| mvqqplot {mvMORPH} | R Documentation |
Quantile-Quantile plots for multivariate models fit with mvgls or mvols
Description
The quantile-quantile plots of the Chi square distribution is used to assess multivariate normality and detect outliers using the squared Mahalanobis distances from the models residuals.
Usage
mvqqplot(object, conf=0.95, ...)
Arguments
object |
A model fit obtained by the |
conf |
Confidence interval for the approximate envelope. Default is 0.95. |
... |
Graphical options. |
Details
The empirical quantiles of standardized Mahalanobis distances (Caroni 1987) estimated from models fit by mvgls (or mvols) are compared to the quantiles of a Chi square distribution with 'p' degrees of freedom (where 'p' is the number of dimensions) when models are fit by maximum likelihood (method='LL'). For penalized likelihood model fit (regularized covariance), a matching moments method is used to map the standardized Mahalanobis distances to the Chi square distribution (Clavel, in prep.). This last option is experimental and still under development.
Value
a list with components
squared_dist |
the squared Mahalanobis distances (standardized) |
chi2q |
the chi squared quantiles |
Note
Chi square Q-Q plots may be outperformed by F based Q-Q plots for identifying outliers (Hardin & Rocke 2005). The function is still under development.
Author(s)
J. Clavel
References
Caroni, C. 1987. Residuals and Influence in the multivariate linear model. Journal of the Royal Statistical Society 36(4): 365-370.
Clavel, J., Aristide, L., Morlon, H., 2019. A Penalized Likelihood framework for high-dimensional phylogenetic comparative methods and an application to new-world monkeys brain evolution. Systematic Biology 68(1): 93-116.
Clavel, J., Morlon, H. 2020. Reliable phylogenetic regressions for multivariate comparative data: illustration with the MANOVA and application to the effect of diet on mandible morphology in phyllostomid bats. Systematic Biology 69(5): 927-943.
See Also
Examples
data(phyllostomid)
# Fit a linear model by PL
fit <- mvgls(mandible~grp1, data=phyllostomid, phyllostomid$tree, model="lambda", method="PL")
# QQ plots
mvqqplot(fit, lty=2, conf=0.99)