Tidying methods for mixed effects models

These methods tidy the coefficients of mixed effects models of the lme class from functions of the nlme package.

Usage

# S3 method for class 'lme'
tidy(
  x,
  effects = c("var_model", "ran_pars", "fixed"),
  scales = NULL,
  conf.int = FALSE,
  conf.level = 0.95,
  ...
)

# S3 method for class 'lme'
augment(x, data = x$data, newdata, ...)

# S3 method for class 'lme'
glance(x, ...)

# S3 method for class 'gls'
tidy(x, conf.int = FALSE, conf.level = 0.95, ...)

# S3 method for class 'gls'
augment(x, data = nlme::getData(x), newdata, ...)

Arguments

x: An object of class lme, such as those from lme or nlme
effects: One or more of "var_model", "ran_pars", "fixed", "ran_vals", and/or "ran_coefs".
scales: scales on which to report the variables: for random effects, the choices are ‘"sdcor"’ (standard deviations and correlations: the default if scales is NULL) or ‘"vcov"’ (variances and covariances). NA means no transformation, appropriate e.g. for fixed effects.
conf.int: whether to include a confidence interval
conf.level: confidence level for CI
...: extra arguments (not used)
data: original data this was fitted on; if not given this will attempt to be reconstructed
newdata: new data to be used for prediction; optional

Value

All tidying methods return a data.frame without rownames. The structure depends on the method chosen.

tidy returns one row for each estimated effect, either random or fixed depending on the effects parameter. If effects = "ran_vals" (or "ran_pars"), it contains the columns

group: the group within which the random effect is being estimated
level: level within group
term: term being estimated
estimate: estimated coefficient
estimated: This column is only included if some parameters are fixed. TRUE if the residual error is estimated and FALSE if the residual error is fixed.

If effects="fixed", tidy returns the columns

term: fixed term being estimated
estimate: estimate of fixed effect
std.error: standard error
statistic: t-statistic
p.value: P-value computed from t-statistic

If effects="var_model" (the weights argument to the model), tidy returns the columns defined in the help for tidy.varFunc.

augment returns one row for each original observation, with columns (each prepended by a .) added. Included are the columns

.fitted: predicted values
.resid: residuals
.fixed: predicted values with no random effects

glance returns one row with the columns

sigma: the square root of the estimated residual variance
logLik: the data's log-likelihood under the model
AIC: the Akaike Information Criterion
BIC: the Bayesian Information Criterion
deviance: returned as NA. To quote Brian Ripley on R-help https://stat.ethz.ch/pipermail/r-help/2006-May/104744.html, "McCullagh & Nelder (1989) would be the authorative [sic] reference, but the 1982 first edition manages to use 'deviance' in three separate senses on one page."

Details

When the modeling was performed with na.action = "na.omit" (as is the typical default), rows with NA in the initial data are omitted entirely from the augmented data frame. When the modeling was performed with na.action = "na.exclude", one should provide the original data as a second argument, at which point the augmented data will contain those rows (typically with NAs in place of the new columns). If the original data is not provided to augment and na.action = "na.exclude", a warning is raised and the incomplete rows are dropped.

Examples


if (require("nlme") && require("lme4")) {
    data("sleepstudy", package="lme4")
    ## original model
    if (FALSE) { # \dontrun{
         lmm1 <- lme(Reaction ~ Days, random=~ Days|Subject, sleepstudy)
    } # }
    ## load stored object
    load(system.file("extdata","nlme_example.rda", package="broom.mixed"))
    tidy(lmm1)
    tidy(lmm1, effects = "fixed")
    tidy(lmm1, conf.int = TRUE)
    tidy(lmm1, conf.int = TRUE, conf.level = 0.8)
    tidy(lmm1, effects = "ran_pars")
    tidy(lmm1, effects = "ran_vals")
    tidy(lmm1, effects = "ran_coefs")
    head(augment(lmm1, sleepstudy))
    glance(lmm1)

    startvec <- c(Asym = 200, xmid = 725, scal = 350)
    nm1 <- nlme(circumference ~ SSlogis(age, Asym, xmid, scal),
                  data = Orange,
                  fixed = Asym + xmid + scal ~1,
                  random = Asym ~1,
                  start = startvec)
    tidy(nm1)
    tidy(nm1, effects = "fixed")
    head(augment(nm1, Orange))
    glance(nm1)

    gls1 <- gls(follicles ~ sin(2*pi*Time) + cos(2*pi*Time), Ovary,
                         correlation = corAR1(form = ~ 1 | Mare))
    tidy(gls1)
    glance(gls1)
    head(augment(gls1))
}
#> Loading required package: nlme
#> 
#> Attaching package: ‘nlme’
#> The following object is masked from ‘package:lme4’:
#> 
#>     lmList
#> The following object is masked from ‘package:glmmADMB’:
#> 
#>     VarCorr
#> The following object is masked from ‘package:dplyr’:
#> 
#>     collapse
#> Warning: ran_pars not yet implemented for nonlinear or multilevel models
#> # A tibble: 6 × 5
#>   Mare     Time follicles .fitted .resid
#>   <ord>   <dbl>     <dbl>   <dbl>  <dbl>
#> 1 1     -0.136         20   13.7  6.28  
#> 2 1     -0.0909        15   13.0  2.04  
#> 3 1     -0.0455        19   12.1  6.87  
#> 4 1      0             16   11.3  4.68  
#> 5 1      0.0455        13   10.6  2.43  
#> 6 1      0.0909        10    9.96 0.0405

Usage

Arguments

Value

Details

See also

Examples