Chance baseline for each item

The proportion of a rule form's candidate space that contains each item – how often an item would be selected if selection were arbitrary. Enumerated from the candidate space rather than assumed, which matters under a cluster constraint: a two-item cluster contributes far more of its combinations than a seven-item cluster, so the baseline is a step function of the cluster rather than a constant.

Usage

chance_baseline(n_symptoms = 6L, clusters = NULL, n_total = 20L)

Arguments

n_symptoms

Combination size.

clusters

NULL, or a named list of integer vectors.

n_total

Number of items to choose from.

Value

A numeric vector of length n_total, on the 0-1 scale.

Details

The baselines sum to n_symptoms over the items, by construction.

Examples

# A small space keeps the example instant; the pattern is the same as at six
# symptoms, with the two-item cluster far above the seven-item one. Four is
# the smallest size that can cover four clusters.
round(chance_baseline(3) * 100, 1)
#>  [1] 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15
clusters <- list(B = 1:5, C = 6:7, D = 8:14, E = 15:20)
round(chance_baseline(4, clusters) * 100, 1)
#>  [1] 20.0 20.0 20.0 20.0 20.0 50.0 50.0 14.3 14.3 14.3 14.3 14.3 14.3 14.3 16.7
#> [16] 16.7 16.7 16.7 16.7 16.7