check_model {performance} | R Documentation |
Visual check of model various assumptions (normality of residuals, normality of random effects, linear relationship, homogeneity of variance, multicollinearity).
check_model(x, ...) ## Default S3 method: check_model( x, dot_size = 2, line_size = 0.8, panel = TRUE, check = "all", alpha = 0.2, dot_alpha = 0.8, colors = c("#3aaf85", "#1b6ca8", "#cd201f"), theme = "see::theme_lucid", detrend = FALSE, show_dots = NULL, verbose = TRUE, ... )
x |
A model object. |
... |
Currently not used. |
dot_size, line_size |
Size of line and dot-geoms. |
panel |
Logical, if |
check |
Character vector, indicating which checks for should be performed
and plotted. May be one or more of |
alpha, dot_alpha |
The alpha level of the confidence bands and dot-geoms. Scalar from 0 to 1. |
colors |
Character vector with color codes (hex-format). Must be of length 3. First color is usually used for reference lines, second color for dots, and third color for outliers or extreme values. |
theme |
String, indicating the name of the plot-theme. Must be in the
format |
detrend |
Should QQ/PP plots be detrended? |
show_dots |
Logical, if |
verbose |
Toggle off warnings. |
For Bayesian models from packages rstanarm or brms,
models will be "converted" to their frequentist counterpart, using
bayestestR::bayesian_as_frequentist
.
A more advanced model-check for Bayesian models will be implemented at a
later stage.
The data frame that is used for plotting.
The plot Linearity checks the assumption of linear relationship.
However, the spread of dots also indicate possible heteroscedasticity (i.e.
non-constant variance); hence, the alias "ncv"
for this plot.
Some caution is needed when interpreting these plots. Although these
plots are helpful to check model assumptions, they do not necessarily
indicate so-called "lack of fit", e.g. missed non-linear relationships or
interactions. Thus, it is always recommended to also look at
effect plots, including partial residuals.
Plots that check the normality of residuals (QQ-plot) or the homogeneity of
variance use standardized Pearson's residuals for generalized linear models,
and standardized residuals for linear models. The plots for the normality of
residuals (with overlayed normal curve) and for the linearity assumption use
the default residuals for lm
and glm
(which are deviance
residuals for glm
).
For models with many observations, or for more complex models in general,
generating the plot might become very slow. One reason might be that the
underlying graphic engine becomes slow for plotting many data points. In
such cases, setting the argument show_dots = FALSE
might help. Furthermore,
look at the check
argument and see if some of the model checks could be
skipped, which also increases performance.
This function just prepares the data for plotting. To create the plots,
see needs to be installed. Furthermore, this function suppresses
all possible warnings. In case you observe suspicious plots, please refer
to the dedicated functions (like check_collinearity()
,
check_normality()
etc.) to get informative messages and warnings.
## Not run: m <- lm(mpg ~ wt + cyl + gear + disp, data = mtcars) check_model(m) if (require("lme4")) { m <- lmer(Reaction ~ Days + (Days | Subject), sleepstudy) check_model(m, panel = FALSE) } if (require("rstanarm")) { m <- stan_glm(mpg ~ wt + gear, data = mtcars, chains = 2, iter = 200) check_model(m) } ## End(Not run)