import pandas as pd # import matplotlib.pyplot as plt timelog = pd.read_csv("timelog-titled.csv") timelog["started"] = pd.to_datetime(timelog["started"]).dt.tz_convert("US/Eastern") timelog["recorded"] = pd.to_datetime(timelog["recorded"]).dt.tz_convert("US/Eastern") timelog["time"] = 30 # A pomodoro started before 3am Eastern time is considered to be a continuation # of the day before, so we are, effectively, on West Coast time for determining # the day we want to associate a time entry with. PomodoroPrompt saves as UTC. timelog["date"] = timelog["started"].dt.tz_convert("US/Pacific").dt.date timelog["day_of_week"] = pd.to_datetime(timelog["date"]).dt.day_name() # Condense duplicate entries by date, summing the minutes spent. tl = timelog.groupby(["date", "description"]).agg({"time": ['sum']}, {"started": ['max']}).reset_index()