Original aggregate data format from Remiro-Azocar study changed to a long format
Original covariate names from Remiro-Azocar study generalised and user-defined
Input data can now contain any combination of binary and continuous covariates
Treatment names taken from data, rather than hard coded as “A”, “B”, “C” (5119596)
simulate_ALD_pseudo_pop() can now take user-provided
marginal arguments for a target distribution with
marginal_distns, marginal_params. These are
defined optionally in strategy_gcomp_ml() and
strategy_gcomp_bayes()
Included correlation of ALD simulations as optional function argument (d48d0ab)
print method available (983cb2f)
outstandR() call returns absolute values, as well as
contrasts (2f1fbd7)
outstandR() family argument now takes
binary (binomial), continuous (gaussian) and
count data (poisson)
Separate vignettes written for binary, continuous and count outcome data (fc10a68)
pkgdown site built and published on GitHub
(4df16f2)
outstandR() scale argument gives the
ability to select outcome scale from log-odds, risk difference, relative
risk
Moved covariate generation in to separate simcovariates
package