The Product Idea boards have gotten an update to better integrate them within our Product team's idea cycle! However this update does have a few unique behaviors, if you have any questions about them check out our FAQ.

Alteryx Designer Desktop Ideas

Share your Designer Desktop product ideas - we're listening!
Submitting an Idea?

Be sure to review our Idea Submission Guidelines for more information!

Submission Guidelines

Revisit R tools fake errors


I remember a while ago running into a peculiar error:
'The R.exe exit code (4294967295) indicted an error'. This was peculiar, as the data output was still seemingly correct, however, the error made me double-check the community for answers.


There are some very technical sources here:

but in short, this seems to be caused by a return code from C++ libraries, being understood by R as an error. Its a very inconsistent error, typically caused by low memory. This creates what most call a 'fake error' - the code runs perfectly fine, but seems to produce an error that doesn't actually indicate anything wrong.


Within those threads, its also stated that calling the garbage collection function (gc()) does tend to solve the problem on R exit, however this requires a user to understand basic R, and have access to the macro to be able to change the code - thus making predictive analytics more intimidating than it already is for new Alteryx users.


The first occurrence of this error seems to be way back in 2015, however the error is still being reported by users (see posts from 2020 and 2021): 

An important issue of these 'fake errors', is not only that they cause confusion, but also that they will cause analytic apps and server workflows to not work as expected, and stop running depending on the configuration.


My suggestion would be to revisit this issue, as by my understanding it occurs inconsistently, and calling garbage collection does not always seem to fix it. Even if the Error message is still created, it may be worth Alteryx suppressing these errors, in the case they are not real errors.



Steps to reproduce:

(as mentioned, its very inconsistent)

1. Open the Boosted Model example workflow

2. *10 the number of maximum trees in the model, in the boosted model configuration (Model customization)

3. Run the workflow, inspect the results (which are seemingly correct), and the error message in the results window.





Hope this helps!

17 - Castor
17 - Castor

@TheOC Thanks for submitting the idea! Thankfully I've only run into this a couple times, but it was really confusing when I came to the same conclusions as you.

16 - Nebula
16 - Nebula

Thank you for re-addressing this issue.  It's been a pain point for me for many years, and keeps me from "fully" automating our modeling processes.  

15 - Aurora
15 - Aurora

@mbarone @patrick_digan 
Agreed - I believe I received that error on one of my first times of using Alteryx and figured I would come back to the predictive tools when I understand them in more detail.


Turns out I probably received the correct results after all!



7 - Meteor

We've also been seeing over the past few months during our production cycle. We're seeing the same pattern as described above, e.g., sporadic, memory-related etc. We did add in the gc() call and it seemed to make a difference for a while, but eventually the problem reappeared. Aside from the error being annoying and creepy, it appears to be innocuous. The job completes and we see the results we expect. We might switch out our R macro with Python at some point to make the problem go away.