R/uploaddata.R

In corujito/filtrafilet: Filtra o filet da base

rbind.match.columns <- function(input1, input2) {
  n.input1 <- ncol(input1)
  n.input2 <- ncol(input2)
  
  if (n.input2 < n.input1) {
    TF.names <- which(names(input2) %in% names(input1))
    column.names <- names(input2[, TF.names])
  } else {
    TF.names <- which(names(input1) %in% names(input2))
    column.names <- names(input1[, TF.names])
  }
  
  return(rbind(input1[, column.names], input2[, column.names]))
}


uploaddata <- function(csvfile, data, ...){
  #init
  #column_name_options <<- c("Publication.Year", "Authors", "Source.Title", "V5")
  #init end

  #if(substring(tolower(csvfile), nchar(csvfile)-3) != ".csv"){
   # stop('Uploaded data needs to be .csv file. When using MS Excel, try "Save As" and select type "CSV (comma delimited)"');
  #}
  
  #read csv data
  c = tryCatch({
    read.table(csvfile, row.names = NULL, stringsAsFactors=FALSE, header=TRUE, sep=",")
  }, error = function(e) {
    read.table(csvfile, row.names = NULL, skip=4, stringsAsFactors=FALSE, header=TRUE, sep=",")
  })
  
  if (ncol(c) < 18){
    colnames(c) <- c(
    "Authors", "Title", "Publication.Year", "Source.Title", "Volume", "Issue",
    "Article.Number", "Beginning.Page", "Ending.Page", "Page.count", "Total.Citations",
    "DOI", "Link", "Document.Type", "Source", "EID"
    )
  }
  
  c$Source.Title <-toupper(c$Source.Title) #Deixando o título dos Journals todo com letra maiuscula, pois será utilizado como chave
    
  ## Base com JCR ##
  #jcr2 <- read.csv('data/jcr2.csv', header = F, sep = ",")
  #jcr2 <- jcr2[-1,]
  jcr2[,2] <- toupper(jcr2[,2]) #Deixando o título dos Journals todo com letra maiuscula, pois será utilizado como chave
  jcr <- jcr2[!duplicated(jcr2$Full.Journal.Title), ] #retirando duplicidade do arquivo jcr
  
  mydata <- merge(x=c, y=jcr, by.x = "Source.Title", by.y = "Full.Journal.Title", all.x = TRUE)
  #mydata <- mydata[!duplicated(mydata$Title), ]
  
  #convert columns with 7 or less levels to factors
  for (i in seq_along(mydata)) {
    if (length(unique(mydata[[i]])) < 8) {
      mydata[[i]] <- as.factor(mydata[[i]]);
    }
  }
  
  list(message = paste(nrow(jcr), nrow(jcr2), nrow(mydata), nrow(c)))
  
  ## Tratamento de algumas variáveis da base ##
  
  ## Transformando campo jcr em numerico ##
  mydata$Journal.Impact.Factor <- as.numeric(as.character(mydata$Journal.Impact.Factor))
  
  ## Transformando campo total.citations em numerico ##
  mydata$Total.Citations <- as.numeric(as.character(mydata$Total.Citations))
  
  ## Transformando campo publication.year em numerico ##
  mydata$Publication.Year <- as.numeric(as.character(mydata$Publication.Year))
  
  ## Substituindo NA de jcr por -1 e total de citacoes por 0 ##
  for (i in which(is.na(mydata$Journal.Impact.Factor)==T)) { mydata$Journal.Impact.Factor[i] = -1 }  #melhorar codigo sem utilizar for
  for (i in which(is.na(mydata$Total.Citations)==T)) { mydata$Total.Citations[i] = 0 } #melhorar codigo sem utilizar for
  
  if (!is.null(data)) {
    mydata <- rbind.match.columns(mydata,data)
  }
  
  #return dataset
  return(mydata)  
}