smooth_sslCPP: Computes a smoothed projection using C++
In stepcie/sslcov: Semi-Supervised Estimation Of Covariance

Smoothing predictor using Gaussian kernel using a C++ implementation

1	smooth_sslCPP(ri, fi, fnew, wgt = NULL, bw = NULL, cdf_trans = TRUE, rsup)

`ri`	label data correlation
`fi`	the labeled data
`fnew`	the new data to be predicted
`wgt`	optionnal weights (used for perturbations). Default is `NULL` in which case no weighting is performed.
`bw`	kernel bandwith. Default is `NULL` in which case the sd of the new data divided by the total number of observation to the power 0.3 is used.
`cdf_trans`	a logical flag indicating wether the smoothing should be performed on the data transformed with their cdf. Default is `TRUE`.
`rsup`	the supervised estimate of rhat

Smoothing over the CDF transformed data preven_learns some tail estimation issues when the new data are subsequen_learnly large.

a vector of length 3 containing:

rhat.ssl the semi-supervised estimation of rhat
rhat.ssl.bc the semi-supervised estimation of rhat accounting for smoothing bias
bw the value of the bandwith actually used

stepcie/sslcov documentation built on May 30, 2019, 2:39 p.m.