ACCUEIL / HOME

All the scripts described below can also be downloaded from the folder above : Pages & Files / scripts /

Manipulation d'arbres de distances :

• 'stickTips' : function to stick phylogenetic subtrees (eg : sticking genus to a family tree)

Please use now the proper versioning available in github for use and request permissions for modifications https://github.com/iboulangeat/stickTips 

(MAJ 17 décembre 2014) : update with new birth-death function, version 2.0

(MAJ 12 mai 2010) :  Inversion 2 steps in the function (prune after sticking) allows the tree to keep ultrametricity

(MAJ 17 mai 2010) :  Add an option for the subtree type. It may be a random tree instead of a polytomy

(MAJ 20 mai 2010) : debug some problem with new options (about the root edge length of the subtree) and add another subtree option (birthdeath tree)

(MAJ 26 mai 2010) : modify root edge as random (brownian) for the subtree in option "brownian" (old called "birthdeath")

(MAJ 07 janvier 2011) : passage à une nomenclature en version : version 1.0

(MAJ 07 janvier 2011) : version 1.1

(MAJ 01 juin 2011) : description des arguments, passage version html, version 1.2

• 'edgeAttributes' : une fonction qui permet de passer d'attributs des noeuds aux branches (ex : pour colorer les branches selon la valeur du noeud amont ou aval)

Phylogenetics :

• 'evomodels' : function to fit different models for the evolution of a continuous trait in a phylogenetic tree

• 'multiphylotests2' : function to run 3 different calculations for phylogenetic signal and the associated p-values for a continuous trait

• 'subTree' : Create subtree from larger phylogeny using a vector of species names

Community ecology :

• 'RAO_16.12.2011' : The Rao function computes alpha, gamma and beta-components for taxonomic, functional and phylogenetic diversity with the Rao index. The script integrates two functions: "Qdecomp", by Villeger & Mouillot (J Ecol, 2008) and "disc", by S. Pavoine, in the package ade4. The function was introduced in the following publication:

  de Bello F., Lavergne S., Meynard C., Lepš I. & Thuiller W. (2010) The spatial partitioning of diversity: showing Theseus a way out of the labyrinth. Journal of Vegetation Science, 21, 992-1000. PDF

• 'MDNC_MDNN_Invasive' : Function to estimate the distance (phylogenetic, functional, ecological) between some species of interest (invasive, rare) and their co-occuring species in the communities. It estimates the alpha niche of a selected species (mean (weighted) distance between the species and the community). It also estimates the distance between the species of interest and the most abundant species in each community.

• 'Extract_species' : une fonction qui permet d'extraire les coordonnées d'une espèce à partir de données de répartition, de communautés.;;

• 'OMI_CCA_RDA' : function to run three different constrained ordinations in an automatic way (CCA, OMI and RDA)

• 'Niche_properties' : Function to extract species niche characteristics from any constrained ordination (CCA, RDA or OMI).

• 'corratio' : a measure of the relationship between the statistical dispersion within individual categories and the dispersion across the whole population or sample

• 'Overlap_fonctions' : functions to calculate niche overlaps between couple of species in an environmental space

GIS and spatial analyses:

• 'sampleRaster_Fixed' : function who takes a raster/grid/ASCII file as input and extract a random selection of points (size). The function returns the X and Y coordinates of the random selection. The function also controls for NODATA values. The random selection is repeated with NODATA deletion until the required number of points is extracted (not the case of the function sampleRaster)

• 'sampleRaster' :function who taks a raster/grid/ASCII file as input and extract a random selection of points (size). The function returns the X and Y coordinates of the random selection. The function DOES NOT controls for NODATA values. NODATA are removed from the selection leading potentially to a lower number of points than required (see sampleRaster_Fixed function for a solution)

• 'Tips and tricks' : These are a collection of tips and tricks for a appropriate use of the raster package in R (mask, crop, sample). 

Divers statistics :

• 'Person_Cor_Permut' : Cette fonction realise, par permutation, un test du coefficient de correlation de Pearson entre deux vecteurs

• 'RV_Coefficient' : function to calculate RV coefficient for two matrices

• 'RandomSkewers' : script with two functions to run a Random Skewers analysis as described in Cheverud 1996 J. Evol. Biol. This method can be used to compare any set of VCV or correlation matrices.

• 'Combin_pvalues' : function to combine significance levels from multiple experiments or analyses. Kind of Bonferroni but less conservative and more exact.

• 'Nagelkerke R2': function to estimate pseudo-Rsquare from either LM, GLM or GAM. For logLik models, it is a ratio of deviances and ranges from 0 to 1 (analogy to the proper R-square from OLS).  

• 'VarPart': function implementing a variance partitioning method based on Nagelkerke R-square estimations. Works for GLM and GAM. The function returns the variance explained (R2) by each component and the overall unexplained variance.
  The functions also take two other arguments : if the R-squares have to be scaled and/or adjusted. 
  Several versions :
• VarPart2mod (use VarPartUnlimited instead) : Takes three arguments, a model object 1 (a model with a set of variables), a model object 2 (a model with another set of variables) and a model object 12 (a model with ALL variables).
• VarPart3mod (use VarPartUnlimited instead) : Takes 7 arguments, a model object with variable 1, a model object with variable 2, a model object with variable 3, a model object with variables 1 and 2, a model object with variables 1 and 3, a model object with variables 2 and 3, and a model object with ALL variables (1, 2 and 3).
• VarPart (use VarPartUnlimited instead) : Takes 1 argument : a list with all single models. For example, for 3 variables, a list with a model object with variable 1, a model object with variable 2, and a model object with variable 3. LIMITED TO 4 VARIABLES MAX
• VarPartUnlimited : Takes 1 argument : a list with all single models. For example, for 3 variables, a list with a model object with variable 1, a model object with variable 2, and a model object with variable 3. THEORETICALLY UNLIMITED NUMBER OF VARIABLES

ACCUEIL / HOME