This script processes the attendance list csv from a teams meeting, and plots some useful graphs

teams_meeting_attendance_analysis.py

import pandas as pd
import matplotlib.pyplot as plt

# --------------- You may need to change this -------------
file_name = "meetingAttendanceList.csv"
user_fullname_colname = "Nom complet"
user_action_colname = "Action de l’utilisateur"
joined_text = "Rejoint"
left_text = "A quitté l'appel"
date_and_time_colname = "Date et heure"
# ---------------------------------------------------------


# note: make sure to convert your attendance list to utf-8 before hand
df = pd.read_csv(file_name, sep="\t", encoding="utf-8")
df["datetime"] = pd.to_datetime(df[date_and_time_colname], format="%d/%m/%Y à %H:%M:%S")
del df["Date et heure"]

# merging time spans for people who connected several times
for user in df[user_fullname_colname].unique():
    user_mask = df[user_fullname_colname] == user
    joined = df.loc[user_mask & (df[user_action_colname] == joined_text)]
    left = df.loc[user_mask & (df[user_action_colname] == left_text)]
    if len(joined) > 1 or len(left) > 1:
        assert len(joined) == len(left)
        nb_slots = len(joined)
        tot_time = None
        for i in range(nb_slots):
            j = joined.iloc[i]["datetime"]
            l = left.iloc[i]["datetime"]
            if tot_time is None:
                tot_time = (l - j)
            else:
                tot_time += (l - j)
        for j in range(nb_slots):
            if j != 0:
                df.drop(joined.iloc[j].name, axis=0, inplace=True)
            if j != (nb_slots - 1):
                df.drop(left.iloc[j].name, axis=0, inplace=True)
        df.loc[user_mask, "total_time"] = tot_time
    else:
        assert len(joined) > 0
        left = df["datetime"].iloc[-1] if len(left) == 0 else left["datetime"].item()
        df.loc[user_mask, "total_time"] = left - joined["datetime"].item()

# de-pivoting so that there is now one row per attendee
df = df.set_index([user_fullname_colname, user_action_colname])
df = df.unstack()
df.columns = ['left_datetime', 'joined_datetime', '_', 'total_time']
df = df[['total_time', 'joined_datetime', 'left_datetime']]

# Histogram of attendance time
df['total_time_mins'] = df['total_time'].dt.total_seconds() / 60
df['total_time_mins'].hist(bins=20)
plt.title("Attendance total time in minutes (#attendees = %s)" % len(df))
plt.show()

Currently this generates a single plot looking like this: