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

Frontend hardcoded and send codes to Visualization server, it should be refactoryed #3085

Closed
rmgogogo opened this issue Feb 14, 2020 · 10 comments
Assignees
Labels
area/backend area/frontend lifecycle/stale The issue / pull request is stale, any activities remove this label. priority/p1 status/triaged Whether the issue has been explicitly triaged

Comments

@rmgogogo
Copy link
Contributor

cb4f4ba#diff-3403c794353f09e8d1cde18b6bfc3baa

Python codes should be keep inside Visualization server. Frontend should not hardcode python codes. Aka, if CUSTOM visualization means put python codes in frontend, better we disable it and use TYPED visualization.

@rmgogogo
Copy link
Contributor Author

@jingzhang36
@Bobgy

@Bobgy
Copy link
Contributor

Bobgy commented Feb 20, 2020

It's a good point, we should do this. just that I don't feel this is a p0.

This is an example of how to do this: https://github.com/kubeflow/pipelines/pull/3089/files

@eterna2
Copy link
Contributor

eterna2 commented Feb 21, 2020

Hey guys, what is ur opinion of using Vega as the main vis dsl? I am a big fan of Vega as I think it is a very good dsl for describing vis.

@eterna2
Copy link
Contributor

eterna2 commented Feb 21, 2020

And I can implement a purely client-side solution. And all it needs is a json spec of the vis.

@Bobgy
Copy link
Contributor

Bobgy commented Feb 22, 2020

@eterna2 Thanks for offering the idea, current KFP visualization work focuses on integrating existing jupyter notebook visualizations into KFP UI.

For Vega, sounds like an interesting usage, but I don't have much personal experience, can you open a separate issue to discuss about it? Do we expect data scientists using Vega for visualizations directly?

@Ark-kun
Copy link
Contributor

Ark-kun commented Feb 26, 2020

If possible, can we refactor visualization to move the visualization codes to /components/visualizations/<data type>/?
We could have them as proper components that receive the data or URI and output HTML or Markdown.

@Bobgy
Copy link
Contributor

Bobgy commented Feb 26, 2020

@Ark-kun thanks! that's a good idea to me. Current visualization doesn't take any parameters, sounds like it would be better as components instead. The caching problem is also solved.

@eterna2
Copy link
Contributor

eterna2 commented Feb 28, 2020

hey I created a issue to describe more on my idea #3187

I don't think vega will solve the issue with existing integration with jupyter notebook vis (unless they are using py lib like https://altair-viz.github.io/ ).

But I think it is very useful moving forward, for more general use cases (i.e. data that do not have a handy lib to generate the vis) and for ml-pipelines-ui-metadata.json

  • it is language agnostic and fully client-side

  • the dsl makes it easy to provide flexibility to generate any vis on demand, or wrap around a few signals for users to customize the vis

@Bobgy Bobgy added the status/triaged Whether the issue has been explicitly triaged label Mar 4, 2020
@stale
Copy link

stale bot commented Jun 24, 2020

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

@stale stale bot added the lifecycle/stale The issue / pull request is stale, any activities remove this label. label Jun 24, 2020
@stale
Copy link

stale bot commented Jul 1, 2020

This issue has been automatically closed because it has not had recent activity. Please comment "/reopen" to reopen it.

@stale stale bot closed this as completed Jul 1, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
area/backend area/frontend lifecycle/stale The issue / pull request is stale, any activities remove this label. priority/p1 status/triaged Whether the issue has been explicitly triaged
Projects
None yet
Development

No branches or pull requests

4 participants