fit_models: fit_models

In ameztegui/gradients: A package to assess the role of environmental variables on plant traits across ecological gradients

Description

fits several ecological models to test the effect of environmental variables on species traits across ecological gradients

Usage

fit_models(data, dep_var, ind_var, modelname, species = NULL,
  size_var = NULL, sites = NULL, rep_vectors, max_iter = 15000)

Arguments

data	The dataframe that contains the data
dep_var	numeric. The name of the column in 'data' that holds the trait for which to fit the predictive models
ind_var	numeric. The name of the column in 'data' that holds the numeric environmental variable (light, nutrients) that will act as main predictor of the response variable 'dep_var'
modelname	name of the object that contains the scientific function to be fit. User must provide also a list containing the initial values of the parameters of the model, the highest and the lowest possible values for those parameters. The names of these lists must be exactly the same as 'modelname' followed by the suffixes '.par', '.par.hi' and 'par.lo', respectively.
species	factor. Name of the species for which the model is gonna be fitted. If 'species = NULL' then the model is fit for all the dataframe together.
size_var	numeric. Name of the column in 'data' that contains the size of plants. If 'size = NULL' no effect of size is included in the model fitted.
sites	factor. Name of the column in 'data' that contains the name of the different sites along the gradient. This can be different elevations, different latitude, or different climate, for example. If 'sites = NULL' the function will fit one model for for all sites pooled together. Otherwise, the parameters of the model indicated in 'rep_vectors' will be allowed to take a different value for each site.
rep_vectors	vector. In the parameter list (provided by the user), position of the parameters that will be allowed to vary across sites. If sites != NULL and rep_vectors = NULL, all parameters in the function will be allowed to vary across sites.
max_iter	maximum number of iterations allowed for the 'anneal' function. By default 'max_iter = 15000'.

Value

This function fit a user-given scientific function to assess the effect of an environmental variable ('ind_var') on a given plant trait ('dep_var') across an environmental gradient of a second variable ('sites'). The model uses the function ànneal'in the 'likelihood'package to fit the model, and returns a list containing: - the best estimates of the parameters in the function - the upper limits - the lower limits - LL - AICc - slope -R2 - k

(please see the documentation for the 'anneal'function in the 'likelihood' package for more info in the output of this function.

Examples

data(ecophysio)
rdpi(ecophysio,sp,SB, Piso)

ameztegui/gradients documentation built on Nov. 2, 2019, 1:46 p.m.