Generating dataframe with conditional random variables

Question

I'm trying to create a dataframe with two variables. The first one is binary physical activity status (active/non-active), the second one is presence of cardiovascular disease (also binary variable - 1/0).

The problem is that the presence of disease must occur with a given probability. It is 0.5 probability of disease in people with active status and 0.6 probability in people who doesn't physically active.

Please, really need help. I don't have any ideas how to do that. Looking into the sample and rbinom functions but to no avail.

Answer 1

You can do this in base R with an ifelse statement and the probs argument in sample. Below is an example with 10,000 observations (to help check that it works with prop.table)

set.seed(123)
n <- 1e4
df <- data.frame(physical_activity = sample(c("Active", "Non-Active"), n, replace = TRUE))

df$cardio_disease <- ifelse(df$physical_activity %in% "Active", 
                            sample(0:1, n, replace = TRUE), # P(disease|active) = 0.5
                            sample(0:1, n, replace = TRUE, prob = c(0.4, 0.6))) # P(Disease|non-active) = 0.6

Check results:

prop.table(table(df), margin = 1)

#physical_activity         0         1
#       Active     0.5062787 0.4937213
#       Non-Active 0.4021674 0.5978326