Is there a multiple change point analysis for both the mean and the variance simultaneously, implemented in R without a distribution assumption? I know the changepoint packet http://cran.r-project.org/web/packages/changepoint/index.html but as far as I see, it only allows change point analysis (mean & variance simultaneously) with distribution assumptions (normal, exponential, gamma).
There may be some more general non-parametric methods in the cpm package: http://cran.r-project.org/web/packages/cpm/index.html. The manual associated with the package lists three methods that are supposedly used to check for more general distributional changes (such as simultaneous mean and variance changes). However, I haven't spent a lot of time on change-point analysis, so I can't speak for the quality of the package.
There is also the
ecp package which does more general nonparametric distributional changes and even allows for multivariate data: