- new exemplary dataset
data.timss1 borrowed from the
BIFIEsurvey package for illustration purposes of the
checkBackgroundVariables() function
- new function
checkBackgroundVariables() checks the
consistency of the variables in a data set if the variables belong to
different groups, which in turn have different properties, for example
ID variables, weighting variables, replication weights or
covariates.
existsBackgroundVariables() now allows to check the
number of variables entered if, for example, several items but only one
ID may be defined for a data record. It also checks which variable
groups may overlap and which may not.
- new function
cleanifyString() removes non-alphanumeric
characters and special characters from a string or factor variable
- minor bugfixes when integers are very large in
addLeadingZerosToCharInt() (Previously an error message was
displayed.)
- minor bugfixes in the messages of
mergeAttr()
recodeLookup() return the original vector if the the
user does not provide any recode table (i.e., if the argument is NULL).
(Previously an error message was displayed.)- minor bugfixes in the display of
makeTria()
- messages can be turned off in
makeDataFrame()
- package sticker added
- bugfixes in
gsubAll(): Does not longer allow using
regular expressions (internally, gsub(..., fixed = TRUE) is
used). Function now begins with the longest string to avoid repeated
modifications if strings in “old” match each other. See function
documentation for further details.
- add argument checks using the checkmate package
- bugfix in
halveString() for splitting at the last
instance of a regular expression
- new function
rbind_fill_vector() allows
rbind() for unidimensional vectors of unequal length,
filling empty entries with NA
- new operator
%$$% works for lists similar to
$ but gives error instead of NULL if the corresponding list
element does not exist
- internal fixes to tests (removed language dependency)
- new function
makeTria() for internal use in
eatTools and eatModel reshapes
covariance/correlation matrices into triangular shape
- added
seq2() for sequence generation
- adapt
halveString() for patterns with more than 1
character
- add messages to
mergeAttr() if combinations of merge
variables from one data set do not occur in the other data set
- add messages to
mergeAttr() if the two
data.frames intended to merge contain additional common
variables (beyond the by-variables)
makeDataFrame() gives warning/errors if data.frame has
less rows than expected at the least.existsBackgroundVariables() gives warning/errors if
variables have missing values, depending on how critical the missing
values are considered to be
- new function
rbind_common() rbinds a list of
data.frames and selects only common columns
- bug fix in
mergeAttr() when variables have more than
one class
mergeAttr() gives warning if merging variables have
NAs- optional colnames argument in
halveString()
function
- new function
print_and_capture() mainly for internal
use in eatRep and eatCodebook packages allows
for easy integration of (small) tables into (error) messages
- added some unit tests
- new function
makeDataFrame() for internal use in
eatRep and eatModel packages converts
tibbles and data.tables to data.frame
- new function
mergeAttr() mainly for internal use in
eatRep and eatModel packages restores
attributes after merging and provides messages if merging units are not
unique or only partially matches
- new function
pwc() computes part-whole-correlation
between an item and the whole scale (except this item)
- new function
roundDF() rounds all numeric variables in
a data.frame
- new function
na_omit_selection() drops rows containing
missing values in selected columns
existsBackgroundVariables(): add optional warning if
variables contains missing values
- new function
readMultisep() reads in data.frames with
separator characters >=1Byte
- new function
addLeadingZerosToCharInt() adds leading
zeros to all columns that can be identified as integers in a character
data.frame
- bug fix in
descr(). missing handling in variable when
additional weights are used
- bug fix in
tablePattern() when additional weights are
used
- new function
recodeLookup() allows recoding of a
variable based on a look up table
- functionality of
wideToLong() extended
- Switch to Github Action for CI
- new function
do_call_rbind_withName() allows row
binding a list of data.frames while keeping their list
names
- bug fix in
wideToLong()
asNumericIfPossible() now works with matrices