Associate per-sample values with clinical variables

Usage

clin_associate(d, key_col, value_col, clin, patient_of = NULL, min_n = 4L)

d: long data frame with sample_id, a key column and a value column.
key_col: column naming the thing measured: "population" or "marker".
value_col: the per-sample quantity.
clin: named list: variable -> named vector of sample_id -> value.
patient_of: named vector sample_id -> patient. When given, a variable that is CONSTANT within a patient is tested on one value per patient rather than one per sample. See clin_variable_unit().
min_n: fewest observations for a test to run at all.

data frame, one row per key x variable, BH-adjusted within variable.