R/intra_phyglm.R

In sensiPhy: Sensitivity Analysis for Comparative Methods

#' Intraspecific variability - Phylogenetic Logistic Regression
#'
#' Performs Phylogenetic logistic regression evaluating
#' intraspecific variability in predictor variables.
#'
#' @param formula The model formula: \code{response~predictor}. 
#' @param data Data frame containing species traits with species as row names.
#' @param phy A phylogeny (class 'phylo', see ?\code{ape}).
#' @param Vx Name of the column containing the standard deviation or the standard error of the predictor 
#' variable. When information is not available for one taxon, the value can be 0 or \code{NA}
#' @param x.transf Transformation for the predictor variable (e.g. \code{log} or \code{sqrt}). Please use this 
#' argument instead of transforming data in the formula directly (see also details below).
#' @param n.intra Number of times to repeat the analysis generating a random value for the predictor variable.
#' If NULL, \code{n.intra} = 2
#' @param distrib A character string indicating which distribution to use to generate a random value for the
#' predictor variable. Default is normal distribution: "normal" (function \code{\link{rnorm}}).
#' Uniform distribution: "uniform" (\code{\link{runif}})
#' Warning: we recommend to use normal distribution with Vx = standard deviation of the mean.
#' @param btol Bound on searching space. For details see \code{phyloglm}
#' @param track Print a report tracking function progress (default = TRUE)
#' @param ... Further arguments to be passed to \code{phyloglm}
#' @details
#' This function fits a phylogenetic logistic regression model using \code{\link[phylolm]{phyloglm}}.
#' The regression is repeated \code{n.intra} times. At each iteration the function generates a random value
#' for each row in the dataset using the standard deviation or error supplied and assuming a normal or uniform distribution.
#' To calculate means and se for your raw data, you can use the \code{summarySE} function from the 
#' package \code{Rmisc}.
#' 
#' All phylogenetic models from \code{phyloglm} can be used, i.e. \code{BM},
#' \code{OUfixedRoot}, \code{OUrandomRoot}, \code{lambda}, \code{kappa},
#' \code{delta}, \code{EB} and \code{trend}. See ?\code{phyloglm} for details.
#'
#' Currently, this function can only implement simple logistic models (i.e. \eqn{trait~
#' predictor}). In the future we will implement more complex models.
#'
#' Output can be visualised using \code{sensi_plot}.
#'
#' @section Warning:  
#' When Vx exceeds X negative (or null) values can be generated, this might cause problems
#' for data transformation (e.g. log-transformation). In these cases, the function will skip the simulation. This problem can
#' be solved by increasing \code{n.intra}, changing the transformation type and/or checking the target species in output$sp.pb.
#'  
#' @return The function \code{intra_phyglm} returns a list with the following
#' components:
#' @return \code{formula}: The formula
#' @return \code{data}: Original full dataset
#' @return \code{sensi.estimates}: Coefficients, aic and the optimised
#' value of the phylogenetic parameter (e.g. \code{lambda}) for each regression.
#' @return \code{N.obs}: Size of the dataset after matching it with tree tips and removing NA's.
#' @return \code{stats}: Main statistics for model parameters.\code{CI_low} and \code{CI_high} are the lower 
#' and upper limits of the 95% confidence interval.
#' @return \code{all.stats}: Complete statistics for model parameters. \code{sd_intra} is the standard deviation 
#' due to intraspecific variation. \code{CI_low} and \code{CI_high} are the lower and upper limits 
#' of the 95% confidence interval.
#' @return \code{sp.pb}: Species that caused problems with data transformation (see details above).
#' 
#' @author Caterina Penone & Pablo Ariel Martinez
#' @seealso \code{\link[phylolm]{phyloglm}}, \code{\link{sensi_plot}}
#' @references 
#' 
#' Paterno, G. B., Penone, C. Werner, G. D. A. 
#' \href{http://doi.wiley.com/10.1111/2041-210X.12990}{sensiPhy: 
#' An r-package for sensitivity analysis in phylogenetic 
#' comparative methods.} Methods in Ecology and Evolution 
#' 2018, 9(6):1461-1467
#'
#' Martinez, P. a., Zurano, J.P., Amado, T.F., Penone, C., Betancur-R, R., 
#' Bidau, C.J. & Jacobina, U.P. (2015). Chromosomal diversity in tropical reef 
#' fishes is related to body size and depth range. Molecular Phylogenetics and 
#' Evolution, 93, 1-4
#' 
#' Ho, L. S. T. and Ane, C. 2014. "A linear-time algorithm for 
#' Gaussian and non-Gaussian trait evolution models". Systematic Biology 63(3):397-408.
#' @examples
#' # Simulate Data:
#' set.seed(6987)
#' phy = rtree(150)
#' x = rTrait(n=1,phy=phy)
#' x_sd = rnorm(150,mean = 0.8,sd=0.2)
#' X = cbind(rep(1,150),x)
#' y = rbinTrait(n=1,phy=phy, beta=c(-1,0.5), alpha=.7 ,X=X)
#' dat = data.frame(y, x, x_sd)
#' 
#' # Run phylogenetic logistic regression accounting for intraspecific variation:
#' intra_glm <- intra_phyglm(y~x,Vx = "x_sd",data = dat,phy=phy,distrib = "normal")
#' 
#' #Print summary of sensitivity analysis
#' summary(intra_glm)
#' head(intra_glm$sensi.estimates)
#' #Visual output
#' sensi_plot(intra_glm)
#' @export


intra_phyglm <- function(formula,
                         data,
                         phy,
                         Vx = NULL,
                         n.intra = 30,
                         x.transf = NULL,
                         distrib = "normal",
                         btol = 50,
                         track = TRUE,
                         ...) {
  #Error check
  if (is.null(Vx))
    stop("Vx must be defined")
  if (!inherits(formula, "formula"))
    stop("formula must be class 'formula'")
  if (!inherits(data, "data.frame"))
    stop("data must be class 'data.frame'")
  if (!inherits(phy, "phylo"))
    stop("phy must be class 'phylo'")
  if (formula[[2]] != all.vars(formula)[1] ||
      formula[[3]] != all.vars(formula)[2])
    stop("Please use argument x.transf for data transformation")
  if (distrib == "normal")
    warning ("distrib=normal: make sure that standard deviation is provided for Vx")
  
  #Matching tree and phylogeny using utils.R
  datphy <- match_dataphy(formula, data, phy, ...)
  full.data <- datphy[[1]]
  phy <- datphy[[2]]
  
  resp1 <- all.vars(formula)[1]
  if (length(all.vars(formula)) > 2) {
    resp2 <- all.vars(formula)[2]
  }
  pred <- all.vars(formula)[length(all.vars(formula))]
  
  if (!is.null(Vx) && sum(is.na(full.data[, Vx])) != 0) {
    full.data[is.na(full.data[, Vx]), Vx] <- 0
  }
  
  #Function to pick a random value in the interval
  if (distrib == "normal")
    funr <- function(a, b) {
      stats::rnorm(1, a, b)
    }
  else
    funr <- function(a, b) {
      stats::runif(1, a - b, a + b)
    }
  
  
  #Create the results data.frame
  sensi.estimates <-
    data.frame(
      "n.intra" = numeric(),
      "intercept" = numeric(),
      "se.intercept" = numeric(),
      "pval.intercept" = numeric(),
      "estimate" = numeric(),
      "se.estimate" = numeric(),
      "pval.estimate" = numeric(),
      "aic" = numeric(),
      "optpar" = numeric()
    )
  
  
  #Model calculation
  counter = 1
  errors <- NULL
  species.NA <- list()
  if (track == TRUE)
    pb <- utils::txtProgressBar(min = 0, max = n.intra, style = 3)
  for (i in 1:n.intra) {
    ##Set predictor variable
    #Vx is not provided or is not numeric, do not pick random value
    if (!inherits(full.data[, pred], c("numeric", "integer")) ||
        is.null(Vx)) {
      full.data$predV <- full.data[, pred]
    }
    
    #choose a random value in [mean-se,mean+se] if Vx is provided
    if (!is.null(Vx) && is.null(dim(Vx)))
    {
      full.data$predV <-
        apply(full.data[, c(pred, Vx)], 1, function(x)
          funr(x[1], x[2]))
    }
    
    full.data$resp1 <-
      full.data[, resp1] #try to improve this in future
    if (length(all.vars(formula)) > 2) {
      full.data$resp2 <- full.data[, resp2]
    }
    
    #transform Vx if x.transf is provided
    if (!is.null(x.transf))
    {
      suppressWarnings (full.data$predV <- x.transf(full.data$predV))
    }
    
    #skip iteration if there are NA's in the dataset
    species.NA[[i]] <-
      rownames(full.data[with(full.data, is.na(predV)), ])
    if (sum(is.na(full.data[, "predV"]) > 0))
      next
    
    #model
    if (length(all.vars(formula)) > 2) {
      mod = try(phylolm::phyloglm(
        cbind(resp1, resp2) ~ predV,
        data = full.data,
        phy = phy,
        method = "logistic_MPLE",
        btol = btol
      ),
      FALSE)
    }
    else
      mod = try(phylolm::phyloglm(
        resp1  ~  predV,
        data  =  full.data,
        phy  =  phy,
        method  =  "logistic_MPLE",
        btol  =  btol
      )
      ,
      FALSE)
    
    if (isTRUE(class(mod) == "try-error")) {
      error <- i
      errors <- c(errors, error)
      next
    }
    
    
    else{
      intercept            <-
        phylolm::summary.phyloglm(mod)$coefficients[[1, 1]]
      se.intercept         <-
        phylolm::summary.phyloglm(mod)$coefficients[[1, 2]]
      estimate             <-
        phylolm::summary.phyloglm(mod)$coefficients[[2, 1]]
      se.estimate          <-
        phylolm::summary.phyloglm(mod)$coefficients[[2, 2]]
      pval.intercept       <-
        phylolm::summary.phyloglm(mod)$coefficients[[1, 4]]
      pval.estimate        <-
        phylolm::summary.phyloglm(mod)$coefficients[[2, 4]]
      aic.mod              <- mod$aic
      n                    <- mod$n
      #d                   <- mod$d
      optpar               <- mod$alpha
      
      
      if (track == TRUE)
        utils::setTxtProgressBar(pb, i)
      
      #write in a table
      estim.simu <-
        data.frame(
          i,
          intercept,
          se.intercept,
          pval.intercept,
          estimate,
          se.estimate,
          pval.estimate,
          aic.mod,
          optpar,
          stringsAsFactors = F
        )
      sensi.estimates[counter,]  <- estim.simu
      counter = counter + 1
      
    }
  }
  if (track == TRUE)
    on.exit(close(pb))
  
  #calculate mean and sd for each parameter
  #variation due to intraspecific variability
  statresults <- data.frame(
    min = apply(sensi.estimates, 2, min),
    max = apply(sensi.estimates, 2, max),
    mean = apply(sensi.estimates, 2, mean),
    sd_intra = apply(sensi.estimates, 2, stats::sd)
  )[-1,]
  
  statresults$CI_low  <-
    statresults$mean - stats::qt(0.975, df = n.intra - 1) * statresults$sd_intra / sqrt(n.intra)
  statresults$CI_high <-
    statresults$mean + stats::qt(0.975, df = n.intra - 1) * statresults$sd_intra / sqrt(n.intra)
  
  #species with transformation problems
  nr <- n.intra - nrow(sensi.estimates)
  sp.pb <- unique(unlist(species.NA))
  if (length(sp.pb) > 0)
    warning (
      paste(
        "in",
        nr,
        "simulations, data transformations generated NAs, please consider using another function
  for x.transf and check output$sp.pb",
        sep = " "
      )
    )
  
  res <- list(
    call = match.call(),
    formula = formula,
    data = full.data,
    sensi.estimates = sensi.estimates,
    N.obs = n,
    stats = round(statresults[c(1:6), c(3, 5, 6)], digits = 3),
    all.stats = statresults,
    sp.pb = sp.pb
  )
  class(res) <- c("sensiIntra", "sensiIntraL")
  return(res)
}