process_data: Process the data

In EQUALPrognosis: Analysing Prognostic Studies

Process the data

Description

This takes a dataset and the metadata for the dataset and creates R data frames in a format required for the subsequent steps.

Usage

process_data(data_file_path, metadata_file_path)

Arguments

data_file_path	Path to the dataset
metadata_file_path	Path to the metadata file

Details

The metadata should contain the following information as a minimum. variable: this is the name of the variable and should match the column names of the dataset. data_type: 'numerical' for continuous variables, 'count' for count variables, 'binary for binary categorical variables, 'nominal' for unordered categorical variables with more than 2 levels, 'ordinal' for ordered categorical variables, 'date' for variables stored as date, and 'time' for variables stored containing the time of the day.

Optional information includes the following. reference: Reference category for binary and nominal variables. This should be a category existing in the variable. ordinal levels: the levels of ordinal data from lower to higher order, separated by ";". This must include all the levels in the data.

You can use guess_data_types as a starting point for the metadata, which is included in the output list of the guess_data_types function.

Value

outcome	Whether the operation was successfully performed
message	Any information, particularly when the operation fails.
data_processed	The data which has been modifed according to the metadata when correct parameters are provided

.

any_type	All fields.
quantitative	Fields recognised as quantitative.
numerical	Fields recognised as continuous.
count	Fields recognised as count.
categorical	Fields recognised as categorical data.
nominal	Fields recognised as nominal data
binary	Fields recognised as binary data.
ordinal	Fields recognised as ordinal data.
date	Fields recognised as date.
time	Fields recognised as time.

Author(s)

Kurinchi Gurusamy

See Also

guess_data_types

Examples

library(survival)
# Use the dataset colon as example
# Select only the survival for these examples (etype == 2)
data_file_path <- paste0(tempdir(), "/df.csv")
write.csv(colon[colon$etype == 2, ], data_file_path, row.names = FALSE, na = "")
metadata <- {data.frame(
  variable = c("id","study","rx","sex","age",
               "obstruct","perfor","adhere","nodes","status",
               "differ","extent","surg","node4","time",
               "etype"),
  data_type = c("nominal", "nominal", "nominal", "binary", "numerical",
                "binary", "binary", "binary", "count", "binary",
                "ordinal", "ordinal", "binary", "binary", "numerical",
                "nominal"),
  reference = c(NA, NA, "Obs", 0, NA,
                0, 0, 0, NA, 0,
                NA, NA, 0, 0, NA,
                NA),
  ordinal_levels = c(NA, NA, NA, NA, NA,
                     NA, NA, NA, NA, NA,
                     "1;2;3", "1;2;3;4", NA, NA, NA,
                     NA),
  comments = NA
)}
metadata_file_path <- paste0(tempdir(), "/metadata.csv")
write.csv(metadata, metadata_file_path, row.names = FALSE, na = "")
processed_data <- process_data(data_file_path, metadata_file_path)

EQUALPrognosis documentation built on Feb. 4, 2026, 5:15 p.m.