Makes a specification for data.frame. Creates a template based on the data.frame. Uses column names for labels where columns do not have a label attribute. Factors will be encoded. numerics will be rounded to digits and like integers will be expressed as ranges in guide column. Integers and character with less than or exactly tol unique values will be encoded.
# S3 method for data.frame
specification(x, tol = 10, digits = 20, ...)spec data.frame with columns as follows.
Column name.
A descriptive label. Save and edit as necessary using external tool.
A guide to interpretation. NA for arbitrary character; range [low:high] for integer and numeric; an encoding e.g. //0/no//1/yes// for factor-like items ... save and edit factor labels as necessary using external tool.
For numeric ranges you can add text, such as units. E.g. if default guide is '[0:100]' you can edit to give 'mg [0:100]'. Or you can just substitute 'mg'. guidetext extracts just the character portion, and matches enforces the numeric range.
An R expression that can be coerced to logical. TRUE means item cannot be NA.
Arbitrary comment, e.g. derivation of the item given by column.
object
integer
integer
passed arguments
link{read.spec} write.spec respecify.character write.spec matches
Other specification:
specification.comment(),
specification.default(),
specification()
data(drug)
file <- tempfile()
spec <- specification(drug, tol = 3)
Run the code above in your browser using DataLab