stoffelmauro
/
ConsultancyProject_2_ETL
1
0
Fork 1
Code Issues Pull Requests Packages Projects Releases Wiki Activity

update des etl_capacities_weekdays.py

main
mmaurostoffel 2025-01-09 15:07:38 +01:00
parent 2508b34ceb
commit a6cbe3bc29
1 changed files with 10 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
etl/src/data/etl_property_capacities_weekdays.py
View File

@ -8,6 +8,7 @@ d = data.load()
def property_capacities_weekdays(id: int, scrapeDate: str): def property_capacities_weekdays(id: int, scrapeDate: str):
extractions = d.extractions_propId_scrapeDate(id, scrapeDate).pl() extractions = d.extractions_propId_scrapeDate(id, scrapeDate).pl()
weekdays = ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']
df_calendar = pl.DataFrame() df_calendar = pl.DataFrame()
numWeeks = 0 numWeeks = 0
@ -19,12 +20,15 @@ def property_capacities_weekdays(id: int, scrapeDate: str):
df_calendar = df_calendar.with_columns(pl.Series(name="dates", values=columnTitles)) df_calendar = df_calendar.with_columns(pl.Series(name="dates", values=columnTitles))
df_calendar = df_calendar.with_columns((pl.col("dates").str.to_date())) df_calendar = df_calendar.with_columns((pl.col("dates").str.to_date()))
numWeeks = round((df_calendar.get_column("dates").max() - df_calendar.get_column("dates").min()).days / 7, 0) numWeeks = round((df_calendar.get_column("dates").max() - df_calendar.get_column("dates").min()).days / 7, 0)
df_calendar = df_calendar.with_columns(pl.col("dates").dt.weekday().alias("weekday")) df_calendar = df_calendar.with_columns(pl.col("dates").dt.weekday().alias("weekday_num"))
df_calendar = df_calendar.with_columns(pl.col("dates").dt.strftime("%A").alias("weekday"))
df_calendar = df_calendar.drop("dates") df_calendar = df_calendar.drop("dates")
df_calendar = df_calendar.group_by("weekday").agg(pl.col("column_0").sum()) df_calendar = df_calendar.group_by(["weekday", "weekday_num"]).agg(pl.col("column_0").sum())
df_calendar = df_calendar.with_columns((pl.col("column_0") / numWeeks).alias("weekday")) df_calendar = df_calendar.with_columns((pl.col("column_0") / numWeeks).alias("column_0"))
df_calendar = df_calendar.sort('weekday_num')
df_calendar = df_calendar.drop('weekday_num')
print(df_calendar)
result = {"scraping-date": scrapeDate, "weekday": df_calendar['weekday'].to_list(), 'capacities': df_calendar['column_0'].to_list()} result = {"scraping-date": scrapeDate, "weekday": df_calendar['weekday'].to_list(), 'capacities': df_calendar['column_0'].to_list()}
return result return result