fix: Fix discrepancies in WFH score [DHIS2-22201]

[enricocolasante]

Fix discrepancies in d2:zScoreWFH DHIS2-22201

Problem

• Wrong reference table between 87 and 110 cm. The weight-for-height tables used WHO weight-for-length values (meant for children measured lying down) all the way up to 110 cm, and only switched to weight-for-height above 110 cm. When age is unknown, the usual convention is to switch at 87 cm, so children 87–110 cm tall were compared against a reference 0.1–0.4 kg too light. Their z-scores came out too high, so a malnourished child could score above a cut-off like −2 SD.
• Corrupted row. The female row at 112 cm had an LMS parameter (0.09507) as its −3 SD value, with every other column shifted by one.
• Truncated results. Interpolated z-scores were cut off after 2 decimals instead of rounded, so float noise turned e.g. −0.2 into −0.19.

Changes

• Regenerated both WFH tables from the WHO reference data (LMS parameters), rounded to 1 decimal like WHO's published tables:
  • weight-for-length below 87 cm, weight-for-height from 87 cm
  • 0.1 cm steps from 45.0 to 120.0 cm (751 rows per sex, up from 151)
  • every existing row is unchanged except 87–110 cm and the female 112 cm row
• Fixed the female 112 cm row.
• ZScore: round interpolated z-scores to 2 decimals instead of truncating them, and remove the string-based formatting helpers.

This also affects d2:zScoreWFA and d2:zScoreHFA, which share the rounding code: their results between two SD values can change by up to 0.01.

Known limitations

• Z-scores are interpolated linearly between SD weights rounded to 0.1 kg, so results can differ from WHO Anthro by up to ~0.1.
• Values beyond ±3 SD are still reported as ±3.5; Anthro keeps counting past ±3.
• Heights must match a 0.1 cm row exactly.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
100.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud