jinilkim/ds_new_user_analy
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Enhance new user cohort optimization by adding language fallback for missing user data and including country in the information retrieval process

Browse Source
This commit is contained in:
Gnill82
2025-08-29 17:45:50 +09:00
parent c08c4cf483
commit a45c7581d5
1 changed files with 11 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
ds_new_user_analy.py
View File

@ -713,8 +713,9 @@ def get_new_user_cohort_optimized(
logger.info(f"login_comp에서 {len(login_comp_collected)}명의 정보 수집 완료")
# Step 4: log_return_to_lobby 인덱스에서 차선 정보 수집 (auth.id 2순위)
missing_uids = [uid for uid in uid_list if uid not in login_comp_collected]
# Step 4: log_return_to_lobby 인덱스에서 차선 정보 수집 (auth.id 2순위, language fallback)
# login_comp에서 수집되지 않은 유저 + language가 N/A인 유저들 처리
missing_uids = [uid for uid in uid_list if uid not in login_comp_collected or cohort[uid]['language'] == 'N/A']
if missing_uids:
logger.info(f"log_return_to_lobby 인덱스에서 {len(missing_uids)}명의 차선 정보 수집 중 (auth.id 2순위)...")
@ -743,7 +744,7 @@ def get_new_user_cohort_optimized(
"top_hits": {
"size": 1,
"sort": [{"@timestamp": {"order": "desc"}}],
"_source": ["body.nickname", "auth.id"]
"_source": ["body.nickname", "auth.id", "country"]
}
}
}
@ -765,12 +766,18 @@ def get_new_user_cohort_optimized(
info_hit = bucket["info"]["hits"]["hits"][0]["_source"] if bucket["info"]["hits"]["hits"] else {}
# 차선 정보 업데이트 (login_comp에서 수집되지 않은 유저만)
# 차선 정보 업데이트
if uid in cohort:
# nickname 업데이트 (없는 경우에만)
if cohort[uid]['nickname'] == 'N/A':
cohort[uid]['nickname'] = info_hit.get('body', {}).get('nickname', 'N/A')
# language fallback: country 값 사용 (language가 N/A인 경우에만)
if cohort[uid]['language'] == 'N/A':
country = info_hit.get('country')
if country:
cohort[uid]['language'] = f"country-{country}"
# auth.id 수집 (2순위, 없는 경우에만)
if cohort[uid]['auth_id'] == 'N/A':
auth_id = info_hit.get('auth', {}).get('id')