ScoreFilter calculates the model and score for the given network structure.

This ignores model information in the input network, and re-fits (re-calculates) the B-spline non-parametric model, or other scoring model of the score function, and then replaces the old ones with new ones.

Arguments

file=file

data=file

Input data file.

file_args={ key = [ value,... ] }

data_args={ key = [ value,... ] }

Arguments for the input file.

score_args={ key = [ value,... ] }

Arguments for the score function.

dynamic

Dynamic model.

sdhybrid

Static-dynamic hybrid model.

j=n

target=n

Caculates the score only for the node n where n is the 0-origin index of the node (variable) in the data set.

ignore

Ingores the invalid scores such as infinity and/or NaN.

node_score

Stores node specific scores as the "score" node property.

save=file

Saves the data set consisting of variables only in the network in the specified file. The file is stored in the GDF format.

MPI support

MPI parallelization supported. Parallelization is based on distributing the child nodes among processes. The calculated model parameters are gathered in the root process, and the gathered parameters are compiled into the single network and distributed among the processes by the main program after the application of this filter.

Filters | INGOR Manual