Provides functions to estimate the size-controlled phenotypic integration index, a novel method by Torices & Méndez (2014) to solve problems due to individual size when estimating integration (namely, larger individuals have larger components, which will drive a correlation between components only due to resource availability that might obscure the observed measures of integration). In addition, the package also provides the classical estimation by Wagner (1984), bootstrapping and jackknife methods to calculate confidence intervals and a significance test for both integration indices.
Released Jun 02, 2017
.jackknife and bootstrap functions now report the resampled values, if requested using the verbose=T option.
Released Apr 27, 2016
.Modifies integration estimation by dividing between the population variance instead of the sample variance.
Released Mar 3, 2016
.Includes manuscript citation for users
Released Apr 28, 2015
.Includes a function to estimate significance of the integration index using jackknife. Contrasting with bootstrap, this method does not repeat observations, thus reducing biases on the covariance matrix.
Released Jan 27, 2015
.Includes a function to estimate significance of the integration index
.Acronym for integration is modified from INT to PINT to avoid confusion with the fundamental programming meaning of an Integer type class
.Estimation of the percentage corrected index ('INTsc.c') is removed.
.New order for arguments in 'pintsc.boot' to be the same in all functions.
.Functions do not print results as a table after computation.
Released Oct 24, 2014
First "official" release of the package including functions for estimating conventional and size-controlled phenotypic integration indices and its confidence intervals by bootstrapping.