Augment accepts a model object and a dataset and adds
information about each observation in the dataset. Most commonly, this
includes predicted values in the
.fitted column, residuals in the
.resid column, and standard errors for the fitted values in a
column. New columns always begin with a
. prefix to avoid overwriting
columns in the original dataset.
Users may pass data to augment via either the
data argument or the
newdata argument. If the user passes data to the
it must be exactly the data that was used to fit the model
object. Pass datasets to
newdata to augment data that was not used
during model fitting. This still requires that all columns used to fit
the model are present.
Augment will often behavior different depending on whether
newdata is specified. This is because there is often information
associated with training observations (such as influences or related)
measures that is not meaningfully defined for new observations.
For convenience, many augment methods provide default
augment(fit) will return the augmented training data. In these
cases augment tries to reconstruct the original data based on the model
object, with some varying degrees of success.
The augmented dataset is always returned as a tibble::tibble with the
same number of rows as the passed dataset. This means that the
passed data must be coercible to a tibble. At this time, tibbles do not
support matrix-columns. This means you should not specify a matrix
of covariates in a model formula during the original model fitting
process, and that
survival::Surv() objects are not supported in input data. If you
encounter errors, try explicitly passing a tibble, or fitting the original
model on data in a tibble.
We are in the process of defining behaviors for models fit with various na.action arguments, but make no guarantees about behavior when data is missing at this time.
# S3 method for rma augment(x, ...)
Additional arguments. Not used. Needed to match generic
signature only. Cautionary note: Misspelled arguments will be
tibble::tibble() with columns:
Upper bound on confidence interval for fitted values.
Lower bound on confidence interval for fitted values.
Upper bound on credible interval for fitted values.
Lower bound on credible interval for fitted values.
Fitted or predicted value.
In meta-analysis, the moderators used to calculate the predicted values.
In meta-analysis, the level of the moderators used to calculate the predicted values.
The difference between fitted and observed values.
Standard errors of fitted values.
The observed values for the individual studies
library(metafor)#>#>#> #>#>#> #>#>#> #>#>#> #>df <- escalc( measure = "RR", ai = tpos, bi = tneg, ci = cpos, di = cneg, data = dat.bcg ) meta_analysis <- rma(yi, vi, data = df, method = "EB") augment(meta_analysis)#> # A tibble: 13 x 6 #> .observed .fitted .se.fit .conf.low .conf.high .resid #> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1 -0.889 -0.801 0.411 -1.61 0.00524 -0.174 #> 2 -1.59 -1.26 0.354 -1.95 -0.561 -0.870 #> 3 -1.35 -0.990 0.437 -1.85 -0.134 -0.633 #> 4 -1.44 -1.40 0.138 -1.67 -1.13 -0.727 #> 5 -0.218 -0.287 0.212 -0.701 0.128 0.497 #> 6 -0.786 -0.785 0.0823 -0.946 -0.623 -0.0711 #> 7 -1.62 -1.25 0.370 -1.97 -0.523 -0.906 #> 8 0.0120 0.00301 0.0626 -0.120 0.126 0.727 #> 9 -0.469 -0.506 0.221 -0.939 -0.0740 0.246 #> 10 -1.37 -1.25 0.246 -1.73 -0.767 -0.656 #> 11 -0.339 -0.353 0.110 -0.568 -0.139 0.376 #> 12 0.446 -0.281 0.460 -1.18 0.621 1.16 #> 13 -0.0173 -0.145 0.244 -0.623 0.333 0.698