PRFilter calculates the partial residuals and then saves them in files. Also, log-likelihood values are calculated and saved in the file. The output files can be used with the INGOR R script to display estimated regression curves by the B-splines.

Arguments

file=file

input data file in GDF format. The file needs to be the same one used to estimate the network.

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

Arguments for the input data file.

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

Arguments for the score function.

dynamic

d

Dynamic model.

sdhybrid

sd

Static-dynamic hybrid model.

output=prefix

The prefix of file names to output calculated partial residuals and so on. Seven files are generated: prefix.X, prefix.Y, prefix.PR.Y, prefix.LL, prefix.Z, prefix.D and prefix.PL. See the explanation of the --output-data option of INGOR Manual for details. For .PL file, see below.

Note that the line order of .LL, .Z, .D, and .PL files depends on the index of the nodes of the network. This may be different from the order of the data file given by the file= argument above. Be carefule this, especially when you read the network file in the TXT format by the --read filter.

The .PL file is generated by only this filter. It contains a primary ID specific log likelihood in each row. Thus if the input file contains N different primary IDs, the .PL file consists of N real values in N lines.

fit

Specifies not to use model attributes (parameters) in the network. Instead, re-fits the model to the input data.

Filters | INGOR Manual