From 4c0567addc163471319cf466d0d70bfae2ded8fd Mon Sep 17 00:00:00 2001
From: Adrien Klose <ikenex3@gmail.com>
Date: Thu, 30 May 2024 12:51:19 +0200
Subject: [PATCH] fix bug that lead to missing questions in the filtering
---
ideas_practice/code/bioasq_extract_yes_no.ipynb | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/ideas_practice/code/bioasq_extract_yes_no.ipynb b/ideas_practice/code/bioasq_extract_yes_no.ipynb
index c71ede4..d688fc1 100644
--- a/ideas_practice/code/bioasq_extract_yes_no.ipynb
+++ b/ideas_practice/code/bioasq_extract_yes_no.ipynb
@@ -125,6 +125,7 @@
" writer.writerow(header_tsv)\n",
" for question in filtered_data:\n",
" id += 1\n",
+ " written = False\n",
" if \"triples\" in question:\n",
" for triple in question[\"triples\"]:\n",
" # check for the filtering criteria\n",
@@ -140,7 +141,12 @@
" relation = base[-1].split(\"#\")[-1] # some of the links include a # for the exact path that needs to be removed\n",
" \n",
" writer.writerow([question[\"id\"], id, triple[\"s\"], triple[\"p\"], triple[\"o\"], \"\", relation, \"\", \"\", question[\"body\"], question[\"exact_answer\"]])\n",
+ " written = True\n",
" else:\n",
+ " writer.writerow([question[\"id\"], id, \"\", \"\", \"\", \"\", \"\", \"\", \"\", question[\"body\"], question[\"exact_answer\"]])\n",
+ " written = True\n",
+ " # if all triples are filtered the question would not be written without this\n",
+ " if not written:\n",
" writer.writerow([question[\"id\"], id, \"\", \"\", \"\", \"\", \"\", \"\", \"\", question[\"body\"], question[\"exact_answer\"]])"
]
},
--
GitLab