Robyn Output Error - One or both dimensions exceed the maximum (50000px)

nikomataaa commented 7 months ago

Project Robyn

Describe issue

When I tried to run the model using conversion (CPA) metric, it was successfully executed, but when I run robyn_output it returned me this error message

>>> Calculating clusters for model selection using Pareto fronts...
>> Auto selected k = 6 (clusters) based on minimum WSS variance of 5%
Error : One or both dimensions exceed the maximum (50000px).
- Use `options(ragg.max_dim = ...)` to change the max
  Warning: May cause the R session to crash
In addition: There were 36 warnings (use warnings() to see them)
Error in clusterCollect$data : $ operator is invalid for atomic vectors

Provide reproducible example

Issues are often related to custom input data that is difficult to debug without. If necessary, please modify your data to mask real values and share a dataset that is able to reproduce the issue. Please also share your model configuration and exported JSON files if available.

the configuration that I used was the same like demo.R in the repo just different dataset

Environment & Robyn version

Make sure you're using the latest Robyn version before you post an issue.

Robyn version : 3.10.5.9007
R version 4.2.3 (2023-03-15)
MacOS Sonoma

gufengzhou commented 7 months ago

you opened and closed the same issue yesterday, right? Have you seen this who has the same problem?

nikomataaa commented 7 months ago

@gufengzhou yess, somehow it was duplicated (requested twice). Anw did you aware or ever find about this problem? somehow when I tried to change the dep_var_type from conversion to revenue it can hold more features, but when I tried to randomly elimiate the channel features, it only hold 8 to be able to run successfully

ghltk commented 6 months ago

@gufengzhou @nikomataaa After updating to latest dev version which is ‘3.10.5.9009’, I had the same issue. I'm running into the following error when I build initial model and refresh both. Did you find something to solve the problem?

If I set export = FASLE the error message is not showing. And then I can export onepager using robyn_onepagers(). But it's not very intuitive becasue I need to plot for all clusters.

jstrome-lmp commented 5 months ago

Hi there,

I am encountering this issue as well when modeling for conversions. Is there any update here?

Thank you!

CJ2407 commented 4 months ago

@gufengzhou Recently I have been encountering the same error that too when modeling for Revenue (which is what I have been modeling since last year. I did set export = FALSE and error message went away and I can get the desired solID's onepagers, but I use the alldecomp_matrix and pareto_aggregated CSVs as well, but none of those have been generated with export = FALSE.

Is there a more permanent solution to this?

R version is 4.2.2 and Robyn version 3.10.5 is installed on my end.

CJ2407 commented 4 months ago

@gufengzhou @laresbernardo Sorry for following up here on this issue. Is there a permanent solution?

CJ2407 commented 3 months ago

@gufengzhou @laresbernardo I am sorry for following up again, but this issue has reduced been bothering me since last 2 weeks and has immensely slowed me down. I have to run model twice to get to all one_pager PNGs and pareto CSVs. You see when I run the model with export = FALSE, the directory Robyn_date_init doesn't get created which then given me error when trying to run one_pager() as looks for that directory and it's not available. So, then I run model first with export = TRUE (this ends in error and hence doesn't generate Inputcollect and Outputcollect) and then run model with export = FALSE.

So, if you could please provide a permanent solution to this "dimensions exceed the maximum" error, that would be a huge help!

laresbernardo commented 3 months ago

Hi @CJ2407 @jstrome-lmp @ghltk

Not sure if you are aware but, even though I am actively collaborating and improving Robyn's code, I am not working at Meta anymore, so I don't have an active agenda to solve issues; all the changes I've submitted are for the sake of open-source, because I need it for my current projects, and for the love I still have for Robyn as an amazing MMM solution.

That said, I've found this same issue while running the code on my end and I've just created a branch that I think could fix it for you too. I would love for you to test it and see if it gets solved. Please, update Robyn by running Robyn::robyn_update(ref = "fix874"), refresh your R session, and retry to run your code. Do let me know how it goes and I'll merge it with main if it's fixed for you too.

CJ2407 commented 3 months ago

Thank you @laresbernardo for letting me know. I was honestly not aware that you are not working at Meta now. Apologies for reaching out so incessantly. Let me try this update and let you know.

CJ2407 commented 3 months ago

Thank you @laresbernardo. I tested the update you suggested and it worked. That error I have been getting is gone and the model run finished successfully and generated all the one_pagers and pareto CSVs. Thanks so much!

laresbernardo commented 3 months ago

Fantastic! Thanks for confirming. Will land in main branch once @gufengzhou approves the PR #939

facebookexperimental / Robyn