test_hypothesis() has been overhauled to allow selection of the test to be used and return of a list which includes the test name.
The calc_cumsum() function has been removed.
Introduced functions to facility creation of CUSUM charts. These include cusum_failure(), cusum_loglike(), cusum_ome(), and cusum_sprt().
P-value formatting has been moved to an exported function, paste_pval().
ggplot2 added as a suggested dependency.
paste_efs() now sets the names of the returned vector using the times provided.
Created a pkgdown website for the package.
Transitioned to inherits inside paste_efs.
Fix to calc_duration casting for upcoming purrr 1.0.0 release (Thanks Hadley!)
Fixed faulty class check in test_hypothesis and updated documentation for clarity.
Added calc_cumsum which provides additional NA handling when calculating cummulative sums.
Fixed LazyData CRAN error.
Added dplyr to suggests.
The final tabulate function, tabulate_at_risk, has been removed after being internalized by utile.visuals package.
Unused dplyr dependency removed.
vctrs dependency added.
Fixed incorrect estimate rounding in <1 checks for paste_median and paste_mean.
Fixed issue where test_hypothesis would not accept logical stratification data.
Added more reliable type checking and recycling to calc_duration.
All functions now use a consistent x/y argument style.
Added paste() & paste0(). Base functions with added option for NA removal. Default behavior
identical to base functions.
Removed tabulate_model(). Feature parity has been provided in utile.tables::build_table().
Pruned dependency list.
Overhauled documentation.
calc_duration(): Now supports the return of a lubridate::duration() object when the units parameter is left unspecified.Improved NA value handling of calc_duration().
paste_freq():
paste_ functions: Fixed parameter naming [BREAKING CHANGE]
Consolidated test_ functions into test_hypothesis() which has S3 methods for numeric, factor, and logical data. Numeric methods now support >2 strata. [BREAKING CHANGE]
Added calc_chunks() & chunk_data_(). Tools for breaking tibbles, data frames, and vectors into smaller, usable chunks of data.
tabulate_model(): Formatted tables now combine the point estimate and confidence interval into a single column to be more consistent with the output of utile.tables:: functions.
paste_freq(): Non-numeric data is now tallied and documentation has been updated.
paste_mean()/paste_median()/paste_efs(): Added checks for numeric data.
calc_duration(): Added support for multiple calculations against a single timepoint.tabulate_model(): Rethought the purpose of this function.
format= option to return a human-readable, ready-to-export table.tabulate_model.glm() which only supports the 'binomial' family of glm functions for now.tabulate_model.lm() for linear regression model support.tabulate_model.survfit() as other packages (i.e. 'broom') already provide this functionality.tabulate_logit(): Removed now that support has been added to tabulate_model().
tabulate_at_risk(): Now returns the strata column as factor instead of character type.
paste_freq(): Switched from an S3 object to a regular function.
dplyr::mutate()).paste_median()/paste_mean(): Fixed logic for returning '<1' if a statistic rounds to 0 and less.than.one=TRUE.
calc_duration():
dplyr::mutate()).