Tidy summarizes information about the components of a model. A model component might be a single term in a regression, a single hypothesis, a cluster, or a class. Exactly what tidy considers to be a model component varies cross models but is usually selfevident. If a model has several distinct types of components, you will need to specify which components to return.
# S3 method for manova tidy(x, test = "Pillai", ...)
x  A 

test  One of "Pillai" (Pillai's trace), "Wilks" (Wilk's lambda), "HotellingLawley" (HotellingLawley trace) or "Roy" (Roy's greatest root) indicating which test statistic should be used. Defaults to "Pillai". 
...  Arguments passed on to

Depending on which test statistic is specified only one of pillai
,
wilks
, hl
or roy
is included.
tidy()
, stats::summary.manova()
Other anova tidiers: tidy.TukeyHSD
,
tidy.anova
, tidy.aovlist
,
tidy.aov
A tibble::tibble()
with columns:
TODO
Degrees of freedom
The twosided pvalue associated with the observed statistic.
The value of a Tstatistic to use in a hypothesis that the regression term is nonzero.
The name of the regression term.
Pillai's trace.
Wilk's lambda.
HotellingLawley trace.
Roy's greatest root.
npk2 < within(npk, foo < rnorm(24)) m < manova(cbind(yield, foo) ~ block + N * P * K, npk2) tidy(m)#> # A tibble: 8 x 7 #> term df pillai statistic num.df den.df p.value #> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1 block 5 1.03 2.56 10 24 0.0287 #> 2 N 1 0.514 5.81 2 11 0.0190 #> 3 P 1 0.0690 0.408 2 11 0.675 #> 4 K 1 0.401 3.68 2 11 0.0598 #> 5 N:P 1 0.164 1.08 2 11 0.373 #> 6 N:K 1 0.213 1.49 2 11 0.268 #> 7 P:K 1 0.00795 0.0441 2 11 0.957 #> 8 Residuals 12 NA NA NA NA NA