P281424/AMR
1
0
Fork 0
mirror of https://github.com/msberends/AMR.git synced 2026-05-14 03:10:50 +02:00
Code Releases Activity

Revert mo_matching_score.R; fix #288 in mo.R pre-filter instead

Browse Source 
Issue #288: when the abbreviated-genus pre-filter (≤3 char genus) yields
exactly one candidate whose species (and subspecies) word(s) exactly match
the input, narrow filtr to that single candidate and set
minimum_matching_score = 0. This bypasses the automatic 0.55 cutoff that
only runs in the is.null() branch, so "S. apiospermum" resolves to
Scedosporium apiospermum without touching the validated scoring formula.

https://claude.ai/code/session_01VH4Ju4Xq9aW1AHuoVbjGEo
This commit is contained in:
Claude
2026-05-06 18:07:37 +00:00
parent b3b8d301ff
commit d23004641f
2 changed files with 14 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
R/mo.R
View File

@@ -352,6 +352,18 @@ as.mo <- function(x,
(MO_lookup_current$species_first == substr(x_parts[2], 1, 1) |
MO_lookup_current$subspecies_first == substr(x_parts[2], 1, 1) |
MO_lookup_current$subspecies_first == substr(x_parts[3], 1, 1)))
# Issue #288: if the species (and subspecies) word(s) in the input exactly match
# exactly one candidate, use only that candidate and bypass the 0.55 cutoff.
# This prevents prevalent bacteria from outranking a rarer organism whose species
# epithet is an unambiguous exact match, e.g. "S. apiospermum" → Scedosporium.
sp_exact <- tolower(MO_lookup_current$species[filtr]) == x_parts[2]
if (length(x_parts) == 3) {
sp_exact <- sp_exact & tolower(MO_lookup_current$subspecies[filtr]) == x_parts[3]
}
if (sum(sp_exact) == 1) {
filtr <- filtr[sp_exact]
minimum_matching_score <- 0
}
} else {
filtr <- which(MO_lookup_current$full_first == substr(x_parts[1], 1, 1) |
MO_lookup_current$species_first == substr(x_parts[2], 1, 1) |