Optimal kernel widths output by ma are employed to
recompute the weighted joint distribution for two variables in
a data set, and a contour plot for this distribution is drawn.
Usage
pd(d,iv=1,jv=2)
Arguments
d
an n x m data frame with m > 1.
iv
the column index of the independent variable
jv
the column index of the dependent variable
Value
An n x n distribution of weighted likelihoods is returned.
Details
A data set of two variables is extracted from the user's data set and a
full distribution is calculated
using weighted marginal and joint likelihoods. The optimal kernel sizes and
weighting are first computed via a call to ma.