Error: object 'knime.in' not found

[harishpk]

I am being driven crazy by this error. The data I am trying to manipulate in the R script is taking a while to load (circa 20-25 seconds) and hence the execution of the code is failing.

I built a 25 seconds delay in the code anticipating the data to load into R snippet from the previous node.

The error on the Console is ERROR R Snippet            0:12       Execute failed: R evaluation failed.

When I open the node and evaluate the script then error is Error: object 'knime.in' not found. If i wait for 15 seconds, knime.in gets loaded and then the script works fine

Can someone please help?

• capture.png

[jonathan.hale]

Hi harishpk,

sorry for the late reply!

When I open the node and evaluate the script

You will have to wait for the data to be loaded (knime.in show up in the workspace list on the right hand side), if you have large amounts of data this will take a while. (How many rows/columns is your input data?)

But the node should definitely not fail. I am currently working on improvements which should allow the R Snippet node to handle larger amounts of data. In Preferences > KNIME > R, you will find an option to increase the buffer size for Rserve, which may already help to prevent the

 Execute failed: R evaluation failed.

error.

If you believe this doesn't have anything to do with the data size, it would be amazing if you could send me the workflow via email: jonathan.hale@knime.com. Please also include which operating system you are running.

Please tell me if anything helped!

Regards, Jonathan.

[jpfeuffer]

Lately we are experiencing this kind of problems more and more, too.

Often associated with error 127. Sometimes data shows up in the scripting dialog only after the node was executed.

This happens with only a few rows with the random Cluster Data Generator.

[son]

I am experiencing the same kind of problem.

My workflow containing table to R nodes used to work until couple of days ago, but now it fails with the error in the KNIME console :"Execute failed: ("NullPointerException"): null".

If I open the node and evaluate the script I get "Error: object 'knime.in' not found". Moreover, I noticed that the table (1368 rows, 127 columns) doesn't appear in the workspace, no matter how long i wait for  it.

 

 

[joshuahoran]

I also get the "Error: object 'knime.in' not found" on a routine basis with the most recent version on KNIME. Even when you allow sufficient time for the data to load and wait for the knime.in object to become listed in the Workspace list -- I still get the error. Usually hitting "Reset Workspace" will fix it. If not, then closing the config window and re-opening will fix it. I've noticed that the frequency of this error is correlated with another error:

"The dialog pane for node 'R View (Table) 3:796' has thrown a 'IndexOutOfBoundsException'. That is most likely an implementation error."

I get this error about 50% of the time when I try to open the config window for an R node.

[Judit]

Replacing 'knime.in' with 'kIn' and 'knime.out' with 'rOut' solved the issue for me.

[Ryon]

Hello, 

I am having the same error: Object 'knime.in' not found when using R snippet nodes.  I tried using the aforementioned workarounds but it is still not working. I am running KNIME version 3.5.1 for Mac. The data is not particularly big: less than 20 columns and less than 500 rows.

Are there additional workarounds or things I can do?  

Much thanks in advance! 

Ryon

 

 

[pogacsa]

Hi All, I'm having the same issue too. 

I have KNIME 3.5.1, have already set the R buffer size to unlimited.

At first, if I evaluate the script (which works perfectly in RStudio), all goes well, up until the point, where I try using the nmf from the NMF package. Then I get the error message from above, which is:

ERROR: R evaluation failed.: "knime.tmp.ret<-NULL;printError<-function(e) message(paste('Error:',conditionMessage(e)));for(exp in tryCatch(parse(text=knime.tmp.script),error=printError)){tryCatch(knime.tmp.ret<-withVisible(eval(exp)),error=printError)
if(!is.null(knime.tmp.ret)) {if(knime.tmp.ret$visible) print(knime.tmp.ret$value)}};rm(knime.tmp.script,exp,printError);knime.tmp.ret$value"ERROR: R evaluation failed.: "knime.tmp.ret<-NULL;printError<-function(e) message(paste('Error:',conditionMessage(e)));for(exp in tryCatch(parse(text=knime.tmp.script),error=printError)){tryCatch(knime.tmp.ret<-withVisible(eval(exp)),error=printError)
if(!is.null(knime.tmp.ret)) {if(knime.tmp.ret$visible) print(knime.tmp.ret$value)}};rm(knime.tmp.script,exp,printError);knime.tmp.ret$value"ERROR: eval failedERROR: Could not capture output of command.ERROR: R evaluation failed.: "sink();sink(type='message')
close(knime.stdout.con);close(knime.stderr.con)
knime.output.ret<-c(paste(knime.stdout,collapse='\n'), paste(knime.stderr,collapse='\n'))
knime.output.ret"ERROR: R evaluation failed.: "sink();sink(type='message')
close(knime.stdout.con);close(knime.stderr.con)
knime.output.ret<-c(paste(knime.stdout,collapse='\n'), paste(knime.stderr,collapse='\n'))
knime.output.ret"ERROR: eval failed, request status: error code: 127ERROR: Could not execute internal command.ERROR: R evaluation failed.: "dev.off()"ERROR: R evaluation failed.: "dev.off()"ERROR: eval failed, request status: error code: 127

After which all objects are forgotton, knime.in not found, functions not found, everything not found, even though the objects are still listed on the right.

Any ideas? (I'm happy to send the wf, if you'd like to take a look.)

[jonathan.hale]

Hi all!

The data is not particularly big: less than 20 columns and less than 500 rows.

The size of the data should not be an issue with newer KNIME AP versions (e.g. 3.4 or 3.5).

 

Any ideas? (I'm happy to send the wf, if you'd like to take a look.)

Please make sure that the nmf package is properly installed. Also: Which OS are you using? In case you're on Windows, please also check that KNIME is using your own (not the bundled) R installation.
If that doesn't help, it would be indeed great if you could send me the workflow.

Cheers, Jonathan.

 

[pauloscheidegger]

It seems an old issue is live again. A script that was working fine yesterday started to show ‘Error: object ‘knime.in’ not found’ at the very start. I tried all tricks in the thread to no avail. It is a very small table input, less than 100 rows by 4 columns.
I can send WF and Excel input.

[ScottF]

Hi @pauloscheidegger -

Sorry for the trouble. Feel free to post an example workflow and dataset here - but if you would prefer it to be confidential, you can email those directly to scott.fincher@knime.com and I will have a look.

[pauloscheidegger]

Thank you for your prompt answer. Please find attached both the exported workflow and Excel file.Tabela Indicadores.xlsx (19.7 KB)
I’m running KNIME 4.1.2 in a Win7 machine.

Thanks in advance for the help.

[pauloscheidegger]

It seems something got wrong with the .knwf file, I’l try again.X-Am dados normais.knwf (18.4 KB)

It shows a different error now. But I’m unable to understand it.
You may (probably) find some errors in R script, I was tweaking it when the error appeared and it is not yet ready.

[ScottF]

Thanks for uploading the workflow and data.

The reason the R View node was failing was because the input to the node contained an unknown variable type (which shows up as a question mark in the table header). This was happening because the Cell Replacer was trying to replace a string from your Table Creator over an integer from your Excel Reader.

I explicitly changed the column2 field in the Table Creator from string to integer, and this solved the downstream knime.in problem in the R View node. Hope that helps!

[pauloscheidegger]

Wow! Thank you so much! I’m 68 and certainly not so clever a programmer as I was in my twenties. I am barely learning how to use Knime and R, it is so different from Algol… But back then this kind of error (type mismatch) was promptly detected and reported by the compiler at compile time. Maybe I became too lazy to drive myself around.
Thank you again!

[ScottF]

I don’t think it’s due to laziness on your part at all! If anything, the R View node is not generating a useful error message to help diagnose the problem. I think this behavior is probably a bug, so I’ll run it by one of our developers for clarification.

Thanks for bringing it to our attention.

(EDIT: ticket added - internal AP-14015)