Calculate R and Rsquare with missing values in R-CodePudding

This is how my data looks like:

> dput(head(NDVINS,30))
structure(list(NDVIN = c(0.473423556691161, 0.477054534599308, 
0.476981842446724, 0.534850236167682, 0.494749776839649, 0.487290542051824, 
0.520846067679237, 0.480150934504084, 0.481785217497498, 0.48192643529568, 
0.481434727772958, 0.500768986540367, 0.479271005904589, 0.471903306084303, 
0.491082681011792, 0.504112964380976, 0.47717537029362, 0.48095597078326, 
0.481278024224638, 0.49770426200673, 0.46549746851826, 0.493456827694478, 
0.478417734411996, 0.527051712023175, 0.527993175833221, 0.533868783798418, 
NA, 0.586107357729844, 0.546745185860292, 0.557100901289835), 
    NDVIN_GPR = c(0.4905, 0.4905, 0.4912, 0.4922, 0.4933, 0.4942, 
    0.4947, 0.4946, 0.4939, 0.4925, 0.4906, 0.4882, 0.4857, 0.4833, 
    0.4812, 0.4799, 0.4797, 0.4807, 0.4831, 0.4871, 0.4926, 0.4994, 
    0.5072, 0.5158, 0.5247, 0.5335, 0.5418, 0.5492, 0.5555, 0.5604
    )), row.names = c(NA, 30L), class = "data.frame")

I'm trying to calculate the R and Rsquare but since I've missing values I don't know how to do it. I've tried the following code, but had no success:

rsq <- function (NDVIN, NDVIN_GPR, na.rm=TRUE) cor(NDVIN, NDVIN_GPR) ^ 2
rsq(NDVIN, NDVIN_GPR, na.rm=TRUE)

Any help will be much appreciatted.

CodePudding user response：

In the cor, there is an argument to remove the missing values i.e. use, by default, it is "everything". We can change it to complete.obs

rsq <- function (NDVIN, NDVIN_GPR, na.rm=TRUE) cor(NDVIN, NDVIN_GPR, use = "complete.obs") ^ 2