target_market_share(..., by_company = TRUE) outputs unexpected columns

[jdhoffa]

In particular, when comparing the output of target_market_share(..., by_company = FALSE) and target_market_share(..., by_company = TRUE), I would expect to see only the added column name_ald.

What I see in practice is three added columns: name_ald, and then the two unexpected columns sector_weighted_production and technology_weighted_production.

library(r2dii.data)
library(r2dii.match) 
library(r2dii.analysis)
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union

matched <- match_name(loanbook_demo, ald_demo) %>%
  prioritize()

out_false <- target_market_share(
  matched,
  ald_demo,
  scenario_demo_2020,
  region_isos_demo,
  by_company = FALSE
)

out_true <- target_market_share(
  matched,
  ald_demo,
  scenario_demo_2020,
  region_isos_demo,
  by_company = TRUE
  )
#> Warning: You've supplied `by_company = TRUE` and `weight_production = TRUE`.
#> This will result in company-level results, weighted by the portfolio
#> loan size, which is rarely useful. Did you mean to set one of these
#> arguments to `FALSE`?

# similarly, as expected, if by_company = TRUE, the output has the same columns 
# regardless of if weight_production = T or F
setdiff(names(out_true), names(out_false))
#> [1] "name_ald"                       "sector_weighted_production"    
#> [3] "technology_weighted_production"

# I would expect the only added column to be `name_ald`
setdiff(names(out_true), names(out_false)) == "name_ald"
#> [1]  TRUE FALSE FALSE

^{Created on 2021-05-03 by the reprex package (v2.0.0)}

[maurolepore]

For the record, the test for this may come near the one that avoids the regression to #186:

test_that("outputs same names regardless of the value of `weight_production` (#186)", {
  out_weighted <- target_market_share(
    fake_matched(),
    fake_ald(),
    fake_scenario(),
    region_isos_stable,
    weight_production = TRUE
  )

  out_unweighted <- target_market_share(
    fake_matched(),
    fake_ald(),
    fake_scenario(),
    region_isos_stable,
    weight_production = FALSE
  )

  diff_names <- setdiff(names(out_unweighted), names(out_weighted))

  expect_equal(diff_names, character(0))
})