[Feature Request]: Pre/post-processing mechanism with teal modules

[lcd2yyz]

Feature description

Consider two workflow cases below

1. Original dataset -> filter data via teal filter panel -> perform additional data processing, data manipulation, or calculation based on the filtered datasets -> summary/visualization with existing teal modules
2. Original dataset -> filter data via teal filter panel -> summary/visualization with existing teal modules -> perform additional processing or formatting based on the analysis result returned from teal modules -> display in teal app

It will be a very useful feature to create a mechanism to allow customizable pre-processing of filtered data before entering teal modules or post-processing of data/results coming out of the teal modules.

Examples of potential use cases for pre-processing can be

• Performing simulation or statistical bootstrapping on filtered datasets
• Creating new variables on-the-fly based on filtered datasets
• Flexible pooling and/or grouping based on user interaction
• Passing-in additional data processing code to manipulate filtered datasets
• Additional UI selectors different from the out-of-box teal modules

Examples of potential use cases for post-processing can be

• Re-formatting of analysis results from teal modules
• Performing additional derivations and calculation of analysis results
• Extracting selective statistics from teal modules

Tasks:

• [Research] Propose the design and PoC to fulfill the use case
• Implementation

[chlebowa]

It sounds like having a GUI to an interactive R session. This may be exceeding the scope of teal.

[gogonzo]

Consider two workflow cases below

I'm closing this issue in favour of two more specific.

1. Original dataset -> filter data via teal filter panel -> perform additional data processing, data manipulation, or calculation based on the filtered datasets -> summary/visualization with existing teal modules

For the first item we have teal_transform_module https://github.com/insightsengineering/nestdevs-tasks/issues/58 or the milestone https://github.com/insightsengineering/teal/milestone/7

2. Original dataset -> filter data via teal filter panel -> summary/visualization with existing teal modules -> perform additional processing or formatting based on the analysis result returned from teal modules -> display in teal app

Second one is called "output decorator" https://github.com/insightsengineering/teal.modules.clinical/issues/1187