Query Editor R Script Text Type Issues

[alexandrevv]

Hello!

Query Editor R Script is changing the data type of text columns to integer automatically.

By business rule, the data must remain with the same type that was specified before running the R script.

Fields as text type:

 

R script passing the data without any transformation:

Data output with data type changed:

Code generated automatically by R saved in folder 'C:/Users/x/AppData/Local/Radio/RScriptWrapper_83e78459-95e6-46bb-b903-5cf7ae45dd6a':

# Prolog - Auto Generated #
# To run this script, open it with an R editor with UTF-8 encoding and uncomment the following line(s):
# Sys.setenv(RPackagesLibrariesDirectory = 'C:/Users/x/AppData/Local/Radio/Library')
# Sys.setenv(RScriptWrapperWorkingDirectory = 'C:/Users/x/AppData/Local/Radio/RScriptWrapper_83e78459-95e6-46bb-b903-5cf7ae45dd6a')
.libPaths( c( Sys.getenv('RPackagesLibrariesDirectory'), .libPaths() ))
# Input load. Please do not change #
setwd(Sys.getenv('RScriptWrapperWorkingDirectory'))
`dataset` = read.csv('input_df_6ebc0c50-ee6b-475f-952f-c15af44c5915.csv', check.names = FALSE, encoding = "UTF-8", blank.lines.skip = FALSE);
# Original Script. Please update your script content here and once completed copy below section back to the original editing window #

output <- dataset

# Epilog - Auto Generated  #
rm(dataset)
setwd(Sys.getenv('RScriptWrapperWorkingDirectory'))
rScriptWrapper_dataFrames <- list()
if (length(sapply(mget(ls(), .GlobalEnv), is.data.frame)) > 0 )
{
rScriptWrapper_dataFrames <- ls()[sapply(mget(ls(), .GlobalEnv), is.data.frame)]
for(x in rScriptWrapper_dataFrames) {assign(x, as.data.frame(get(x)))}
}
rScriptWrapper_options = getOption('save.image.defaults')
save(list = rScriptWrapper_dataFrames, file = "globalEnv.rdata", version = rScriptWrapper_options$version, ascii = FALSE, compress = TRUE)

Power BI's prolog code is using the function read.table to load the flat file without any parameter about data types of the source. The problem is being circumvented by reloading the file by the read_delim function with column definitions parameters for char type.

[v-haibl-msft]

@alexandrevv

I can repro the same issue as you. I’ve reported it internally to Power BI Team: CRI 37520685
I’ll post here once I get any update about it.

Best Regards,
Herbert

[v-haibl-msft]

@alexandrevv

Currently a workaround for this is to convert back integers to string in R script by using toString method.

Best Regards,
Herbert

[alexandrevv]

I think that in this case I'll lose padding zeros. Am I correct?