[Question]: Delayed data loading clarification

[mschaffer-incyte]

What is your question?

I've explored teal pretty deeply and was looking for clarification of the delayed data loading approach illustrated by the delayed gallery code: https://github.com/insightsengineering/teal.gallery/blob/main/delayed-data/app.R.

While this is useful, it still seems to require fully loading the (sometimes very large) analysis data sets into memory. I was wondering if there is an abstraction that allows a more efficient column-level data retrieval based only on what module analyses calls for (e.g. for populating the UI dropdowns, unique content of a column may be all that is needed).

Instead of loading all data into memory, is there a recommended pattern for data retrieval from a database or parquet files that dynamically reacts to a module's data requests and only fetches the relevant filtered content needed for the UI and current visual?

Code of Conduct

• I agree to follow this project's Code of Conduct.

Contribution Guidelines

• I agree to follow this project's Contribution Guidelines.

Security Policy

• I agree to follow this project's Security Policy.

[llrs-roche]

Hi, apologies for not answering sooner.

I think I replied you on the slack channel. In addition, I don't think there is anything specific on teal that makes teal applications eager for data. One could pass data that is loaded when a user clicks a button or certain conditions are met.

Current modules on teal.modules.general depend on what data they receive to react differently which requires collecting it.