phenology (version 5.1)

map_phenology: Generate a likelihood map varying Phi and Delta.

Description

This function generates a map of likelihood varying Phi and Delta. Parameters are the same than for the fit_phenology() function except for trace that is disabled. If Alpha, Beta or Tau are not indicated, Alpha and Tau are set to 0 and 1 and Beta is fitted. Only one set of Alpha, Beta, Tau, Phi and Delta are used for all timeseries present in data. Note that it is possible to fit or fixed Alpha[n], Beta[n], Tau[n], Phi[n] and Delta[n] with [n]=1 or 2 and then it is possible to use this function to establish the likelihood map for a second or third sinusoids added to the global pattern. If Delta is not specified, it is estimated from Phi and the same precision as Phi is used.

Usage

map_phenology(data = NULL, parametersfit = NULL, parametersfixed = NA, Phi = seq(from = 0.2, to = 20, length.out = 100), Delta = NULL, infinite = 50, method_incertitude = "convolution", zero_counts = TRUE, progressbar = TRUE)

Arguments

data

dataset generated with add_format

parametersfit

Set of parameters to be fitted

parametersfixed

Set of fixed parameters

Phi

Phi values to be analyzed

Delta

Delta value to be analyzed

infinite

Number of iterations for dmnbinom() used for method_incertitude='convolution'

method_incertitude

'combinatory' estimates likelihood of all combinations for nest numbers; 'convolution' [default] uses the exact likelihood of the sum of negative binomial distribution.

zero_counts

Example c(TRUE, TRUE, FALSE) indicates whether the zeros have been recorded for each of these timeseries. Defaut is TRUE for all.

progressbar

If FALSE, do not show the progress bar

Value

Display a likelihood map

Details

map_phenology generates a likelihood map.

Examples

library("phenology")
# Read a file with data
## Not run: 
# Gratiot<-read.delim("http://max2.ese.u-psud.fr/epc/conservation/BI/Complete.txt", header=FALSE)
# ## End(Not run)
data(Gratiot)
# Generate a formatted list nammed data_Gratiot 
data_Gratiot<-add_phenology(Gratiot, name="Complete", 
		reference=as.Date("2001-01-01"), format="%d/%m/%Y")
# Generate initial points for the optimisation
parg<-par_init(data_Gratiot, parametersfixed=NULL)
# Run the optimisation
## Not run: 
# result_Gratiot<-fit_phenology(data=data_Gratiot, 
# 		parametersfit=parg, parametersfixed=NULL, trace=1)
# ## End(Not run)
data(result_Gratiot)
# Extract the fitted parameters
parg1<-extract_result(result_Gratiot)
# Add constant Alpha and Tau values 
# [day d amplitude=(Alpha+Nd*Beta)^Tau with Nd being the number of counts for day d]
pfixed<-c(parg1, Alpha=0, Tau=1)
pfixed<-pfixed[-which(names(pfixed)=="Theta")]
# The only fitted parameter will be Beta
parg2<-c(Beta=0.5, parg1["Theta"])
# Generate a likelihood map 
# [default Phi=seq(from=0.1, to=20, length.out=100) but it is very long]
# Take care, it takes 20 hours ! The data map_Gratiot has the result
## Not run: 
# map_Gratiot<-map_phenology(data=data_Gratiot, Phi=seq(from=0.1, to=20, length.out=100), 
# 		parametersfit=parg2, parametersfixed=pfixed)
# ## End(Not run)
data(map_Gratiot)
# Plot the map
plot(map_Gratiot, col=heat.colors(128))
# Plot the min(-Ln L) for Phi varying at any delta value
plot_phi(map=map_Gratiot)
# Plot the min(-Ln L) for Delta varying with Phi equal to the value for maximum likelihood
plot_delta(map=map_Gratiot)
# Plot the min(-Ln L) for Delta varying with Phi the nearest to 15
plot_delta(map=map_Gratiot, Phi=15)