butter.parallel: The butter function takes in a dataframe with 'words' and...

In csqsiew/samr: Spreading Activation Machine in R

Usage

butter.parallel(start_run, decay = 0.2, retention = 0, suppress = 0.1,
  network = gc.net, time = 10)

Arguments

start_run

A non-empty dataframe with 'words' and 'activation' columns. Must be specified.

decay

Proportion of activation that is lost at each time step. Default is 20 \item retentionProportion of activation that remains in the node, ranges from 0 to 1. Default is 0. \item suppressSuppress nodes with total final activation of < x units at each time step. Recommended value of x is 0.1 \item networkNetwork where the spreading occurs. Must be specified. Default is gc.net. \item timeNumber of time steps to run spread function for. Default is 10. A compiled dataframe with 'words', 'activation' and 'time' columns showing the spread of activation in the network over time. Note: butter.decay() is a modified function of butter.retention() and calls the parallelized spread.decay.parallel() function for faster processing. The main difference is that spread.decay() specifies decay rate, d, which the rate at which activation is lost at each time step. Use the estimate.time() function to figure out the number of time steps required for total activation in the network to decrease to 10 See Vignette for examples.

csqsiew/samr documentation built on May 28, 2019, 7:49 p.m.