数える

baseline characteristicsを数えるとき、Excelで'=COUNTIF(Sheet1!$XXX,"*AMI*")'などやってたけれど
同じことを R でやる

	value <- matrix(,17,3)

	value[1,1] <- "n"
	n <- dim(data)[1]
	value[1,2] <- n
	value[1,3] <- ""
...
	value[2,1] <- "AMI"
	v <- grep("AMI", data$disease, value=FALSE)
	value[2,2] <- length(v)
	value[2,3] <- sprintf("%.1f%%",length(v)/n*100)
...
	value[7,1] <- "Age"
	v<-mean(d$age,na.rm=TRUE)
	value[7,2] <- sprintf("%.1f",v);
	ve<-sd(d$age,na.rm=TRUE)/sqrt(n)
	value[7,3] <- sprintf("%.1f",ve)
...
	value[10,1] <- "IHD"
	v <- which(d$AP>0)
	value[10,2] <- length(v)
	value[10,3] <- sprintf("%.1f%%",length(v)/n*100)

で

      [,1]      [,2]    [,3]   
 [1,] "n"       "134"  ""     
 [2,] "AMI"     "25"   "18.7%"
 [7,] "Age"     "68.0" "1.1"  
[10,] "IHD"     "86"   "64.2%"

こんな表が作れましたよ