info.nests {embryogrowth}R Documentation

Calculte statistics about nests

Description

This function calculates many statistics about nests.
The embryo.stages is a named vector with relative size as compared to final size at the beginning of the stage. Names are the stages.
For example for SCL in Caretta caretta:
embryo.stages=structure(c(8.4, 9.4, 13.6, 13.8, 18.9, 23.5, 32.2, 35.2, 35.5, 38.5)/39.33),
.Names = c("21", "22", "23", "24", "25", "26", "27", "28", "29", "30", "31"))
indicates that the stages 21 begins at the relative size of 8.4/39.33 as compared to the final size.
Series can be indicated as the name of the series, their numbers or series or succession of TRUE or FALSE. "all" indicates that all series must be analyzed.
The likelihood object is just the total likelihood of the data in the model.
If one parameter is named "pipping_emergence" it is used as the number of days between pipping and emergence to calculate the 1/3 and 2/3 of incubation.
The summary object is a data.frame composed of these elements with the suffix .mean, .se or .quantile_x with x from the parameter probs.

If out is equal to summary, the return is a list with:

If out is equal to metric, the return is a list with:

The object summary.dynamic.metric or indices.dynamic.metric is a data.frame with the following columns:

If you indicate new set of temperatures, you must probably also indicate new hatchling.metric values.
Note: four species have predefined embryo stages. embryo.stages parameter can take the values:

But remember that mass is not the best proxy to describe the growth of an embryo because it can decrease if the substrate becomes dry.
The progress bar is based on both replicates and timeseries progress. It necessitates the pbapply package.
If replicate.CI is null or 0, only maximum likelihood is used and no confidence interval is calculated.
If replicate.CI is 1, one random value for the parameters is used but no confidence interval is calculated.
In other cases, replicate.CI random samples are used to estimate confidence interval.

Usage

info.nests(
  x = NULL,
  parameters = NULL,
  NestsResult = NULL,
  resultmcmc = NULL,
  hessian = NULL,
  GTRN.CI = NULL,
  fixed.parameters = NULL,
  SE = NULL,
  temperatures = NULL,
  integral = NULL,
  derivate = NULL,
  hatchling.metric = NULL,
  stop.at.hatchling.metric = FALSE,
  M0 = NULL,
  series = "all",
  TSP.borders = NULL,
  embryo.stages = NULL,
  TSP.begin = 0,
  TSP.end = 0.5,
  replicate.CI = 0,
  weight = NULL,
  out = "likelihood",
  fill = NULL,
  probs = c(0.025, 0.5, 0.975),
  SexualisationTRN = NULL,
  SexualisationTRN.mcmc = NULL,
  SexualisationTRN.CI = NULL,
  metric.end.incubation = "observed",
  metabolic.heating = 0,
  temperature.heterogeneity = 0,
  progressbar = FALSE,
  warnings = TRUE,
  parallel = TRUE,
  tsd = NULL,
  tsd.CI = NULL,
  tsd.mcmc = NULL,
  zero = 1e-09,
  verbose = FALSE
)

Arguments

x

A set of parameters to model the embryo growth thermal reaction norm or a NestsResult object.

parameters

A set of parameters to model the embryo growth thermal reaction norm. It will replace the parameters included in NestsResult (same as x).

NestsResult

A NestsResult object generated by searchR to model the embryo growth thermal reaction

resultmcmc

A mcmc result for embryo growth thermal reaction norm

hessian

An hessian matrix for embryo growth thermal reaction norm. It will replace the hessian matrix included in NestResult object.

GTRN.CI

How to estimate CI for embryo growth thermal reaction norm; can be NULL, "SE", "MCMC", or "Hessian".

fixed.parameters

A set of fixed parameters to model the embryo growth thermal reaction norm. It will replace the fixed parameters included in NestsResult.

SE

Standard error for each parameter. It will replace the SE in NestsResult. Use SE=NA to remove SE from NestResult

temperatures

Timeseries of temperatures formatted using formatNests(). It will replace the one in NestsResult.

integral

Function used to fit embryo growth: integral.Gompertz, integral.exponential or integral.linear. It will replace the one in NestsResult.

derivate

Function used to fit embryo growth: dydt.Gompertz, dydt.exponential or dydt.linear. It will replace the one in NestsResult.

hatchling.metric

Mean and SD of size of hatchlings. It will replace the one in NestsResult.

stop.at.hatchling.metric

TRUE or FALSE. If TRUE, the model stops when proxy of size reached the mean hatchling.metric size.

M0

Measure of hatchling size proxi at laying date. It will replace the one in NestsResult.

series

The name or number of the series to be estimated.

TSP.borders

The limits of TSP in stages. See embryo.stages parameter.

embryo.stages

The embryo stages. At least TSP.borders stages must be provided to estimate TSP borders. See note.

TSP.begin

Where TSP begin during the stage of beginning? In relative proportion of the stage.

TSP.end

Where TSP begin during the stage of ending? In relative proportion of the stage.

replicate.CI

Number of replicates to estimate CI. See description

weight

Weights of the different nests to estimate likelihood. It will replace the ones in NestsResult.

out

Can take the values of "likelihood", "summary", "metric" or "dynamic".

fill

Number of minutes between two records. Create new one if they do not exist. NULL does not change the time of temperature recordings.

probs

Probabilities for metric quantiles.

SexualisationTRN

A set of parameters used to model sexualisation thermal reaction norm during TSP or a result of STRN()

SexualisationTRN.mcmc

A mcmc object obtained from STRN_MHmcmc() to generate variability for sexualisation thermal reaction norm during TSP

SexualisationTRN.CI

How to estimate CI of sexualisation thermal reaction norm. Can be NULL, "SE", "MCMC", or "Hessian".

metric.end.incubation

The metric at the end of incubation used to calibrate TSP size. Can be "hatchling.metric", or "observed".

metabolic.heating

Degrees Celsius to be added at the end of incubation due to metabolic heating.

temperature.heterogeneity

SD of heterogeneity of temperatures. Can be 2 values, sd_low and sd_high and then HelpersMG::r2norm() is used.

progressbar

If FALSE, the progress bar is not shown (useful for using with sweave or knitr)

warnings

If FALSE, does not show warnings

parallel

If TRUE use parallel version for nests estimation

tsd

A object from tsd() that describe the thermal react norm of sex ratio at constant temperatures

tsd.CI

How to estimate CI for sex ratio thermal reaction norm; Can be NULL, "SE", "MCMC", or "Hessian".

tsd.mcmc

A object from tsd_MHmcmc() .

zero

Value to replace 0 or 1.

verbose

If TRUE, show more information.

Details

Calculate statistics about nests

Value

Return or the total likelihood or a list with $metric and $summary depending on out parameter

Author(s)

Marc Girondot marc.girondot@gmail.com

Examples

## Not run: 
library(embryogrowth)
data(resultNest_4p_SSM)
# Some basic calculations to show the advantage of parallel computing
system.time(summary.nests <- info.nests(x=resultNest_4p_SSM, out="summary", 
  embryo.stages="Caretta caretta.SCL", replicate.CI=0, parallel=FALSE))
system.time(summary.nests <- info.nests(x=resultNest_4p_SSM, out="summary", 
  embryo.stages="Caretta caretta.SCL", replicate.CI=0, parallel=TRUE))
system.time(summary.nests <- info.nests(x=resultNest_4p_SSM, out="summary", 
  embryo.stages="Caretta caretta.SCL", replicate.CI=0, parallel=TRUE, progressbar=TRUE))
system.time(summary.nests <- info.nests(x=resultNest_4p_SSM, out="likelihood", 
  embryo.stages="Caretta caretta.SCL", replicate.CI=0, parallel=TRUE, progressbar=FALSE))
  
# By default parallel computing is TRUE but progressbar is FALSE
# When out is "likelihood", it returns only the likelihood
# otherwise, it returns a list with 3 objects "summary", 
#        "dynamic.metric", and "summary.dynamic.metric".

summary.nests <- info.nests(resultNest_4p_SSM, out="summary", 
  embryo.stages="Caretta caretta.SCL", 
  replicate.CI=100, 
  resultmcmc=resultNest_mcmc_4p_SSM, 
  GTRN.CI="MCMC", 
  progressbar=TRUE)
  
summary.nests <- info.nests(resultNest_4p_SSM, 
  embryo.stages="Caretta caretta.SCL", 
  out="summary", replicate.CI=100, 
  GTRN.CI="Hessian", 
  progressbar=TRUE)
  
summary.nests <- info.nests(resultNest_4p_SSM, 
  series = 1, 
  embryo.stages="Caretta caretta.SCL", 
  out="summary", replicate.CI=100, 
  GTRN.CI="SE", 
  progressbar=TRUE)
  
# Example of use of embryo.stages and TSP.borders:
  summary.nests <- info.nests(resultNest_4p_SSM, out="summary", 
                            embryo.stages=c("10"=0.33, "11"=0.33, "12"=0.66, "13"=0.66), 
                            TSP.borders = c(10, 12), 
                            replicate.CI=100,
                            progressbar=TRUE)
                            
#########################################
# Sex ratio using Massey et al. method PM
#########################################

# Massey, M.D., Holt, S.M., Brooks, R.J., Rollinson, N., 2019. Measurement 
# and modelling of primary sex ratios for species with temperature-dependent 
# sex determination. J Exp Biol 222, 1-9.
  
CC_Mediterranean <- subset(DatabaseTSD, RMU=="Mediterranean" & 
Species=="Caretta caretta" & (!is.na(Sexed) & Sexed!=0))
tsdL <- with (CC_Mediterranean, tsd(males=Males, females=Females, 
                                    temperatures=Incubation.temperature, 
                                    equation="logistic", replicate.CI=NULL))
                                    
PM <- info.nests(x=resultNest_4p_SSM, 
  GTRN.CI="Hessian", tsd.CI="Hessian", 
  embryo.stages="Caretta caretta.SCL", replicate.CI=100, 
  out="summary", progressbar=TRUE, tsd=tsdL)
  

plot_errbar(x=PM$summary$TimeWeighted.temperature.mean, 
            y=PM$summary$TSP.PM.GrowthWeighted.mean, 
            y.minus=PM$summary$TSP.PM.GrowthWeighted.quantile_0.025, 
            y.plus=PM$summary$TSP.PM.GrowthWeighted.quantile_0.975, 
            xlab="CTE SCL growth", 
            ylab="PM Massey et al. 2016", xlim=c(26, 32), ylim=c(0, 1), las=1)

# Relationship between growth and growth rate

infoall.df <- info.nests(x=resultNest_4p_SSM, out="summary", 
  embryo.stages="Caretta caretta.SCL", 
  replicate.CI=100, 
  resultmcmc=resultNest_mcmc_4p_SSM, 
  GTRN.CI="MCMC", 
  progressbar=TRUE)
  
  layout(1)
plot(x=infoall.df$dynamic.metric[[1]][, "Time"], 
     y=infoall.df$dynamic.metric[[1]][, "Metric_50%"], 
     type="l", las=1, bty="n", 
     xlab="Time in minute", ylab="Growth", ylim=c(0, 39), xlim=c(0, 100000))
lines(x=infoall.df$dynamic.metric[[1]][, "Time"], 
     y=infoall.df$dynamic.metric[[1]][, "Metric_2.5%"], lty=2)
lines(x=infoall.df$dynamic.metric[[1]][, "Time"], 
     y=infoall.df$dynamic.metric[[1]][, "Metric_97.5%"], lty=2)

## End(Not run)
[Package embryogrowth version 9.1 Index]