tidycomm 0.4.2
CRAN release: 2025-08-27
Bugfixes
- Updated GGally usage to rely on
ggmatrix
class instead ofgg
, ensuring compatibility with the latest ggplot2 changes. - Migrated tests to testthat v3 to maintain correct test behavior.
Documentation
- Merged README.Rmd with index for pkgdown site.
- Updated links to reflect the new GitHub organization (https://github.com/tidycomm/tidycomm).
- Enabled Bootstrap 5 and light/dark switch for the pkgdown documentation site.
tidycomm 0.3.0
New features
- Converted
tibble
returns intotdcmm/tibble
return objects (they behave just like tibbles but are in essence our own objects now) - Added partial correlation in
correlate(..., partial = z_var)
- Added correlation with a focus variable
correlate(..., with = focus_var)
- Added linear regression
regress()
- Added one-sample t-test
t_test(..., mu = ...)
- Added
reverse_scale()
,minmax_scale()
,z_scale()
,center_scale()
,setna_scale()
,recode_cat_scale()
,recode_scale()
, anddummify_cale()
to shift and modify continuous and categorical scales - Added
tab_percentiles()
- Added
visualize()
to visualize almost everything - Added
snscomments
andincvlcomments
as additional data sets
Minor changes
- Changed
unianova()
andt_test()
to build onleveneTest()
- Allowed
test_icr()
to work with grouped data - Converted all code examples in documentation to be built on
tidycomm
-provided data sets - Added
omega_squared
,Levene_p
, andvar_equal
columns to default return fromunianova()
- Added
d
,se
,t
, anddf
- Removed
null.value
from list of return values inunianova()
post-hoc test - Renamed
unianova()
return column names toVariable
(previously:Var
),Group_Var
(prev.term
),Delta_M
(prev.estimate
),p
(prev.adj.p.value
),conf.lower
andconf.upper
(prev.conf.low
andconf.high
)
tidycomm 0.2.0
New features
- Added Fretwurst’s Lotus and S-Lotus intercoder reliability coeffecients to
test_icr()
function -
describe_cat()
function added to describe categorical variables
Minor changes
- More descriptive error messages for common errors
-
describe()
now also reports 95% confidence intervals -
describe()
now reports valid N instead of full N -
test_icr()
now works withtidyselect
selection
Bugfixes
- Empty groups are dropped if describing with more than one grouping variable
- Krippendorff’s Alpha returns 1 if variable has only one category
-
unianova()
now works with variable names containing whitespace - Groups are dropped for
test_icr()
to avoid computational issues