R Loop: Perform a spearman correlation for Every 12 Rows-CodePudding

I have a matrix which contains 984 rows and 2 columns (M and B), I want to do a spearman correlation of the 2 columns for every 12 rows, so in total 82 correlation. This is what I have

df ->read(......)
M = df$M
B = df$B

cnt <- 0
for (i in seq(1, nrow(df), by = 12)) {
M = df$[M,cnt]
B = df$[B, cnt]
resultM-B = cor(M, B, method = "spearman")
cat("Spearman correlation MB", resultM-B,"\n")
cnt <- cnt   12
}

Thanks

CodePudding user response：

Looks like a job for by

Make up an indexing factor

idx <- ceiling((1:nrow(df))/12)

Then tell by to use that as a guide:

res <- by(df, idx, function(x){cor(x$M, x$B, method ="spearman")})

CodePudding user response：

For dummy data with 24 rows, you may try

df1 <- data.frame(
  M = rnorm(24, 1, 1),
  B = rnorm(24, 1, 5)
)

cnt <- 12
res <- c()
for (i in seq(1, nrow(df1), by = 12)){
  M1 <- df1$M[i:cnt]
  B1 <- df1$B[i:cnt]
  res_MB <- cor(M1, B1, method = "spearman")
  res <- c(res, res_MB)
  cnt <- cnt   12
}
res
[1] -0.4755245 -0.2727273