Perhaps I cannot see this, but why would you use pivot() when you can use other table widgets in R and shiny?
What would be the use case?

My team also has this use case!

I've taught teammates to use the syntax of rpivotTable.

For QA'ing our data, it's great. Slice-split-repeat is a great way to drill down and find anomalies in quantitative data.

Now we want to be able to put two, three, or four (static) tables in a single Rmarkdown document, each of which is a different tabulation of the same dataset. (For reporting purposes.)

pivotUI(), as far as I know, is a big cat. It doesn't seem to play well with others on the same HTML page. I'm hoping the static version — because it doesn't need to expand/contract — can let us put multiple tables on the same HTML page.

A final, seemingly minor, requirement: We also need (want?) at least two levels of row headings. Some other packages will accommodate that. But we'd like to get by with learning and using just one (fairly simple!) syntax for spec'ing tables. The syntax for rpivotTable is great for our needs.

The underlying javascript package added "showUI = FALSE" as a parameter in June 2018, for this purpose: https://github.com/nicolaskruchten/pivottable/wiki/Frequently-Asked-Questions#params

Unfortunately, rpivotTables has not been updated since that was added, so is using an old version of the javascript.

@smartinsightsfromdata or @nicolaskruchten - apologies for this blast from the past - would you be able to update the htmlwidget files to use the latest version of pivottables?

+1. There seem to be some other things breaking that might be fixed by an update. For example, in the UI, checked checkboxes now look blank rather than checked. (?)

@dholstius I've made a fork and done it myself, but in a very haphazard manner. It works, but I'm sure it's not best practice. Feel free to try it out - see PR #121.