get_cox_res: Run Cox Regression on a Single or Multiple Groups of Data

In survutils: Utility Functions for Survival Analysis

Description

get_cox_res is a wrapper around coxph. It can run univariate or multivariate cox regression. If the group parameter is used, then cox regression is run for each group separately.

Usage

get_cox_res(in.df, endpoint, endpoint.code, features, group = NULL,
  broom.fun = c("tidy", "glance"))

Arguments

in.df	Input data.frame.
endpoint	Column name of the endpoint.
endpoint.code	Column name of the endpoint status code.
features	Vector containing the features to run cox regression on.
group	Column name containing the groups to run cox regression on. If, specified, cox regression is run separately for each group.
broom.fun	Which broom function to run on the cox regression results.

Details

The data is returned in a broom::tidy data.frame format.

Value

Cox regression results returned in a tidy data.frame format.

Examples

library("survival")
endpoint <- "time"
endpoint.code <- "status"

# Run Univariate Cox Regression on Single Feature
features <- "age"
test.df <- get_cox_res(colon, endpoint, endpoint.code, features)

# Run Univariate Cox Regression on Multiple Features
multi.features <- c("age", "obstruct")
get_cox_res(colon, endpoint, endpoint.code, multi.features)

# Run Univariate Cox Regression on Multiple Features For Each rx group
group <- "rx"
get_cox_res(colon, endpoint, endpoint.code, multi.features, group)

# Run Multivariate Cox Regression 
get_cox_res(colon, endpoint, endpoint.code, multi.features)

# Run Multivariate Cox Regression For Each rx Group
get_cox_res(colon, endpoint, endpoint.code, multi.features, group)

Example output

Detected only one feature. Running univariate cox regression
Detected multiple features. Running multivariate cox regression
      term  estimate   std.error  statistic     p.value  conf.low conf.high
1      age 0.9983432 0.002804002 -0.5913434 0.554290364 0.9928717  1.003845
2 obstruct 1.2677379 0.080804466  2.9359039 0.003325774 1.0820531  1.485287
  test_type
1  multicox
2  multicox
Detected multiple features. Running multivariate cox regression
    group     term  estimate   std.error  statistic    p.value  conf.low
1     Obs      age 1.0026174 0.004603197  0.5678581 0.57013133 0.9936124
2     Obs obstruct 1.2123725 0.131970513  1.4592591 0.14449382 0.9360576
3     Lev      age 1.0042268 0.004875449  0.8651343 0.38696506 0.9946764
4     Lev obstruct 1.4151910 0.129394314  2.6837694 0.00727973 1.0981822
5 Lev+5FU      age 0.9869403 0.005186566 -2.5345800 0.01125822 0.9769584
6 Lev+5FU obstruct 1.0844978 0.167766933  0.4835103 0.62873343 0.7805940
  conf.high test_type
1 1.0117040  multicox
2 1.5702528  multicox
3 1.0138689  multicox
4 1.8237097  multicox
5 0.9970242  multicox
6 1.5067186  multicox
Detected multiple features. Running multivariate cox regression
      term  estimate   std.error  statistic     p.value  conf.low conf.high
1      age 0.9983432 0.002804002 -0.5913434 0.554290364 0.9928717  1.003845
2 obstruct 1.2677379 0.080804466  2.9359039 0.003325774 1.0820531  1.485287
  test_type
1  multicox
2  multicox
Detected multiple features. Running multivariate cox regression
    group     term  estimate   std.error  statistic    p.value  conf.low
1     Obs      age 1.0026174 0.004603197  0.5678581 0.57013133 0.9936124
2     Obs obstruct 1.2123725 0.131970513  1.4592591 0.14449382 0.9360576
3     Lev      age 1.0042268 0.004875449  0.8651343 0.38696506 0.9946764
4     Lev obstruct 1.4151910 0.129394314  2.6837694 0.00727973 1.0981822
5 Lev+5FU      age 0.9869403 0.005186566 -2.5345800 0.01125822 0.9769584
6 Lev+5FU obstruct 1.0844978 0.167766933  0.4835103 0.62873343 0.7805940
  conf.high test_type
1 1.0117040  multicox
2 1.5702528  multicox
3 1.0138689  multicox
4 1.8237097  multicox
5 0.9970242  multicox
6 1.5067186  multicox

survutils documentation built on May 2, 2019, 6:38 a.m.