SEDesign: experiment design and contrasts object (S7 class)

SEDesign enforces the relationship between individual samples, design groups, and groups involved in statistical contrasts.

Usage

SEDesign(
  design = matrix(nrow = 0, ncol = 0),
  contrasts = matrix(nrow = 0, ncol = 0),
  samples = character(0),
  factors = character(0)
)

design: matrix with rownames as samples, colnames as design groups, containing 0/1 sample-to-group association values.
contrasts: matrix with rownames matching colnames(design), and colnames as contrast names, containing contrast coefficients.
samples: character vector of sample identifiers, typically equal to rownames(design).
factors: character vector of labels for the underlying experimental factors, equivalent to colnames(design_df), i.e. the columns produced by splitting each design group name (groups(object)) on "_". When empty, or when its length does not match the number of underlying factors, it is auto-populated with "factor1", "factor2", etc. Editing factors() only renames these labels (used primarily by print()); it has no effect on design, contrasts, or groups().

In addition to the four constructor arguments below, SEDesign objects carry two internal-use properties, accessible via @:

design_df: data.frame with one row per design group (groups(object), i.e. colnames(design)), derived by splitting each group name on "_" into one column per underlying experimental factor. Column names default to "factor1", "factor2", etc., and can be customized via factors()<-.
contrasts_df: data.frame, cached result of contrasts_to_factors(), refreshed whenever design or contrasts are updated.