R/covDataRead.R

In TreeBUGS: Hierarchical Multinomial Processing Tree Modeling

# read, check and mean-center covData
covDataRead <- function(covData, N, binaryToNumeric = FALSE) {
  if (!is.null(covData)) {
    # list / path to file
    if (is.character(covData)) {
      covData <- read.csv(covData, header = T, sep = ",", strip.white = T)
    }
    try(covData <- as.data.frame(covData))


    if (nrow(covData) != N) {
      stop("Number of individuals in 'data' and 'covData' differs!")
    }

    if (is.null(colnames(covData))) {
      stop("Check names of covariates in covData!")
    }

    if (binaryToNumeric) {
      for (k in 1:ncol(covData)) {
        if (inherits(covData[, k], c("factor", "ordered", "character"))) {
          if (length(unique(covData[, k])) == 2) {
            covData[, k] <- as.numeric(as.factor(covData[, k]))
          }
        }
      }
    }
  }
  covData
}



# get default values for predType
predTypeDefault <- function(covData, predType = NULL) {
  if (!is.null(covData)) {
    # default: continuous / random covariates
    if (missing(predType) || is.null(predType)) {
      cov.class <- sapply(covData, class)
      predType <- ifelse(cov.class %in% c("character", "factor"), "f",
        ifelse(cov.class %in% c("integer", "numeric", "matrix"), "c", "")
      )
    }
    for (cc in 1:length(predType)) {
      if (!all(predType %in% c("f", "c", "r"))) {
        stop(
          "Check definition of predType: should be a vector of the same length\n  ",
          "as there are columns in covData. Possible values are:\n",
          "  'c' (continuous variable),",
          "\n  'f' (fixed effects factor; only traitMPT), or",
          "\n  'r' (random effects factor; only traitMPT)."
        )
      }
    }
  } else {
    predType <- NULL
  }
  predType
}


# mean-centered variables as default (does not matter much for correlational analyses)
covDataCenter <- function(covData, predType) {
  if (!is.null(covData)) {
    for (i in 1:ncol(covData)) {
      if (predType[i] == "c") {
        scaled <- scale(covData[, i], center = TRUE, scale = FALSE) # centering of continuous variables
        if (any(scaled != covData[, i])) {
          covData[, i] <- c(scaled)
        }
      }
    }
  }

  covData
}