reshape_all_samples: Wrapper for reshape_samples, applying too many fits

Description Usage Arguments Value

View source: R/reshape.R

Description

We might want to extract the true and fitted values across several nmf experiments. Each experiment individually would require calling reshape_samples.

Usage

1
reshape_all_samples(fits, config_path, param, dims, sim_seed = 1112017)

Arguments

fits

[list of paths] A list of paths to the fitted STAN samples. They will be loaded and the associated samples will be extracted.

config_path

[character] The path to the configuration JSON containing all the experiment information for these fits. It is assumed that the names in fits include the (necessarily numerical) IDs specified in this JSON object.

param

[character] The name of the parameter to extract data for.

dims

[character vector] The names to use for the rows and columns of the parameter matrix.

sim_seed

[integer] The seed used in simulating the parameter of interest. This is necessary because we don't read the true values anywhere, we regenerate the parameter values.

Value

reshaped [data.table] The same kind of object as the output of reshape_samples(), except over experiments with multiple configurations.


krisrs1128/nmfSim documentation built on May 20, 2019, 1:30 p.m.