Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

❓❔ FAQ & Troubleshooting ❔❓ #54

Open
hbaniecki opened this issue Feb 17, 2020 · 6 comments
Open

❓❔ FAQ & Troubleshooting ❔❓ #54

hbaniecki opened this issue Feb 17, 2020 · 6 comments

Comments

@hbaniecki
Copy link
Member

hbaniecki commented Feb 17, 2020

modelStudio FAQ & Troubleshooting

Most of the information is covered in the documentation: https://modelstudio.drwhy.ai/


Please, submit a new issue when dealing with potential bugs. Thanks!


  • Error occurred during the modelStudio() computation
  • foo plot doesn't show up on the dashboard
  1. Read the console output of DALEX::explain(). There could be a warning message pointing to the solution of this problem.
  2. Read the console output of modelStudio(). There could be an error message (printed as a warning) pointing to the origin and solution of this problem.
  3. Make sure to update these R packages to their latest versions: DALEX, ingredients, iBreakDown.
  • modelStudio() output shows up as a white window in the RStudio Viewer

Solve this by updating the RStudio. Please, check if the output shows up properly in the browser (e.g. use viewer = "browser" argument in modelStudio()).

  • y-axis labels go outside of the plot

Use modelStudio(..., options = ms_options(margin_left = 200)).

@hbaniecki hbaniecki pinned this issue Feb 17, 2020
@hbaniecki hbaniecki changed the title ❓❔ FAQ ❔❓ ❓❔ FAQ & Troubleshooting ❔❓ Feb 17, 2020
@ModelOriented ModelOriented locked and limited conversation to collaborators Feb 17, 2020
@ModelOriented ModelOriented unlocked this conversation Feb 17, 2020
@hbaniecki hbaniecki added short term ⏰ TODO short term and removed short term ⏰ TODO short term labels Apr 9, 2020
@CoolShades
Copy link

4X4 modelStudios the only way to get a studio?

just a half scroll down from https://modelstudio.drwhy.ai/, we are greeted by a very nice 1x2 panel modelStudio.

All variations of modelStudio I am able to create are only 4x4 panels. I can't see any option/argument to choose the sleeker version of modelStudio.

@hbaniecki
Copy link
Member Author

4X4 modelStudios the only way to get a studio?

just a half scroll down from https://modelstudio.drwhy.ai/, we are greeted by a very nice 1x2 panel modelStudio.

All variations of modelStudio I am able to create are only 4x4 panels. I can't see any option/argument to choose the sleeker version of modelStudio.

Hi, you can use the facet_dim parameter in modelStudio() https://modelstudio.drwhy.ai/articles/ms-perks-features.html#grid-size

@CoolShades
Copy link

Wow. 5 mins response time. Thanks so much!!

@BalintKomjati
Copy link

@hbaniecki MS is a very cool tool, love it! I have a complex model trained on a large dataset and MS html report generation fails I guess due to resource constrains... Is it possible to restrict MS to generate only preselected reports? E.g. exclude partial dependency completely. I have not found this option at all.

@hbaniecki
Copy link
Member Author

Hi @BalintKomjati, I am glad you find MS useful. Unfortunately, "MS html report generation fails" is not specific enough for me to be able to help troubleshoot the issue. Answering the question: partially, you can omit to generate EDA plots using the eda=FALSE parameter, but there is currently no option to exclude a selected plot. In general, consider restraining the size of data passed to explain() to 10K observations, as no more are effectively used to create visualizations. There are other options allowing you to decrease the required computation resources https://modelstudio.drwhy.ai/articles/ms-perks-features.html#more-calculations-means-more-time Feel free to open a new issue or write me an email describing the "generation fail".

@BalintKomjati
Copy link

@hbaniecki thanks for the quick reply! By no means I wanted to ask you to troubleshoot my issue :) I have not dig into it and was just looking for a confirmation that this option is not available (bc it would be a great tool to check which plot might cause the trouble...) Thanks again! :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants