phenology (version 10.1)

par_init: Calculate initial set of parameters.

Description

This function is used to generate an initial set of parameters for fitting that is expected to be not to far from the final.
The parameters can be:

• Min, MinE, MinB, PMin, PMinB, PMinE;

• Max;

• Begin, Peak, Flat, End;

• Length, LengthB, LengthE;

• Theta;

• Alpha, Beta, Tau, Phi, Delta;

• Alpha1, Beta1, Tau1, Phi1, Delta1;

• Alpha2, Beta2, Tau2, Phi2, Delta2;

• Alpha3, Beta3, Tau3, Phi3, Delta3;

And the name of level if a cofactor is used.
The parameters Max, Min, MinE, MinB, Length, LengthB, LengthE, and Peak can be followed with _ and part of the name of the rookery.
The model for scale effect of sinusoid is: Alpha + Beta * n(t) ^ Tau where n(t) is the expected number for the day t without the sinusoid effect.

Usage

par_init(
  data = stop("A dataset must be provided"),
  fixed.parameters = NULL,
  add.cofactors = NULL
)

Value

The initial set of parameters

Arguments

data

Dataset generated with add_phenology()

fixed.parameters

Set of fixed parameters

add.cofactors

Names of cofactors that will be used (see fit_phenology)

Author

Marc Girondot

Details

par_init calculates initial set of parameters.

See Also

Other Phenology model: AutoFitPhenology(), BE_to_LBLE(), Gratiot, LBLE_to_BE(), LBLE_to_L(), L_to_LBLE(), MarineTurtles_2002, MinBMinE_to_Min(), adapt_parameters(), add_SE(), add_phenology(), extract_result(), fit_phenology(), likelihood_phenology(), logLik.phenology(), map_Gratiot, map_phenology(), phenology(), phenology2fitRMU(), phenology_MHmcmc(), phenology_MHmcmc_p(), plot.phenology(), plot.phenologymap(), plot_delta(), plot_phi(), print.phenology(), print.phenologymap(), print.phenologyout(), remove_site(), result_Gratiot, result_Gratiot1, result_Gratiot2, result_Gratiot_Flat, result_Gratiot_mcmc, summary.phenology(), summary.phenologymap(), summary.phenologyout()

Examples

if (FALSE) {
library(phenology)
# Read a file with data
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, fixed.parameters=NULL)
# Run the optimisation
result_Gratiot <- fit_phenology(data=data_Gratiot, 
		fitted.parameters=parg, fixed.parameters=NULL)
data(result_Gratiot)
# Plot the phenology and get some stats
output<-plot(result_Gratiot)

## When a series has only 0, it should be used in two steps
## Let see an example

# Let create a times series with only 0
data0 <- data.frame(Date=c("11/3/2015", "12/3/2015", "13/3/2015-18/3/2015", "25/3/2015"), 
                    Number=c(0, 0, 0, 0), 
                    Beach=rep("Site0", 4), stringsAsFactors=FALSE)
data1 <- data.frame(Date=c("15/3/2015", "16/3/2015", "20/3/2015-22/3/2015", "25/3/2015"), 
                    Number=c(1, 0, 3, 0), 
                    Beach=rep("Site1", 4), stringsAsFactors=FALSE)
data <- rbind(data0, data1)

# Here I include timeseries with no observation
try1 <- add_phenology(data, format="%d/%m/%Y", month_ref=1, include0=TRUE)
pfixed <- c(Min=0, Flat=0)
parg <- par_init(try1, fixed.parameters=pfixed)
# The Max value for the series without observations should not be fitted. The ML is for Max being 0
pfixed <- c(pfixed, parg[(substr(names(parg), 1, 4)=="Max_") & (parg == 0)])
parg <- parg[!(names(parg) %in% names(pfixed))]

}