Fix extra columns with column groups (#106).
Fix bug in t1flex() (#103).
Handle NA in parse.abbrev.render.code().
Change space in as.data.frame.table1() (#60).
Add NMISS to output of stats.default() for categorical variables.
Handle NMISS properly in stats.apply.rounding() (no decimals).
Fix handling of abbreviated code with render.missing.
Correct default rounding of PCTnoNA (ie, using digits.pct).
Fix issue with t1kable (#82).
Fix a problem with t1kable when the row with the label contains data.
Choose between qflextable, flextable and regulartable in t1flex.
Add extra columns to groupspan headers in t1flex and t1kable (#93).
Option to place overall on the left (#88).
Missing values in stratification variable(s) raises and error (#80).
Add option rounding.fn to stats.apply.rounding() to allow changing the
rounding behaviour (thanks to Niels Hagenbuch for the suggestion)
Add geometric standard deviation (GSD) to the output of stats.default().
Fixed error when passing digits (thanks to Niels Hagenbuch for reporting this).
Fixed integer formatting (issue #55).
Fixed issue #53.
Headline feature: ability to convert table1 output to a data.frame,
flextable or kableExtra object (though with some limitations). As a
direct consequence, table1 can now be used within LaTeX documents, and
there is a new vignette to document this feature. Output to .docx is also
possible with this feature (through pandoc or flextable).
Fomulas containing dot ('.') are now handled correctly (the '.' means "all variables other than those that appear elsewhere in the formula").
Arguments to formatC (e.g. big.mark) will apply to all numbers by default (previously, only continuous variables had this).
New setLabel() function provides an alternative syntax for setting labels
(useful for setting a label inside a formula).
The headline feature is the extra.col option, which gives a flexible way to
add extra columns (e.g. P-values, SMD, ...) to the table (thanks to Matt
Leipzig for helping out with suggestions and testing on this).
Footnotes now appear in a
tag for better placement.General improvements to the documentation.
Fixed issue that was causing vignette build to fail (due to changes made to the MatchIt package's lalonde data set).
Allow multiple footnotes.
Added caption argument to generate a caption for the table.
Set render.missing=NULL to ignore missing values.
Added SUM to stats.default for continuous variables.
Added PCTnoNA to stats.default and na.is.category option to
render.categorical.default.
Changed (n=XX) to (N=XX) in default table header.
Improved error handling in function eqcut.
Function signif_pad can now use different decimal marks to support
different locales.
Added function round_pad (similar to signif_pad but for fixed rounding).
Added function t1read.
Added quantile.type argument to stats.default to allow selecting which
quantile type is computed.
Special handling for "0%" and "100%" in stats.apply.rounding so that no
decimals are shown.
droplevels now works properly with nested grouping.
table1 now returns a character string (containing HTML). A print method
handles the printing. There is no longer a need to use capture.output to
embed table1 in a shiny app.
Removed standalone option as it is no longer needed.
Use htmltools functionality to print.
Added knit_print method. No need to specify results="asis" chunk option
anymore when knitting.
Default CSS is included automatically. It can still be overridden by specifying a custom stylesheet in R Markdown.
Changed notation from upper case "Q" to small case "q" for percentiles in
stats.default.
Q1 and Q3 are now aliases for q25 and q75 respectively in
stats.default.
Tertiles added to stats.default.
New function eqcut for creating equal-sized categories from continuous
variables.
New footnote argument for adding a footnote to the table.
The vignette has been expanded with new examples and the documentation improved.