| naresid {stats} | R Documentation |
Use missing value information to adjust residuals and predictions.
naresid(omit, x, ...) napredict(omit, x, ...)
omit |
an object produced by an |
x |
a vector, data frame, or matrix to be adjusted based upon the missing value information. |
... |
further arguments passed to or from other methods. |
These are utility functions used to allow predict,
fitted and residuals methods for modelling
functions to compensate for the removal of NAs in the fitting
process. They are used by the default, "lm", "glm" and
"nls" methods, and by further methods in packages MASS,
rpart and survival. Also used for the scores returned by
factanal, prcomp and princomp.
The default methods do nothing. The default method for the na.exclude
action is to pad the object with NAs in the correct positions to
have the same number of rows as the original data frame.
Currently naresid and napredict are identical, but
future methods need not be. naresid is used for residuals, and
napredict for fitted values, predictions and weights.
These return a similar object to x.
In the early 2000s, packages rpart and survival5 contained
versions of these functions that had an na.omit action
equivalent to that now used for na.exclude.