junco (version 0.1.4)

h_odds_ratio: Helper functions for odds ratio estimation

Description

Functions to calculate odds ratios in s_odds_ratio_j().

Usage

or_glm_j(data, conf_level)
or_clogit_j(data, conf_level, method = "exact")
or_cmh(data, conf_level)

Value

A named list of elements or_ci, n_tot and pval.

Arguments

data

(data.frame)
data frame containing at least the variables rsp and grp, and optionally strata for or_clogit_j().

conf_level

(numeric)
confidence level for the confidence interval.

method

(string)
whether to use the correct ('exact') calculation in the conditional likelihood or one of the approximations, or the CMH method. See survival::clogit() for details.

Functions

• or_glm_j(): Estimates the odds ratio based on stats::glm(). Note that there must be exactly 2 groups in data as specified by the grp variable.

• or_clogit_j(): Estimates the odds ratio based on survival::clogit(). This is done for the whole data set including all groups, since the results are not the same as when doing pairwise comparisons between the groups.

• or_cmh(): Estimates the odds ratio based on CMH. Note that there must be exactly 2 groups in data as specified by the grp variable.

See Also

odds_ratio

Examples

data <- data.frame(
  rsp = as.logical(c(1, 1, 0, 1, 0, 0, 1, 1)),
  grp = letters[c(1, 1, 1, 2, 2, 2, 1, 2)],
  strata = letters[c(1, 2, 1, 2, 2, 2, 1, 2)],
  stringsAsFactors = TRUE
)

or_glm_j(data, conf_level = 0.95)

data <- data.frame(
  rsp = as.logical(c(1, 1, 0, 1, 0, 0, 1, 1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0, 0)),
  grp = letters[c(1, 1, 1, 2, 2, 2, 3, 3, 3, 3, 1, 1, 1, 2, 2, 2, 3, 3, 3, 3)],
  strata = LETTERS[c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2)],
  stringsAsFactors = TRUE
)

or_clogit_j(data, conf_level = 0.95)

set.seed(123)
data <- data.frame(
  rsp = as.logical(rbinom(n = 40, size = 1, prob = 0.5)),
  grp = letters[sample(1:2, size = 40, replace = TRUE)],
  strata = LETTERS[sample(1:2, size = 40, replace = TRUE)],
  stringsAsFactors = TRUE
)

or_cmh(data, conf_level = 0.95)