RGet() error: Data type is not supported

hardner

I'm getting this Data Type is Not Supported error trying to pull an R dataframe into JMP using RGet().  As in  "dt=R Get(df);"

 

The data types of the columns in the dataframe are all either character or numeric as reported by R.  I'm going to start trying to sift though the haystack to see what specific entries are apparently causing this problem on the assumption it's something in one particular data entry.. 

 

But wondering whether there are some usual suspects or whether I can get advice on R code for cleansing the dataframe to make JMP consistently happy (I'm much more experienced in JSL than R). Maybe this is a routine happening and there is some sharable code to fix it up?

 

The context is using ROracle to pull data into the dataframe.  I can see this work just in R it creates an apparently unproblematic data frame and the same SQL makes an unobjectionable data table in JMP when used with OpenDatabase().  My JSL/R combo does work for some other example SQL as well, just certain ones give the error.   So apparently if there is some fishy character in the original Oracle table that JMP doesn't like but R is OK with then OpenDatabase() is coping with it in a way RGet() does not?

 

Appreciate any tips for troubleshooting this problem...

 

uday_guntupalli

@hardner, 
        If there is a problem with the data that you are pulling and you are more experienced in JSL than R, I would think it would be easier to identify the source of the error using JMP than bring R into the mix. I have used R in JMP 13 when JMP did not have a well defined API calling and handling ability which was fixed in JMP 14. 

        Now, let us look at the data that you are pulling which seems to be causing issues. When you say, it only happens with some queries, do you know if it happens:  

1. When you try to query from a specific table ? 
2. Can you try isolating the issue to a column or a combination of columns in your data table? 

        If you can isolate the issue in your source data, then it would be possible to build conditions to filter those rows out in SQL and continue to be able to query the data using JMP. While I am not opposed to the idea of using the JMP - R interface, given the problem, I am inclined to believe that this approach might give you speedy results.