Skip to contents

Bayes-Hilbert space operations on distributional data:

  • +.dd(p, q): density "addition" \(p \oplus q = pq / \int pq\); clr is clr(p) + clr(q).

  • -.dd(p, q): density "subtraction" \(p \ominus q = (p/q) / Z\); clr is clr(p) - clr(q).

  • *.dd(alpha, p): scalar multiplication \(\alpha \odot p = p^{\alpha} / Z\); clr is alpha * clr(p). One operand must be a numeric scalar; dd * dd is not defined in Bayes space and errors.

  • [.dd(x, i): select realizations from a multi-column dd.

Usage

# S3 method for class 'dd'
e1 + e2

# S3 method for class 'dd'
e1 - e2

# S3 method for class 'dd'
e1 * e2

# S3 method for class 'dd'
x[i]

# S3 method for class 'ddl'
e1 + e2

# S3 method for class 'ddl'
e1 - e2

# S3 method for class 'ddl'
e1 * e2

Arguments

e1, e2

dd (or ddl) operands. For * exactly one of them must be a length-1 numeric scalar.

x

A dd object to subset.

i

Index (vector, logical, or name) selecting realizations.

Value

A dd object (or ddl list for ddl operands).