Organizations have their Shiny Server installations with custom Shiny applications available. The standardized analytics are separately available in ATLAS after a particular Analysis is executed and its results are available as visuals (tables, graphs)
As soon as Analysis Results data is available there is a wish to have Shiny Applications which will display this data being deployed on a Shiny Server. The following Analysis types are essential for the initial version:
Cohort
Cohort Characterization
Incidence Rate
Cohort Pathway (Sankey diagram visualization which is currently not available in ATLAS)
Design Ideas and Functional Requirements
An application option should be added which will have the corresponding Shiny Applications being available and "hydrated" with associated Analysis execution results data for its further publishing to a configured Posit Connect Shiny Server
The Configuration will be extended with a possibility to integrate with Posit Connect Server (URL, API credentials). There is no intention to have multiple instances configured in the initial implementation
For each Analysis type from the requirements block a specific Shiny Application will be designed with "placeholders" to be populated with data after their availability as the result of an Analysis execution. The Shiny Applications should as much as possible functionally and by their look and feel coincide with the appropriate ATLAS visualizations. The source code will be a part of WebAPI or a separate repository which will be accessible as a dependency during the build process
Each Analysis Execution item will get an option to publish its results to a configured Posit Connect Server via corresponding publishing API. Once being successfully published there should be a marker for that in WebAPI application data so that any further publications of the same Analysis Execution Results are (im)possible
There is no intention to change the result data format which is used for visualizations by ATLAS and prepared by WebAPI so that the Shiny Applications to be designed should not neglect it
Open Questions
an open source version of Posit Shiny Server may be supported in future releases
additionally to the main publishing functionality there might be an option to download a Shiny Application together with the Analysis Execution Results so that it can be manually deployed to a Shiny Server or used directly in RStudio
it might be necessary to design an auxiliary navigation Shiny Application so that it will be possible then on Posit Connect user interface to easily find a published Shiny Application of interest filtering by Analysis type, publication date, data source name or any other available discriminator
multiple connected Posit Connect Servers support
any specific unpublish / republish Shiny Application scenarios
Business Requirements
Organizations have their Shiny Server installations with custom Shiny applications available. The standardized analytics are separately available in ATLAS after a particular Analysis is executed and its results are available as visuals (tables, graphs)
As soon as Analysis Results data is available there is a wish to have Shiny Applications which will display this data being deployed on a Shiny Server. The following Analysis types are essential for the initial version:
Design Ideas and Functional Requirements
An application option should be added which will have the corresponding Shiny Applications being available and "hydrated" with associated Analysis execution results data for its further publishing to a configured Posit Connect Shiny Server
The Configuration will be extended with a possibility to integrate with Posit Connect Server (URL, API credentials). There is no intention to have multiple instances configured in the initial implementation
For each Analysis type from the requirements block a specific Shiny Application will be designed with "placeholders" to be populated with data after their availability as the result of an Analysis execution. The Shiny Applications should as much as possible functionally and by their look and feel coincide with the appropriate ATLAS visualizations. The source code will be a part of WebAPI or a separate repository which will be accessible as a dependency during the build process
Each Analysis Execution item will get an option to publish its results to a configured Posit Connect Server via corresponding publishing API. Once being successfully published there should be a marker for that in WebAPI application data so that any further publications of the same Analysis Execution Results are (im)possible
There is no intention to change the result data format which is used for visualizations by ATLAS and prepared by WebAPI so that the Shiny Applications to be designed should not neglect it
Open Questions
@mrechkem @Sathyaraos