Skip to contents

Stack indicator outcomes into one long-format table

Usage

anlz_indicators(..., bay_segments = c("OTB", "HB", "MTB", "LTB"))

Arguments

...

named data.frames, each with bay_segment, yr, and outcome columns. A data.frame may already have its own indicator column if it bundles more than one indicator (e.g. anlz_wq_load's output already has indicator = "tn_load"/"tnhy_load"). Otherwise every row is labeled with that argument's name.

bay_segments

chr vector of bay segments to keep, defaults to c('OTB', 'HB', 'MTB', 'LTB')

Value

A data.frame with columns bay_segment, yr, indicator, and outcome (0-1, 1 = best)

Details

Stacks (bind_rows) rather than joins the inputs, since bay segment/year coverage differs by indicator. A segment/ year missing from one indicator's source data simply has no row for it, rather than an NA-filled one. Rows outside bay_segments or with an NA outcome are dropped.

This same stacking step is used at every level of the indicator hierarchy, once per category to build a category's indicator table from its raw anlz_* indicator outputs, and again inside anlz_score (via anlz_category) to stack the four category scores into one table before the final average. A category score is just another single-indicator input.

Examples

anlz_indicators(
  wq_attain = data.frame(bay_segment = 'OTB', yr = 2020, outcome = 0.8),
  load = data.frame(bay_segment = 'OTB', yr = 2020,
    indicator = c('tn_load', 'tnhy_load'), outcome = c(1, 0))
)
#>   bay_segment   yr indicator outcome
#> 1         OTB 2020 wq_attain     0.8
#> 2         OTB 2020   tn_load     1.0
#> 3         OTB 2020 tnhy_load     0.0