Skip to main content
. Author manuscript; available in PMC: 2018 Apr 3.
Published in final edited form as: Electron J Stat. 2016 Jul 18;10(2):1807–1828. doi: 10.1214/15-EJS1032

Fig 1.

Fig 1

Illustration of the latent variable used to capture the regression dependence. In (a) we show the distribution of the conditional latent variable, Z, at various points in X assuming a log-linear dependence. In (b) we see the corresponding predictive distributions using a Gaussian mixture model for the marginal, FY, shown as the black line. The points in Z shown in (a) are mapped to the points in Y shown in (b) where the ordering is preserved.