fml_interpret {flowml}

fml_interpret

Description

Pipeline function that sets up and runs a post-hoc interpretation of an ml experiment. All results are written to rds files.

Usage

fml_interpret(parser_inst)

Arguments

parser_inst

instance of fml_parser class that comprises command line arguments.

Value

none

Author(s)

Sebastian Malkusch

Examples

## Not run: 
parser_inst <-  flowml::create_parser()

parser_inst$pipeline_segment <- "interpret"
parser_inst$config <- flowml::fml_example(file = "reg_config.json")
parser_inst$data <- flowml::fml_example(file = "reg_data.csv")
parser_inst$samples_train <- flowml::fml_example(file = "reg_samples_train.txt")
parser_inst$samples_test <- flowml::fml_example(file = "reg_samples_test.txt")
parser_inst$features <- flowml::fml_example(file = "reg_features.txt")
parser_inst$extended_features <- flowml::fml_example(file = "reg_features_extended.txt")
parser_inst$trained <- flowml::fml_example(file = "reg_fit.rds")
parser_inst$interpretation <- "shap"
parser_inst$result_dir <- tempdir()

flowml::fml_interpret(parser_inst = parser_inst)

## End(Not run)