Convert wlr() to S3 generic to accept tte_data or counting_process

[jdblischak]

Closes #275

Added the class "tte_data" to the output of cut_data_by_event() and cut_data_by_date(), then converted wlr() to an S3 generic. The "counting_process" class was already added to the output of counting_process() in #267

[jdblischak]

I rebased onto #273, which has broken my strategy for converting wlr() to an S3 generic. The goal was to pre-compute counting_process() prior to running multiple iterations of wlr(), but now the pre-counting-process data is required to compute the new ratio

simtrial/R/wlr.R

Lines 89 to 95 in 2ed558b

	x <- data |> counting_process(arm = "experimental")
	# calculate the sample size and randomization ratio
	n <- nrow(data)
	ratio <- sum(data$treatment == "experimental") / sum(data$treatment == "control")
	q_e <- ratio / (1 + ratio)
	q_c <- 1 - q_e

[LittleBeannie]

Got you! I guess one possible way is to get ratio as an attribute of data.

ratio <- sum(data$treatment == "experimental") / sum(data$treatment == "control")

[jdblischak]

I guess one possible way is to get ratio as an attribute of data.

Ok, I proceeded with that strategy. Though instead of passing the ratio, I passed the number of experimental and control treatments.

[LittleBeannie]

Thank you, @jdblischak! This PR has been merged.