Excess Memory Utilization in BQ for Encounter Grouper

[chase-jones]

---

Describe the bug - Required

Utilization of window functions in encounters__combined_claim_line_crosswalk model blows up bigquery, makes it mad, and Google refuses to do the work.

---

Environment - Required

• Tuva project package version (e.g. 0.6.0): below
• dbt version (e.g. 1.7): blow
• dbt type (e.g. dbt cloud or dbt CLI): dbt core
• Data warehouse (e.g. Snowflake): BQ

packages:
  - package: dbt-labs/dbt_utils
    version: 1.2.0
  - package: calogica/dbt_expectations
    version: 0.10.3
  - package: dbt-labs/dbt_project_evaluator
    version: 0.14.3
  - package: dbt-labs/audit_helper
    version: 0.12.0
  - package: tuva-health/the_tuva_project
    version: 0.12.3
  - package: calogica/dbt_date
    version: 0.10.1
sha1_hash: e81d71351d6fb0adf5455874c8cd4c25b9d57f8c

requirements

dbt-bigquery==1.8.2
dbt-core==1.8.3
dbt-extractor==0.5.1
dbt-invoke==1.0.0
dbt-adapters==1.4.0

---

To Reproduce
Steps to reproduce the behavior:

1. Use a large medical claims data into Tuva in BQ (arbitrary)
2. Run encounters__combined_claim_line_crosswalk and upstream dependencies
3. Error due to memory

---

Expected behavior
No error

---

Screenshots

[chase-jones]

Not sure where this will land on Tuva priority list, but if it's low, it's something I might take. Just don't want to mess with if there's a lot of work arounds the encounter grouper

[chase-jones]

I didn't realize there there were a lot of dependencies on views. If upstream dependencies are materialized as tables, this will not happen. We can override in dbt_project.yml. Because of this, no longer recommending changes in the models as this can be solved with configuration. Closing in 2 days.

[aneiderhiser]

Sounds good @chase-jones thanks for digging in.

[aneiderhiser]

Re-opening as this is still an issue. Link to Slack convo describing details here: https://thetuvaproject.slack.com/archives/C03DET9ETK3/p1737862652160469

[aneiderhiser]

@colin-tuva can you take a look at this one?