randomize_minimal: Minimal randomisation procedure

In franzikoch/Feedbackloops: Feedback loop analysis for competition networks

Minimal randomisation procedure

Description

In this version, pairs are kept together and their above-below diagonal orientation is also preserved.

Usage

randomize_minimal(df, ij_col, ji_col)

Arguments

df	interaction table (created by interaction_strenghts())
ij_col	column of a_{ij} values to randomise (choose scaled or unscaled)
ji_col	column of a_{ji} values to randomise (scaled or unscaled)

Details

In the interaction table this means the following: a_{ij} and a_{ji} values that appear in the same row in the original table, will also be in the same row in the pairwise randomised table. In contrast to the pairwise (weak) randomisation, they cannot switch from a_{ij} to a_{ji} or vice versa. Only the order of rows in the table is randomised.

The function returns a new interaction table that contains two new columns $a_ij_min and $a_ji_min that contains the same values as the original columns but in a different order.

To get a minimally randomised Jacobian matrix, use assemble_jacobian() and specify the new columns.

Value

Interaction table with two additional columns $a_ij_min and $a_ji_min, containing randomised interaction strengths

franzikoch/Feedbackloops documentation built on July 1, 2023, 12:42 p.m.