dasid with empty fields fails

[LennertSchepers]

Reported by Peter Herman:
flatdatasetrecord(1355) fails

I have discovered the problem in the imis function flatdatasetrecord(). This function tries to reorganise dascontacts into dascontacts2, but dascontacts has no rows for this dataset. That throws the error.
I guess the function will have to check whether dascontacts, daskeywords etc. have information, before the aggregate function is called and everything is put into the data frame. Seems like an easy thing to do.
This function works with the record no. 1355. It just makes an empty tibble if nothing can be found in the database

fdr2<-function(dasid){
  datasetrecords <- datasets(dasid)
  dascitations <- getdascitations(datasetrecords)
  if(nrow(dascitations)==0)dascitations<-tibble(dasid=as.character(dasid),title="",citation="")
  daskeywords <- getdaskeywords(datasetrecords)
  if(nrow(daskeywords)==0)daskeywords<-tibble(dasid=as.character(dasid),title="",keyword="")
  dascontacts <- getdascontacts(datasetrecords)
  if(nrow(dascontacts)==0)dascontacts<-tibble(dasid=as.character(dasid),title="",contact="")
  dastheme <- getdasthemes(datasetrecords)
  if(nrow(dastheme)==0)dastheme<-tibble(dasid=as.character(dasid),title="",theme="")
  dastheme2 <- aggregate(theme ~ dasid, data = dastheme, paste, 
                         collapse = " , ")
  daskeywords2 <- aggregate(keyword ~ dasid, data = daskeywords, 
                            paste, collapse = " , ")
  dascontacts2 <- aggregate(contact ~ dasid, data = dascontacts, 
                            paste, collapse = " , ")
  output <- dascitations %>% left_join(dascontacts2, by = "dasid") %>% 
    left_join(dastheme2, by = "dasid") %>% left_join(daskeywords2, 
                                                     by = "dasid")
  return(output)
}

dasid<-1355
fdr2(1355)

[LennertSchepers]

Peter H: I still had problems with the routine. I had to add more adjustments. The code has become really ugly now, but I don’t care because you are going to find a better structural solutions anyway. I got errors when the tibbles contained one line, but had NA for citation, or keyword, etc. Therefore I also corrected for this condition. You see that it results in quite some redundant code!

fdr2<-function(dasid){
  datasetrecords <- datasets(dasid)
  dascitations <- getdascitations(datasetrecords)
  if(nrow(dascitations)==0)dascitations<-tibble(dasid=as.character(dasid),title="",citation="")
  if(nrow(dascitations)==1) if(is.na(dascitations$citation)) dascitations$citation<-""
  daskeywords <- getdaskeywords(datasetrecords)
  if(nrow(daskeywords)==0)daskeywords<-tibble(dasid=as.character(dasid),title="",keyword="")
  if(nrow(daskeywords)==1) if(is.na(daskeywords$keyword))daskeywords$keyword<-""
  dascontacts <- getdascontacts(datasetrecords)
  if(nrow(dascontacts)==0)dascontacts<-tibble(dasid=as.character(dasid),title="",contact="")
  if(nrow(dascontacts)==1) if(is.na(dascontacts$contact))dascontacts$contact<-""
  dastheme <- getdasthemes(datasetrecords)
  if(nrow(dastheme)==0)dastheme<-tibble(dasid=as.character(dasid),title="",theme="")
  if(nrow(dastheme)==1) if(is.na(dastheme$theme))dastheme$theme<-""
  dastheme2 <- aggregate(theme ~ dasid, data = dastheme, paste, 
                         collapse = " , ")
  daskeywords2 <- aggregate(keyword ~ dasid, data = daskeywords, 
                            paste, collapse = " , ")
  dascontacts2 <- aggregate(contact ~ dasid, data = dascontacts, 
                            paste, collapse = " , ")
  output <- dascitations %>% left_join(dascontacts2, by = "dasid") %>% 
    left_join(dastheme2, by = "dasid") %>% left_join(daskeywords2, 
                                                     by = "dasid")
  return(output)
}