survival_from_data: Title Get survival analysis curves from data

Description Usage Arguments Details Value

View source: R/survival_fit.R

Description

Title Get survival analysis curves from data

Usage

1
2
3
4
5
survival_from_data(location, survival_specs, dists = dists,
  save_fits = TRUE, just_load = FALSE,
  set_definitions = "set_definitions.csv",
  report_file_name = "survival_report", write_report = FALSE,
  write_excel = write_report, report_template)

Arguments

location

base directory for the analysis.

survival_specs

a data.frame containing information on how to perform each fit - see @details.

dists

the distributions to use to fit the survival function

save_fits

should fits be saved to disk? Can be useful for testing.

just_load

If TRUE, data files are ignored in favor of loading fits from fit_files

set_definitions

definitions of different subsets to fit

report_file_name

base name for survival output - Excel and the diagnostic tables and plots on the fits.

write_report

should the call produce an html report with diagnostic tables and plots on the fits?

write_excel

should the call produce a csv file summarizing fits?

report_template

If write_report == TRUE, a template for the report.

Details

By default, the function fits with seven different distribution functions: exponential, Weibull, lognormal, log-logistic, Gompertz, gamma, and generalized gamma.

write_report and write_excel are ignored if just_load = TRUE.

survival_specs contains information about how to create the fits: the directory (or directories) in which files are located, the file names, the names of the columns that should be used for time, events, and censorship status, and where and under what names to save the fits. If data_files is NULL and fit files exists, then fit_files should have the names of files in which survival models are kept. If data_files is not NULL, then survival models will be fit from the data in data_files (using the flexsurvreg package), and if fit_files is also not NULL, the survival models will be saved in fit_files.

If data_directory begins with "\", with the platform file separator (.Platform$file.sep) or with a letter and a colon (to accomodate Windows), it's considered absolute and not appended to location.

Value

a tibble with columns type, treatment, set_name,


MattWiener/heemodFits documentation built on May 19, 2019, 8:21 a.m.