Hi everybody,
I have a data bank in long format, with 3 measures in different locations in the same patient.
I need to calc something like: an average of the correlations among these 3 measures within individual.
For that, I found the rmcorr package (it's like an ANCOVA but the ID is the factor), but doesn't work.
Code:
Bank.out <- rmcorr(Bank$ID, Bank$measureA, Bank$measureB, Bank, CIs=c("analytic", "bootstrap"), nreps=253, bstrap.out = F)
Warning messages:
1: In rmcorr(Bank$ID, Bank$measureA, Bank$measureB, Bank, :
'$' coerced into a factor 'Bank' coerced into a factor'ID' coerced into a factor
2: attempting model selection on an essentially perfect fit is nonsense
My ouput appears:
r: num NA
df: int 496
p: num NA
CI: num [1:2] NA NA
CI.level: num 0.95
I have some missings in measureA and measureB, but the argument useNA="always" was not acceptable.
Someone can help me?
Thank you!