correlation - Using cor() on the entire data set versus two variables in R

Question

I have a dataset called "metals" with many variables and I want to look at each pair's spearman correlation rho. When I run cor(metals, method = "spearman", use = "complete.obs"), I get a value for, lets say, metal1 and metal2 that is .143.

However, if I run cor.test(metals$metal1,metals$metal2, method = "spearman", use = "complete.obs"), it returns .3529. The same discrepancy appears for all other pairs. Can someone explain why or if there is a fundamental difference between the two methods?

question from:https://stackoverflow.com/questions/65600697/using-cor-on-the-entire-data-set-versus-two-variables-in-r

Answer 1

The documentation of cor does not mention the possibility to use formula as input:

Arguments
x   a numeric vector, matrix or data frame.

y   NULL (default) or a vector, matrix or data frame with compatible dimensions to x. The default is equivalent to y = x (but more efficient).