Glance accepts a model object and returns a tibble::tibble()
with exactly one row of model summaries. The summaries are typically
goodness of fit measures, p-values for hypothesis tests on residuals,
or model convergence information.
Glance never returns information from the original call to the modeling function. This includes the name of the modeling function or any arguments passed to the modeling function.
Glance does not calculate summary measures. Rather, it farms out these
computations to appropriate methods and gathers the results together.
Sometimes a goodness of fit measure will be undefined. In these cases
the measure will be reported as NA.
Glance returns the same number of columns regardless of whether the
model matrix is rank-deficient or not. If so, entries in columns
that no longer have a well-defined value are filled in with an NA
of the appropriate type.
# S3 method for lmrob glance(x, ...)
| x | A |
|---|---|
| ... | Additional arguments. Not used. Needed to match generic
signature only. Cautionary note: Misspelled arguments will be
absorbed in |
For tidiers for robust models from the MASS package see
tidy.rlm().
Other robustbase tidiers:
augment.glmrob(),
augment.lmrob(),
tidy.glmrob(),
tidy.lmrob()
A tibble::tibble() with exactly one row and columns:
Residual degrees of freedom.
R squared statistic, or the percent of variation explained by the model. Also known as the coefficient of determination.
Estimated standard error of the residuals.
library(robustbase) # From the robustbase::lmrob examples: data(coleman) set.seed(0) m <- robustbase::lmrob(Y ~ ., data = coleman) tidy(m) #> # A tibble: 6 × 5 #> term estimate std.error statistic p.value #> <chr> <dbl> <dbl> <dbl> <dbl> #> 1 (Intercept) 30.5 6.71 4.54 4.59e- 4 #> 2 salaryP -1.67 0.431 -3.86 1.72e- 3 #> 3 fatherWc 0.0843 0.0147 5.74 5.10e- 5 #> 4 sstatus 0.668 0.0339 19.7 1.30e-11 #> 5 teacherSc 1.17 0.110 10.6 4.35e- 8 #> 6 motherLev -4.14 0.921 -4.49 5.07e- 4 augment(m) #> # A tibble: 20 × 8 #> Y salaryP fatherWc sstatus teacherSc motherLev .fitted .resid #> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1 37.0 3.83 28.9 7.2 26.6 6.19 36.8 0.191 #> 2 26.5 2.89 20.1 -11.7 24.4 5.17 26.7 -0.159 #> 3 36.5 2.86 69.0 12.3 25.7 7.04 40.7 -4.16 #> 4 40.7 2.92 65.4 14.3 25.7 7.1 41.3 -0.625 #> 5 37.1 3.06 29.6 6.31 25.4 6.15 36.3 0.768 #> 6 33.9 2.07 44.8 6.16 21.6 6.41 33.7 0.249 #> 7 41.8 2.52 77.4 12.7 24.9 6.86 42.0 -0.203 #> 8 33.4 2.45 24.7 -0.17 25.0 5.78 33.7 -0.282 #> 9 41.0 3.13 65.0 9.85 26.6 6.51 41.5 -0.466 #> 10 37.2 2.44 9.99 -0.05 28.0 5.57 36.9 0.286 #> 11 23.3 2.09 12.2 -12.9 23.5 5.62 23.7 -0.368 #> 12 35.2 2.52 22.6 0.92 23.6 5.34 34.3 0.912 #> 13 34.9 2.22 14.3 4.77 24.5 5.8 35.8 -0.924 #> 14 33.1 2.67 31.8 -0.96 25.8 6.19 32.6 0.486 #> 15 22.7 2.71 11.6 -16.0 25.2 5.62 22.4 0.266 #> 16 39.7 3.14 68.5 10.6 25.0 6.94 38.6 1.07 #> 17 31.8 3.54 42.6 2.66 25.0 6.33 33.0 -1.19 #> 18 31.7 2.52 16.7 -11.0 24.8 6.01 24.5 7.23 #> 19 43.1 2.68 86.3 15.0 25.5 7.51 42.1 1.03 #> 20 41.0 2.37 76.7 12.8 24.5 6.96 41.4 -0.367 glance(m) #> # A tibble: 1 × 3 #> r.squared sigma df.residual #> <dbl> <dbl> <int> #> 1 0.981 1.13 14 # From the robustbase::glmrob examples: data(carrots) Rfit <- glmrob(cbind(success, total - success) ~ logdose + block, family = binomial, data = carrots, method = "Mqle", control = glmrobMqle.control(tcc = 1.2) ) tidy(Rfit) #> # A tibble: 4 × 5 #> term estimate std.error statistic p.value #> <chr> <dbl> <dbl> <dbl> <dbl> #> 1 (Intercept) 2.39 0.692 3.45 0.000561 #> 2 logdose -2.05 0.368 -5.56 0.0000000268 #> 3 blockB2 0.235 0.212 1.11 0.268 #> 4 blockB3 -0.450 0.241 -1.87 0.0620 augment(Rfit) #> # A tibble: 24 × 4 #> `cbind(success, total - success)`[,"success"] [,""] logdose block .fitted #> <int> <int> <dbl> <fct> <dbl> #> 1 10 25 1.52 B1 -0.726 #> 2 16 26 1.64 B1 -0.972 #> 3 8 42 1.76 B1 -1.22 #> 4 6 36 1.88 B1 -1.46 #> 5 9 26 2 B1 -1.71 #> 6 9 33 2.12 B1 -1.96 #> 7 1 31 2.24 B1 -2.20 #> 8 2 26 2.36 B1 -2.45 #> 9 17 21 1.52 B2 -0.491 #> 10 10 30 1.64 B2 -0.737 #> # … with 14 more rows