cal training data

cal.py

import pandas as pd
from datetime import datetime

cal = pd.read_csv('cal.csv')

# clean summary
cal = cal.query("summary not in ('Mid-week Meditation', 'Coffee Chat: Meet our New Hires!')")
cal = cal[cal['summary'].str.contains("Company Holiday") == False]

# clean dtstart
parsed = pd.to_datetime(cal["dtstart"], errors="coerce").fillna(pd.to_datetime(cal["dtstart"], format="%Y-%d-%m", errors="coerce"))
ordinal = pd.to_numeric(cal["dtstart"], errors="coerce").apply(lambda x: pd.Timestamp("1899-12-30")+pd.Timedelta(x, unit="D"))
 
cal['dtstart'] = parsed.fillna(ordinal)
# still might have problems
cal = cal.dropna()
cal['dtstart'] = cal['dtstart'].apply(lambda x: datetime.replace(x, tzinfo=None))

cal = cal.reset_index(drop=True)
cal['weekday'] = cal['dtstart'].dt.weekday
cal = cal.sort_values(by='dtstart', ascending=False)

#cal.to_csv("out.csv", columns=['dtstart', 'weekday', 'summary'])

gr = cal.groupby(cal['weekday'])['dtstart'].count()
print(gr)