logrnk: Performs Log Rank test on the long and short patient sets

In RCASPAR: A package for survival time prediction based on a piecewise baseline hazard Cox regression model.

Description

This function performs a Chi-square test on the long and short subject sets to determine if their is a significant difference between the survival times in both sets. It returns the p-value.

Usage

logrnk(dataL, dataS)

Arguments

dataL	The set of subjects predicted to fall into the long-survivor set. A data frame containing at least the following columns: “PatientOrderValidation” (the number/order of the subject); “group” (the group into which the patient falls L (for long) or S(for short)); “censored” (the censorship status of the patient $1$ for uncensored and $0$ for censored).
dataS	Same as “dataL” but for the set of short survivors.

Details

Note that the typical arguments to be passed are the results of the “STpredict” functions “long\_survivors” and “long\_survivors”

Value

The estimated p-value is returned

Author(s)

Douaa AS Mugahid

References

Bland JM, Altman DG. Survival probabilities (the Kaplan-Meier method). BMJ 2004;328;1073 http://www.bmj.com/statsbk/12.dtl

See Also

survivAURC

Examples

PatientOrderValidation_L <- c(1, 2, 3, 5, 7)
PatientOrderValidation_S <- c(4, 6, 8)
group_L <- rep("L", 5)
group_S <- rep("S", 3)
censored_L <- c(0, 0, 1, 1, 0)
censored_S <- c(0, 0, 1)
True_STs_L <- c(5, 6, 6 ,7, 8)
True_STs_S <- c(2, 3, 2)
short <- as.data.frame(cbind(PatientOrderValidation_S, group_S, censored_S, True_STs_S))
long <- as.data.frame(cbind(PatientOrderValidation_L, group_L, censored_L, True_STs_L))
names(short) <- c("PatientOrderValidation", "group", "censored", "True_STs")
names(long) <- c("PatientOrderValidation", "group", "censored", "True_STs")
logrnk(dataL=long, dataS=short)

RCASPAR documentation built on Nov. 8, 2020, 6:56 p.m.