R/door_melt.R

In ropensci/DoOR.functions: Integrating Heterogeneous Odorant Response Data into a Common Response Model: A DoOR to the Complete Olfactome

# door_melt
# 
# prepare odor response data for plotting via ggplot2
# 
# @param data data frame containing either subsets of door_response_matrix or
#   identifier and data columns (e.g. Or22a)
# @param ident chemical identifier, if empty rownames of data will be used
# @param datasets column names of data sets to use, important when using
#   response data (e.g. Or22a), if empty all columns will be used
# @param na.rm should NAs be removed
#   
# @return "long" data frame for plotting with ggplot2
# @author Daniel Münch <\email{daniel.muench@@uni-konstanz.de}>
# @examples
# library(DoOR.data)
# data(Or22a)
# data(door_response_matrix)
# head(door_melt(Or22a, datasets = c("Hallem.2004.WT", "Pelz.2006.AntEC50")))
# head(door_melt(door_response_matrix[1:100,], na.rm = TRUE))
# @importFrom stats na.omit
door_melt <- function(data, datasets, ident, na.rm = FALSE) {
  if (missing(datasets))
    datasets <- colnames(data)
  
  if (missing(ident)) {
    data$odorant <- rownames(data)
    ident <- "odorant"
  }
  
  result <- data.frame()
  for (i in seq_along(datasets)) {
    tmp <-
      data.frame(odorant = data[, ident],
                 dataset = datasets[i],
                 value = data[, datasets[i]])
    result <- rbind(result, tmp)
  }
  
  if (na.rm == TRUE)
    result <- na.omit(result)
  
  return(result)
}