My problem is to evaluate several conditional moments nonparametrically.
That is to evaluate E(y|x=x_0) and Var(y|x=x_0) with y a 2x1 variable.
I have tried several commands in "np" package, like "npregbw". I cannot
find a command that can apply to the case with multivariate dependent
variables, i.e. y in my case. I cannot do (y1, y2) separately because
finally I also need to evaluate Var(y|x=x_0).
My thought is to do this:
1. evaluate E(y1|x=x_0) and E(y2|x=x_0) separately;
2. Var(y|x=x_0)=E(yy'|x=x_0)-E(y|x=x_0)E(y|x=x_0)', so I evaluate
E(yy'|x=x_0) by evaluating E(y1y1|x=x_0), E(y1y2|x=x_0), E(y2y2|x=x_0)
separately. Then put them in the right order to form a 2X2 matrix.
Am I right if I do this? And is there direct way to do this?
Thank you for suggestions.
--
Best Regards,
Ying Zheng
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
