摘要: Partial residual plots have a long history and, judging from their prominence in the literature, are frequently used. In this article, I explore structure and usefulness of partial augmented as basic tools for dealing with curvature function selected covariates x 2 regression problems which covariate vector is partitioned T = (x 1 , ). The these obtaining good impression can depend on behavior through conditional expectation E(x 1|x 2). seem to perform best under linear expectations. Augmented allow 2) be quadratic 2. This development leads new class plots, called CERES that includes special cases. may useful an when expectations ...