- Now
forced
andusafe
are possible output columns.- Where unsafe means unsafe primary suppressions due to forced cells. That, is the unsafe primarily suppressed values can be re-calculated from the values of the cells forced to be not suppressed.
- See parameters
forcedInOutput
andunsafeInOutput
toGaussSuppressionFromData()
.
- Now the original variable names, as specified by
freqVar
andweightVar
, are kept in the output.- In previous versions these names were standardized to
"freq"
and"weight"
. - Code relying on previous behavior with other
freqVar
/weightVar
than"freq"
/"weight"
needs to be updated. "freq"
is still default when data is aggregated from microdata withoutfreqVar
specified (see new parameterfreqVarNew
).
- In previous versions these names were standardized to
- Adaption needed after Matrix ver. 1.4-2 (not a user-visible change)
- Last version before any news