我有一个简短的Rscript，它找到了具有以下输出的cohens kappa：
Cohen Kappa and Weighted Kappa correlation coefficients and confidence boundaries
lower estimate upper
unweighted kappa 0.099 0.18 0.27
weighted kappa 0.313 0.51 0.71
Number of subjects = 100
是否可以将cohen.kappa（数据）的输出传递给grep以分配kappa估计值，将下限/上限估计值传递给类似于csh的变量（即cohen.kappa（data）| grep / awk“未加权kappa“......
以下是完整的
脚本 library(psych)
## We want R to look for command line arguments. 1 is path to data csv file. 2 is column1. 3 is column2. 4 is save file name
## example command line...
## KS sas_reporting_summary.csv myo_stress_score sas_stress_score
args <- commandArgs(TRUE)
## Now input data csv file into a data frame in R, and attach column names
data<-data.frame(read.csv(args[1]))
attach(data)
## Now build a matrix for the kappa test to use (might not need to but this seemds to work!specify the column data and save name
col1 <- data[,args[2]]
col1_lab <- args[2]
col2 <- data[,args[3]]
col2_lab <- args[3]
save <- args[4]
## Now build a matrix for the kappa test to use (might not need to but this seems to work!). Name the matrix cols with the specified names and then print on screen
## Not sure if this is needed cohen.kappa(cbind(col1,col2)) works equally well
data_mat <- matrix(c(col1,col2),ncol=2)
colnames(data_mat) = c(col1_lab,col2_lab)
row.names(data_mat) = paste(1:nrow(data_mat))
cat("\n")
data_mat
####################################################################################################
#
# Now for the stats
# This calls cohen.kappa1(x=x,w=w,n.obs=n.obs, alpha=alpha)
# x - 2xn data with categorical values from 1 to p
# w - weights: Set to be:
# 0 (on the diagonal) - unweighted kappa
# (distance from the diagonal) off the diagonal)^2 - weighted kappa
#
# n.obs = number of observations
# alpha = Probability level for confidence intervals.
#
###################################################################################################
cohen.kappa(data_mat)