rowCoxTests {simulatorZ} | R Documentation |
method for performing Cox regression
rowCoxTests(X, y, option = c("fast", "slow"), ...)
X |
Gene expression data. The following formats are available: matrix Rows correspond to observations, columns to variables. data.frame Rows correspond to observations, columns to variables. ExpressionSet rowCoxTests will extract the expressions using exprs(). |
y |
Survival Response, an object of class: Surv if X is of type data.frame or matrix character if X is of type ExpressionSet. In this case y is the name of the survival response in the phenoData of X. If survival time and indicator are stored separately in the phenoData one can specify a two-element character vector the first element representing the survival time variable. |
option |
"fast" loops over rows in C, "slow" calls coxph directly in R. The latter method may be used if something goes wrong with the "fast" method. |
... |
currently unused |
dataframe with two columns: coef = Cox regression coefficients, p.value = Wald Test p-values. Rows correspond to the rows of X.
Yuqing Zhang, Christoph Bernau, Levi Waldron
#test ##regressor-matrix (gene expressions) X<-matrix(rnorm(1e6),nrow=10000) #seed set.seed(123) #times time<-rnorm(n=ncol(X),mean=100) #censoring(1->death) status<-rbinom(n=ncol(X),size=1, prob=0.8) ##survival object y<-Surv(time,status) ## Do 10,000 Cox regressions: system.time(output <- rowCoxTests(X=X,y=y, option="fast"))