Options
"RapidMiner 511 with R Extension 5.2 and Read Database"
tridicular
Member Posts: 4 
Contributor I
Hello, I am trying something very simple with the R Extension. I have been using the R Extension, and it has been working out well, but I have just run into a problem.
I have a simple process with just two operators, a Read Database operator which leads to an Execute Script (R) operator. Both operators work fine independently, but when I connect up the two, the JVM crashes and RapidMiner shuts down.
The Read Database operator simply selects about a thousands rows and passes that Data Table to the next operator.
The Execute Script (R) operator takes in the Data Table from the previous process as a parameter and does the following:
result_data <- input_data
And that's it. Once I run the process RapidMiner crashes when it hits the Execute Script (R) operator.
Anyone else having similar issues?
Thank you!!
Tony
I have a simple process with just two operators, a Read Database operator which leads to an Execute Script (R) operator. Both operators work fine independently, but when I connect up the two, the JVM crashes and RapidMiner shuts down.
The Read Database operator simply selects about a thousands rows and passes that Data Table to the next operator.
The Execute Script (R) operator takes in the Data Table from the previous process as a parameter and does the following:
result_data <- input_data
And that's it. Once I run the process RapidMiner crashes when it hits the Execute Script (R) operator.
Anyone else having similar issues?
Thank you!!
Tony
Tagged:
0
Answers
did you try with several data? Might be there's something in your data you are retrieving from the database that makes this problems. Would it be a possibility to replace the Write database with a generate data? if this works, please write the generated data to the database and read and forward it again.
Greetings,
Sebastian
Thanks for the tip! You are right, it's data set specific. The process works fine with the Iris Repository, yet crashes with my Read Database operator. It also failed when I tried Read Excel. But it did work when I tried other Read Database operators. So I guess this means that the problem is in fact data specific. I still haven't narrowed down the actual cause, but your insight has helped me immensely.
Thank you very much!!
Tony
would be cool if you could trace it down, so that we can avoid this crashing in further versions
Greetings,
Sebastian
You are exactly right. The crash appears to be caused by a date column in the Read Database operator.
If I place a Date to Nominal operator in between the Read Database and the Execute Script, RapidMiner no longer crashes, and everything seems to work properly.
Regards,
Trishul
could you file a bug for this? We will try to address this in the next release of the R Extension. Although I'm not sure whether R supports any date variable?
Greetings,
Sebastian
Not a problem. I have filed a bug here:
http://bugs.rapid-i.com/show_bug.cgi?id=450
Yes, I don't believe R supports a date variable. However, it would suffice for RapidMiner to catch the exception without crashing the JVM and shutting down.
Cheers,
if the JVM crashes I doubt I get any chance to catch any Exception. In fact all exceptions are caught, but this lies much deeper in the native code...
Greetings,
Sebastian