Create a dataset with known values for each parameter (A=1, B=2, C=3, D=4). The outcome is a ratio with some combination of [A,B,C,D] in the numerator and some combination of [A,B,C,D] or "r" (reference) in the denominator. Encode the covariates A, B, C and D as a network, adding 1 when the component is in the numerator and subtracting 1 when the component is in the denominator.

Comparing survival analysis results from Cox proportional hazards and logistic regression. The logistic version models time as a covariate and gets the conditional probability that the event occurs in the next time step. Data 'ovarian' are from a randomized trial of treatments for ovarian cancer

Generate quantiles in (0,1] for two variables with a given correlation structure. These can be used to set a simulant's position in two arbitrary probability distributions