R/featureHistogram.R

In nkdhruw/XploreR:

featureHistogram <- function(data, feature, splitBy = NULL){
  errorMessage <- NULL 
  if(is.null(splitBy)){
    x <- data[, feature]
    p <- plot_ly(x = x, type = 'histogram', histnorm = 'probability') 
    return(list(errorMessage = errorMessage, p = p))
  } else {
    splitByClass <- class(data[,splitBy])
    splitByFactors <- unique(data[, splitBy])
    if(length(splitByFactors) > 50){
      errorMessage <- 'No. of factors in the splitBy column should be less than 50'
      return(list(errorMessage = errorMessage))
    }else{
      p <- plot_ly(type='histogram') %>%
        layout(xaxis = list(title = feature),
               yaxis = list (title = "Probability"))
      
      for(i in 1:length(splitByFactors)){
        x <- data[data[,splitBy]==splitByFactors[i], feature]
        print(length(x))
        p <- add_trace(p, x = x, name = splitByFactors[i], type = 'histogram', histnorm = 'probability')
      }
      return(list(errorMessage = errorMessage, p = p))
    }
  }
  
}