R-Forge Project: vegan - Community Ecology Package - News

vegan moves to GitHub and new release 2.2-0

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Thu, 18 Sep 2014 15:58:36 GMT

Vegan development has moved to GitHub. The front page of vegan is https://github.com/vegandevs/vegan. We do still update R-Forge, but that is now a secondary site. All changes are first pushed to github and then mirrored to R-Forge. This also means that R-Forge is potentially out of date, and may become functional at any time without notice.

We are just in process of preparing vegan release 2.2-0 to CRAN. Among other things, this will use permute package in all permutations, allow parallel processing in several functions, introduce a new and more powerful null model simulation, and remove rgl and scatterplot3d based functions to a separate package vegan3d. The release is scheduled for early October, 2014.

vegan moves to GitHub and new release 2.2-0

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Thu, 18 Sep 2014 15:57:20 GMT

vegan 2.0-6 in CRAN

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Mon, 11 Feb 2013 13:39:15 GMT

Vegan 2.0-6 was released in CRAN on Feb 11, 2013.

The new version contains several small bug fixes and minor improvements of existing functions. More detailed information can be found in its NEWS file at

http://vegan.r-forge.r-project.org/NEWS.html

Trying GitHub

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Sun, 26 Feb 2012 16:27:50 GMT

We have tried GitHub (https://github.com/jarioksa/vegan) as an alternative development site for vegan for a couple of weeks. At the moment, R-Forge is still the official main development environment, but some of the work happens at GitHub first. The interface between GitHub and R-Forge is not completely painless and we are going to drop one of these platforms someday, but so far we try to keep R-Forge and GitHub in sync.

The major advantage of R-Forge is its excellent R support and good CRAN connectivity. The major advantage of GitHub is its better collaborative environment which allows easier contribution to vegan. The final choice between these two platforms will finally depend on the productivity: the developers and contributors will choose which one they prefer.

Trying GitHub

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Sun, 26 Feb 2012 16:26:55 GMT

vegan 2.0-1 on CRAN

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Thu, 20 Oct 2011 12:28:23 GMT

Version 2.0-1 of vegan was published today on CRAN.

The immediate reason for the release was that R 2.14.0 will be released soon, and vegan will give a lot of annoying warnings in R 2.14.0. The sd() function will no longer accept matrix input in R 2.14.0 and vegan was adapted to this change. There are also some minor bug fixes in meandist(), metaMDS(), permatswap() and permutest() for cca() objects.

The 2.0-1 release will contain two new functions: clamtest() implements a test to classify species to habitat specialists and generalists (CLAM of Chazdon et al., Ecology 92, 1332-1343; 2011), and raupcrick() implements a version of Raup-Crick dissimilarity where species co-occurrence probabilities are based on species frequencies (Chase et al. (2011) _Ecosphere_ 2:art24).

vegan 2.0-0 published

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Fri, 09 Sep 2011 13:13:12 GMT

*Vegan* 2.0-0 was published in CRAN on September 8, 2011. This is a major release with changed dependencies, standard NAMESPACE and new functions and features. You can read more detailed news in

http://cran.r-project.org/web/packages/vegan/news.html

but here the main points:

1. *Vegan* depends on *permute* package. This package will provide new, strong schemes of restricted permutations. Currently only permutest() for betadisper() uses the new permutation infrastructure, but with time all vegan permutation will be transferred to use *permute*.

2. *Vegan* has a standard NAMESPACE which only exports the functions that are intended for users. In particular, S3 method functions are not exported, and you cannot any longer use commands line 'cca.default()', but you must only use 'cca()'. However, most *vegan* functions were intended to be visible for users, and are exported. This change may disrupt other packages, scripts etc that use *vegan*. If you think that some non-exported functions should be exported, please contact us.

3. There are several new functions and features. Most importantly, we have new non-metric multidimensional scaling (NMDS) function monoMDS which is the default engine in metaMDS. Function monoMDS can do local NMDS, hybrid MDS and metric MDS in addition to classical global NMDS, it handles tied dissimilarities adequately in heterogeneous data sets, it can handle zero and NA dissimilarities, and it is fast. The FORTRAN code for the function was written by Peter Minchin, who joined the *vegan* author team. For other new functions and features, see the vegan NEWS (link above).

Jari Oksanen

permute an R package for restricted permutations

gsimpson@users.r-forge.r-project.org (Gavin Simpson) — Thu, 08 Sep 2011 08:50:58 GMT

Yesterday, I rolled up the first public release of the *permute* package and submitted it to CRAN. This represents the first beta release of *permute* and presages significant changes to the way permutation tests are performed, and to the sorts of permutation available, in *vegan*.

*permute* allows the following permutation designs:

o free permutation (randomisation)
o permutation of spatial transect or time series data via cyclic wrapping
o permutation of spatial grids via toroidal wrapping

Each of these can be used at one or both of the block level and the within-block level. Blocks of samples are defined by the argument `strata` as has been used in *vegan* for many years. As such, *permute* allows for testing of split-plot designs. For spatial data (transect or grid), the cyclic or toroidal wrapping can be performed forwards or backwards in either spatial dimension, and the same permutation can be used for the samples in each block (for example, in the case a time series data where the samples in each block have been "exposed" to the same "time process").

The main functions in *permute* are the `shuffle()` and `shuffleSet()` functions which generate a single permutation or a set of many permutations for a stated design. Permutation designs are defined using function `permControl()`. Function `allPerms()` can be used to generate the entire set of permutations for a given design, which is useful if there are only a small number of permutations possible for the combination of permutation design and number of observations and/or complete enumeration (an exact test) is required.

With the release of *permute*, the older functions in *vegan* that implemented a previous version of these restricted permutations have been made defunct. Old code that utilised these functions should be ported to use the new functions in *permute*; help is available upon request. These changes are in *vegan* from version 2.0-0, which has also been released to CRAN.

The main reason for splitting the permutation code out from *vegan* was that by doing so we might make this code available to other package authors without having to depend on the entirety of *vegan*.

A nascent Vignette is provided with the package which aims to document the new functions and will, by version 0.8 include a section on using *permute* within functions for package authors wishing to make use of the functionality in *permute*.

If you have questions about or comments on the package, contact me or post a message in the Forums section of the R-forge website[1] for the *vegan* stable of packages/

Gavin Simpson

[1] https://r-forge.r-project.org/forum/?group_id=68

vegan birthday: ten years on CRAN, four years on R-Forge

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Tue, 06 Sep 2011 10:08:45 GMT

First CRAN release of vegan was on September 6, 2001: vegan is now ten-years-old. Vegan entered R-Forge on September 6, 2007 or four years ago.

The first vegan release was made soon after R 1.3.1 was release, and in that there were only about 120 contributed extension packages on CRAN. The first release was made to run NMDS like recommended in vegetation science. NMDS was already available in the MASS package (isoMDS), but vegan added missing pieces:

- data standardization prior to analysis (decostand, wisconsin)
- dissimilarity measures for community ecologists (vegdist)
- starting NMDS with random configurations (initMDS)
- comparing solutions from random starts (procrustes)
- post-standardization of NMDS results (postMDS)
- adding species scores to NMDS ordination (wascores)
- fitting environmental vectors to ordination (vectorfit)

Many of these functions are now incorporated in other functions such metaMDS and envfit, but they are still all parts of vegan, although not very visible. The only living fossil is wisconsin() which has survived unchanged to the latest release (initMDS was changed in rev1792 on Sep 5, 2011).

The other early major version on CRAN were:

1.2-0 (Nov 19, 2001) with decorana().
1.4-0 (May 3, 2002) which added constrained ordination methods cca() and rda().
1.6 series (from Oct 20, 2003) which added diversity analyses, capscale and metaMDS and expanded old features.
1.8 series with incremental changes and wider authorship, and finally moved to R-Forge.

Thanks to all developers, users, bug reporters and testers. Happy birthday vegan!

vegan: 1.17-12 released, 2.0 coming soon

jarioksa@users.r-forge.r-project.org (Jari Oksanen) — Thu, 18 Aug 2011 10:20:11 GMT

A minor maintenance release vegan_1.17-12 was released on August 17. Changes are many but most of them improve the stability in exceptional cases. The only new function is tolerance() which finds the species "tolerances" and sample heterogeneities from a CA or CCA result. There is a small bug fix in metaMDSrotate(), and new arguments were added to make.cepnames(), ordilabel() and swan(), and adonis() is faster.

The maintenance release was made to clean the table for a new major release (vegan 2.0) which is expected in September. This includes several new features:

- Future versions of R require packages to have NAMESPACE, and vegan 2.0 introduces a standard R NAMESPACE. A major source of disruption may be that S3 methods are not exported, and commands like anova.cca(mod) will fail because anova.cca was not exported: you should use anova(mod) instead, and let R decide which function to use. Packages or scripts using vegan may fail because of this change. The current release (1.17-12) also has a NAMESPACE, but that exports every function so that the user experience should be similar as earlier.

- Vegan will start to use new more powerful permutation routines on the 'permute' package released soon. This package will have several alternatives of restricted permutation, and will replace vegan permuted.index2 routines. The shift to 'permute' package will be gradual, and first touch only betadisper(), but all vegan permutation will move to use new routines in next gradual releases. This also means that vegan will be dependent on 'permute' package.

- Peter Minchin joins the vegan team, and he has written for us new NMDS function called monoMDS(). This will replace MASS::isoMDS() as the default ordination engine in metaMDS(). The major advantage of monoMDS() is that it can use 'weak ties' or split tied dissimilarities (particularly dissimilarity = 1 for nothing shared) which hugely improves ordination of heterogeneous data sets. In addition, monoMDS can perform local and hybrid NMDS. It is even faster than the old engine. Some more capabilities will be taken into use in gradual releases, such as adding new points to existing ordinations.

- Another new set of functions will be fitspecaccum() that can fit some non-linear regression models to species accumulation results from specaccum(). In addition, vegan will have several new self-starting non-linear regression models that are used in fitspecaccum(), but can also be used separately. These include Arrhenius, Gleason, Gitay and Lomolino models.

Most of these features are already available in the R-Forge version of vegan (1.90-0 and above). Please test the new features and report the problems to us.

R-Forge Project: vegan - Community Ecology Package - News

vegan moves to GitHub and new release 2.2-0

vegan moves to GitHub and new release 2.2-0

vegan 2.0-6 in CRAN

Trying GitHub

Trying GitHub

vegan 2.0-1 on CRAN

vegan 2.0-0 published

*permute* an R package for restricted permutations

vegan birthday: ten years on CRAN, four years on R-Forge

vegan: 1.17-12 released, 2.0 coming soon

permute an R package for restricted permutations