TODO.md

In southwick-associates/sadash: Prepare Agency Dashboard Data for state-level dashboards

Version 1.2

• Improve documentation, and store in vignettes
  • intro (overview & template code setup)
  • history
  • dashboard
  • dive

Ideas Version 1.3+

• Generate a sample dashboard from salic (rather than the full version)

  • This will also probably provide an opportunity to add more sample data and include examples for certain other functions

• run_visual(), etc.:

  • may want nicer labelling on hover for plotly

More Ideas

We might want tests using testthat (similar to what was done in salic). Making changes to code is quite a pain currently b/c it is laborious to find things like syntax errors. Although even including examples (which get run with R CMD check) would improve this.

It may be that functions will need to be tweaked, improved, etc. on the backend. However, it is difficult to do that reliably without tests (and probably sample data). Therefore, any changes should begin with tests to ensure that functionality is not broken, or a change to a function leads to an error in calculation.

• Consider limiting the county results to dashboard_yrs by making sadash-specific versions of est_part, etc. (would look something like the code below)
  • first make at least one sample dataset (dashboard) and at least one test (output of calc_metrics() matches sample data) to ensure that the primary functionality doesn't get broken with code changes (instead of hoping for the best or laboriously comparing to existing state code)
  • would also want to drop the filter at the end of calc_metrics(), which is nice because the code is then cleaner & fewer threshold warnings would be produced (a benefit on the user end)

#' Estimate participants
#' 
#' This is basically a wrapper for salic::est_part() that filters out non-dashboard_yrs
#' for county-level results
#' 
#' @params dashboard_yrs years for dashboard focus
#' @export
est_part <- function(
    history, segment = "tot", test_threshold = 20, show_test_stat = FALSE,
    suppress_warning = FALSE, outvar = "participants", dashboard_yrs
) {
    if (segment == "county") {
        history <- filter(history, year %in% dashboard_yrs)
    }
    salic::est_part(history, segment, test_threshold, show_test_stat, 
                    suppress_warning, outvar)
}

• maybe make a fill_segs() function to ensure tableau gets res/nonres, etc.
  • filling with zeroes for any missing categories (in res, sex, agecat)
  • probably won't happen for sex, agecat, but might be worth generalizing
• maybe investigate scaleup warnings with VA resident permissions (in counties)
  • the discrepancies are quite small (almost certainly rounding error) but probably worth a closer look

southwick-associates/sadash documentation built on Jan. 6, 2022, 5:17 p.m.