search

osofr / simcausal

Simulating Longitudinal and Network Data with Causal Inference Applications
64 stars 11 forks source link

Evaluation of categorical by variable name #7

Open osofr opened 8 years ago

osofr commented 8 years ago

THIS no longer works, because the dimensions of catprob.W0 and catprob.W1 (1 row) do not match the dims of W (n rows)

  D <- DAG.empty()
  D <- D + node("W", distr = "rbern", prob = 0.3)
  catprob.W0 <- cbind(0.7,0.1,0.2)
  catprob.W1 <- cbind(0.2,0.1,0.7)
  D <- D + node("Cat3", distr = "rcategor.int", probs = (W==0)*.(catprob.W0) + (W==1)*.(catprob.W1))
  checkException(Dset2 <- set.DAG(D))

This works however, because simcausal gets a chance to parse c(0.7,0.1,0.2) into a matrix with appropriate number of rows:

  D <- DAG.empty()
  D <- D + node("W", distr = "rbern", prob = 0.3)
  catprob.W0 <- cbind(0.7,0.1,0.2)
  catprob.W1 <- cbind(0.2,0.1,0.7)
  D <- D + node("Cat3", distr = "rcategor.int", probs = (W==0)*c(0.7,0.1,0.2) + (W==1)*c(0.2,0.1,0.7))
  Dset2 <- set.DAG(D)