observe_data function

Observe data, given a strategy

observe_data(
  complete_data,
  observed = NULL,
  nodes_to_observe = NULL,
  prob = 1,
  m = NULL,
  subset = TRUE
)

Arguments

• complete_data: A data.frame. Data observed and unobserved.

• observed: A data.frame. Data observed.

• nodes_to_observe: A list. Nodes to observe.

• prob: A scalar. Observation probability.

• m: A integer. Number of units to observe; if specified, m

  overrides prob.

• subset: A character. Logical statement that can be applied to rows of complete data. For instance observation for some nodes might depend on observed values of other nodes; or observation may only be sought if data not already observed!

Returns

A data.frame with logical values indicating which nodes to observe in each row of complete_data.

Examples

model <- make_model("X -> Y")
df <- make_data(model, n = 8)
# Observe X values only
CausalQueries:::observe_data(complete_data = df, nodes_to_observe = "X")
# Observe half the Y values for cases with observed X = 1
CausalQueries:::observe_data(complete_data = df,
     observed = CausalQueries:::observe_data(complete_data = df, nodes_to_observe = "X"),
     nodes_to_observe = "Y", prob = .5,
     subset = "X==1")

See Also

Other data_generation: data_helpers, get_all_data_types(), make_data_single()