[Freeze][Do Not Merge] - Created silver predictions table #665
Conversation
…leaned up logic to follow latest implementations and verified tests are working.
…ta as ms. Plots are now charting as expected.
…ms timestamp, not raw_subgraph/Predictions... this cold be improved further
…mmed addresses, ms, and other details that were lost in the test configuration.
…d gets composed correctly. Updated tests to handle current errors. I believe there are columns from subgraph fetch that we can drop
* the function sign has changed * fix table test --------- Co-authored-by: Norbert <[email protected]>
…date stake calculation logic
|
Just did some testing agains real data, unfortunately the table creation with all the calculations take too much time for big number of data. It needs some improvements.. |
| df_subscriptions: DataFrame, | ||
| user_subscriptions: DataFrame, | ||
| df_slots: DataFrame, | ||
| ) -> dict: |
There was a problem hiding this comment.
This looks great @KatunaNorbert!
It provides an outline for all the calculations we'll have to do for the silver table, for each individual parameter... awesome! Now that we know these, we'll be able to more easily write them in SQL.
- Don't worry about cleaning this up, splitting into functions, etc...
- Focus on improving test coverage such that you can validate that the values are calculated correctly
- Once you feel tests are validated well enough, let's move onto streamlit.
I have upated the Epic + streamlit task to reflect how you should move forward. Specifically, look at Calina's PRs and follow the deliverables on #612
|
Hey @KatunaNorbert, I reviewed the implementation, some of the test coverage, the PR in general, and it's shaping up really well! I have provided some feedback (minor) and next steps for how to move forward. [Final Requirement] Please get the |
idiom-bytes
changed the title
Fixes #746