tidyCDISC
Authors: Maya Gans, @Aaron-Clark , Robert Krajcik, Nate Mockler
Affiliations: Biogen
Abstract: The Table Generator is a custom shiny input binding that allows you to drag and drop columns from imported datasets, and pair them with summary statistics based on whether the columns are categorical or numeric. The table can also be broken into groups using our group by dropdown, and the data can even be filtered. You can export the gt table as either an html or csv.
Full Description: Clinical trials produce hundreds of tables, requiring validation and a high amount of time and resources to produce. Creating a product that can create these tables without programming effort while simultaneously allowing the flexibility to subgroup, filter, and reorder the underlying table structure has been a bane of pharma companies for a long time. In response to this, Biogen has created a package called tidyCDISC.
By leveraging this common data structure and the summary statistic patterns seen in clinical submissions, we've created a custom shiny input that generates a {gt} table by dragging pairs of columns and their summary statistic counterpart.
Each file the user uploads into the application creates a dropdown of blocks corresponding to column names. These columns can be pared with either numeric (mean, chg, anova) or categorical (freq) summary blocks to create a table.
The data can be filtered using our filter dropdown, and we even have a dropdown of commonly created tables -- rather than drag in all the block pairs needed for a demographic table, you can select 'DEMOGRAPHY' from the dropdown.
Table Type: interactive-Shiny
Submission Type: Other
Table: https://maya-gans.shinyapps.io/tablecontest-biogen/
Repo: github.com/biogen-inc/tidycdisc/tree/table-contest
RStudio Cloud:
DT package used:
gt package used: true
reactable package used:
flextable package used:
huxtable package used:
kableExtra package used:
Other packages: