Skip to contents

Process raw Independent Monitoring (IM) data into missed-children rates

Source: R/clean_lqas_im.R
process_im.Rd

Cleans the raw POLIS Im table with the same recipe as clean_afp() / clean_sia() – standardise names via the crosswalk, clean strings, parse and sanitise dates, fix admin names and (when a shape is supplied) reconcile admin GUIDs, then dedup by id – and computes the missed-children fraction per district-year separately for in-house and out-of-house monitoring. For each setting missed = 1 - sum(marked)/sum(checked), falling back to mean(result) when no children were checked, with a Valid/Invalid status flag (Invalid when the result is missing or negative).

Usage

process_im(
  im,
  cfg = polis_active_config(),
  shape = NULL,
  verbose = TRUE,
  summary = TRUE
)

Arguments

im

Raw IM table (data.frame/tibble).

cfg

A polis_config() object (default polis_config()); its crosswalk maps the raw column names to canonical snake_case.

shape

Optional district shape (an sf polygon layer or a long ADM2 attribute table) for admin GUID reconciliation via reconcile_admin_guids() (keyed on year). Default NULL.

verbose

Emit cli progress + the one-line roll-up result (default TRUE).

summary

Emit the full cli summary panel (per-setting missed-children). Default TRUE; run_pipeline() passes FALSE to keep the pipeline terse.

Value

A list with district (per-district-year, carrying both missed_frac_inhouse / im_status_inhouse and missed_frac_outhouse / im_status_outhouse plus their checked/marked totals) and meta.

Examples

im <- data.frame(
  Id = 1L, Admin0 = "NIGERIA", Admin1 = "KANO", Admin2 = "NASSARAWA",
  Admin2GUID = "{A2}", ActivityPlannedDateFromYear = 2024L,
  HouseholdsNumberChildrenChecked = 20L,
  HouseholdsNumberChildrenMarked = 18L, HouseholdsResult = NA_real_,
  OutOfHouseNumberChildrenChecked = 10L,
  OutOfHouseNumberChildrenMarked = 8L, OutOfHouseResult = NA_real_
)
process_im(im, verbose = FALSE)$district
#> # A tibble: 1 × 13
#>    year adm2_guid adm0    adm1  adm2      n_checked_inhouse n_marked_inhouse
#>   <int> <chr>     <chr>   <chr> <chr>                 <dbl>            <dbl>
#> 1  2024 {A2}      NIGERIA KANO  NASSARAWA                20               18
#> # ℹ 6 more variables: missed_frac_inhouse <dbl>, n_checked_outhouse <dbl>,
#> #   n_marked_outhouse <dbl>, missed_frac_outhouse <dbl>,
#> #   im_status_inhouse <chr>, im_status_outhouse <chr>