bakery-admin/bakery-ia
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Start fixing forecast service 18

Browse Source
This commit is contained in:
Urtzi Alfaro
2025-07-30 08:41:47 +02:00
parent c10a745695
commit 024290e4c0
2 changed files with 25 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
services/training/app/ml/data_processor.py
View File

@@ -324,15 +324,18 @@ class BakeryDataProcessor:
if 'date' not in weather_clean.columns and 'ds' in weather_clean.columns:
weather_clean = weather_clean.rename(columns={'ds': 'date'})
# ✅ FIX: Ensure timezone consistency
# 🔧 CRITICAL FIX: Ensure both DataFrames have compatible datetime formats
weather_clean['date'] = pd.to_datetime(weather_clean['date'])
daily_sales['date'] = pd.to_datetime(daily_sales['date'])
# Remove timezone info from both to make them compatible
# ✅ NEW FIX: Normalize both to timezone-naive datetime for merge compatibility
if weather_clean['date'].dt.tz is not None:
weather_clean['date'] = weather_clean['date'].dt.tz_localize(None)
# Convert timezone-aware to UTC then remove timezone info
weather_clean['date'] = weather_clean['date'].dt.tz_convert('UTC').dt.tz_localize(None)
if daily_sales['date'].dt.tz is not None:
daily_sales['date'] = daily_sales['date'].dt.tz_localize(None)
# Convert timezone-aware to UTC then remove timezone info
daily_sales['date'] = daily_sales['date'].dt.tz_convert('UTC').dt.tz_localize(None)
# Map weather columns to standard names
weather_mapping = {
@@ -390,15 +393,19 @@ class BakeryDataProcessor:
if 'date' not in traffic_clean.columns and 'ds' in traffic_clean.columns:
traffic_clean = traffic_clean.rename(columns={'ds': 'date'})
# 🔧 FIX: Ensure timezone awareness before merge
# 🔧 CRITICAL FIX: Ensure both DataFrames have compatible datetime formats
traffic_clean['date'] = pd.to_datetime(traffic_clean['date'])
# If timezone-naive, localize to UTC
if traffic_clean['date'].dt.tz is None:
traffic_clean['date'] = traffic_clean['date'].dt.tz_localize('UTC')
# If already timezone-aware but not UTC, convert to UTC
elif str(traffic_clean['date'].dt.tz) != 'UTC':
traffic_clean['date'] = traffic_clean['date'].dt.tz_convert('UTC')
daily_sales['date'] = pd.to_datetime(daily_sales['date'])
# ✅ NEW FIX: Normalize both to timezone-naive datetime for merge compatibility
# This prevents the "datetime64[ns] and datetime64[ns, UTC]" merge error
if traffic_clean['date'].dt.tz is not None:
# Convert timezone-aware to UTC then remove timezone info
traffic_clean['date'] = traffic_clean['date'].dt.tz_convert('UTC').dt.tz_localize(None)
if daily_sales['date'].dt.tz is not None:
# Convert timezone-aware to UTC then remove timezone info
daily_sales['date'] = daily_sales['date'].dt.tz_convert('UTC').dt.tz_localize(None)
# Map traffic columns to standard names
traffic_mapping = {