CoxMK (version 0.1.1)

fit_cox_model_from_files: Step 2: Fit Cox Model from Files

Description

Implements Step 2 of the CoxMK workflow: fitting a null Cox proportional hazards model by reading phenotype and covariate data from files. This function is designed for batch processing and large-scale analysis where data is stored in separate files.

Usage

fit_cox_model_from_files(
  phenotype_file,
  covariate_file,
  output_file = NULL,
  use_spacox = TRUE
)

Value

Invisible path to the output file

Arguments

phenotype_file

Path to CSV file with columns: IID, time, status

covariate_file

Path to CSV file with columns: IID, covar1, covar2, ...

output_file

Path to RDS file to save the fitted null model (default: temporary directory)

use_spacox

Legacy parameter, kept for compatibility (ignored)

Examples

# \donttest{
# Prepare example data files
pheno_data <- data.frame(
  IID = paste0("ID", 1:100),
  time = rexp(100, 0.1),
  status = rbinom(100, 1, 0.3)
)
covar_data <- data.frame(
  IID = paste0("ID", 1:100),
  age = rnorm(100, 50, 10),
  sex = rbinom(100, 1, 0.5)
)

# Use temporary directory for file operations to comply with CRAN policies
temp_dir <- tempdir()
pheno_file <- file.path(temp_dir, "phenotype.csv")
covar_file <- file.path(temp_dir, "covariates.csv")
output_file <- file.path(temp_dir, "null_model.rds")

write.csv(pheno_data, pheno_file, row.names = FALSE)
write.csv(covar_data, covar_file, row.names = FALSE)

# Step 2: Fit null Cox model from files
fit_cox_model_from_files(
  phenotype_file = pheno_file,
  covariate_file = covar_file, 
  output_file = output_file
)

# Load the fitted model for Step 3
model_info <- readRDS(output_file)
# }