University Timelog Analysis

import numpy as np
import pandas as pd
import seaborn as sns
import plotly.express as px
import plotly.graph_objects as go
import matplotlib.pyplot as plt
from datetime import timedelta

Load Raw Data

# Load data
source = {}
source[2019] = pd.read_csv('data/University_Timelogs_2019-01-01_to_2020-01-01.csv')
source[2020] = pd.read_csv('data/University_Timelogs_2020-01-01_to_2020-12-31.csv')
source[2021] = pd.read_csv('data/University_Timelogs_2021-01-01_to_2021-12-31.csv')
source['ba'] = pd.read_csv('data/BA_Timelogs_2019-04-01_to_2020-03-31.csv')

Data Cleanup

# Convert data
for k, df in source.items():
    source[k]['Start'] = pd.to_datetime(df['Start date']) + pd.to_timedelta(df['Start time'])
    source[k]['End'] = pd.to_datetime(df['End date']) + pd.to_timedelta(df['End time'])
    source[k]['Duration'] = pd.to_timedelta(df['Duration'])

# Get distinct lectures and activity types
activity_types = ['Lecture', 'Exercise', 'Reading', 'Project', 'Study', 'Preparation']
# Convert tags to activity type column
for k, df in source.items():
    if k == 'ba':
        source[k]['Lecture'] = 'BA Thesis'
        source[k]['Type'] = None
        continue
    source[k]['Lecture'] = None
    source[k]['Type'] = None
    for i, t in df['Tags'].items():
        if t in activity_types:
            source[k]['Lecture'][i] = df['Description'][i].strip()
            source[k]['Type'][i] = t
     
# List of postfixes to convert to activity types
postfixes = {
    'Projekt': 'Project',
    'Prüfungsvorbereitung': 'Study',
    'Klausurvorbereitung': 'Study',
    'Klausur': 'Study',
    'Prüfung': 'Study',
    'Übungen': 'Exercise',
    'Aufgabe': 'Exercise',
    'Einführung': 'Lecture',
    'Vorbereitung': 'Preparation',
    'Abschlussbericht': 'Exercise',
    'Bericht': 'Exercise',
    'Aufgaben': 'Exercise',
    'Dokumentation': 'Exercise',
    'Lernen': 'Study',
    'Essay': 'Exercise',
    'Homework': 'Exercise',
    'Präsentation': 'Exercise',
    }
# Convert postfixes to type and lecture
for k, df in source.items():
    for i, d in df[df['Lecture'].isnull()]['Description'].items():
        for p, t in postfixes.items():
            if d.lower().endswith(p.lower()):
                source[k]['Lecture'][i] = d[:-len(p)].strip()
                source[k]['Type'][i] = t
                break
        else:
            source[k]['Lecture'][i] = d.strip()
            source[k]['Type'][i] = 'Lecture'
    source[k]['Type'] = pd.Categorical(df['Type'])

Combine Data

combined_data = pd.concat(source)
combined_data

		User	Client	Project	Task	Description	Billable	Start date	Start time	End date	End time	Duration	Tags	Amount ()	Start	End	Lecture	Type
2019	0	Christoph Jabs	NaN	Hochschule	NaN	Akustik	No	2019-01-09	16:30:13	2019-01-09	18:00:13	0 days 01:30:00	NaN	NaN	2019-01-09 16:30:13	2019-01-09 18:00:13	Akustik	Lecture
	1	Christoph Jabs	NaN	Hochschule	NaN	Microcontroller	No	2019-01-10	10:45:17	2019-01-10	13:00:19	0 days 02:15:02	Hochschule	NaN	2019-01-10 10:45:17	2019-01-10 13:00:19	Microcontroller	Lecture
	2	Christoph Jabs	NaN	Hochschule	NaN	Regelungstechnik Projekt	No	2019-01-10	13:55:41	2019-01-10	14:39:30	0 days 00:43:49	Hochschule	NaN	2019-01-10 13:55:41	2019-01-10 14:39:30	Regelungstechnik	Project
	3	Christoph Jabs	NaN	Hochschule	NaN	Technische Mechanik	No	2019-01-10	14:45:19	2019-01-10	18:00:19	0 days 03:15:00	Hochschule	NaN	2019-01-10 14:45:19	2019-01-10 18:00:19	Technische Mechanik	Lecture
	4	Christoph Jabs	NaN	Hochschule	NaN	Regelungstechnik Projekt	No	2019-01-11	15:39:43	2019-01-11	19:13:06	0 days 03:33:23	Hochschule	NaN	2019-01-11 15:39:43	2019-01-11 19:13:06	Regelungstechnik	Project
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
ba	346	Christoph Jabs	NaN	Bacherlorarbeit	NaN	Review: Overall	No	2020-02-09	19:20:05	2020-02-09	19:31:54	0 days 00:11:49	NaN	NaN	2020-02-09 19:20:05	2020-02-09 19:31:54	BA Thesis	NaN
	347	Christoph Jabs	NaN	Bacherlorarbeit	NaN	Review: Overall	No	2020-02-10	07:10:00	2020-02-10	07:34:00	0 days 00:24:00	NaN	NaN	2020-02-10 07:10:00	2020-02-10 07:34:00	BA Thesis	NaN
	348	Christoph Jabs	NaN	Bacherlorarbeit	NaN	Review: Overall	No	2020-02-10	08:04:14	2020-02-10	09:17:25	0 days 01:13:11	NaN	NaN	2020-02-10 08:04:14	2020-02-10 09:17:25	BA Thesis	NaN
	349	Christoph Jabs	NaN	Bacherlorarbeit	NaN	Review: Overall	No	2020-02-10	11:00:00	2020-02-10	12:50:00	0 days 01:50:00	NaN	NaN	2020-02-10 11:00:00	2020-02-10 12:50:00	BA Thesis	NaN
	350	Christoph Jabs	NaN	Bacherlorarbeit	NaN	Review: Overall	No	2020-02-10	13:55:57	2020-02-10	17:44:15	0 days 03:48:18	NaN	NaN	2020-02-10 13:55:57	2020-02-10 17:44:15	BA Thesis	NaN

1443 rows × 17 columns

Manual Changes

combined_data[combined_data['Lecture'] == 'Betriebs- und Kommunikationssysteme']['Lecture'] = 'Betriebssysteme und Echtzeit'
combined_data[combined_data['Lecture'] == 'Buisness English']['Lecture'] = 'Business English'
combined_data[combined_data['Lecture'] == 'Signale und Systeme II']['Lecture'] = 'Digitale Signalverarbeitung'
combined_data[combined_data['Lecture'] == 'Seminar Leitfaden Abschlussarbeit']['Lecture'] = 'Leitfaden Abschlussarbeit'

grouped_data = combined_data.groupby('Lecture')
grouped_data['Duration'].sum()

Lecture
Akustik                                                 0 days 16:07:56
Algorithms and Complexity                               4 days 00:06:28
Approximation Algorithms                                3 days 02:51:25
BA Thesis                                              13 days 20:59:44
BWL/Rechnungswesen                                      0 days 20:06:05
Betriebs- und Kommunikationssysteme                     0 days 01:50:00
Betriebssysteme                                         0 days 00:50:32
Betriebssysteme und Echtzeit                            1 days 04:27:13
Betriebssysteme und Echtzeit Praktikum                  0 days 13:14:22
Building AI                                             0 days 04:26:13
Buisness English                                        1 days 00:41:40
Business English                                        0 days 03:40:33
CS Colloquium                                           0 days 12:20:42
Combinatorial Optimization                              2 days 02:59:31
Data Analysis with Python                               0 days 19:16:09
Design and Analysis of Algorithms                       1 days 18:54:37
Digitale Signalverarbeitung                             0 days 09:34:25
Elektrische Antriebe                                    1 days 06:03:16
Elektrische Antriebe Praktikum                          0 days 21:23:57
Elements of AI                                          0 days 06:00:31
Finnish 1A                                              0 days 21:27:00
Finnish 1B                                              0 days 22:50:17
Finnish 2A                                              1 days 02:15:33
Introduction to Data Science                            3 days 06:03:35
Introduction to the Programming of Quantum Computers    0 days 15:35:44
Kommunikationssysteme                                   1 days 02:23:43
Kommunikationssysteme Praktikum                         0 days 10:40:34
Kreativer Systementwurf                                 1 days 19:02:09
Leistungselektronik                                     1 days 03:37:16
Leitfaden Abschlussarbeit                               0 days 01:30:00
Microcontroller                                         0 days 09:55:16
Microcontroller Praktikum                               0 days 03:15:00
Modellierung und Simulation heterogener Systeme         0 days 05:29:58
Network Analysis                                        1 days 15:12:49
Praktikumsbewerbung                                     0 days 00:42:55
Praxissemester                                          1 days 05:26:12
Präsentation Stipendium                                 0 days 06:39:21
Recht                                                   0 days 19:42:00
Regelungstechnik                                        1 days 17:14:41
Regelungstechnik II                                     0 days 20:48:53
Regelungstechnik II Praktikum                           0 days 19:45:18
Robotersysteme                                          2 days 10:30:34
Robotersysteme Praktikum                                0 days 22:06:53
Seminar Leitfaden Abschlussarbeit                       0 days 01:12:50
Signale und Systeme II                                  0 days 02:00:00
Software Engineering                                    0 days 06:37:13
Statistical Machine Learning                            5 days 09:41:23
Steuerungstechnik                                       0 days 04:42:50
Stipendium                                              0 days 00:08:46
String Processing Algorithms                            2 days 11:16:27
Studienbericht Studienstiftung                          0 days 02:48:39
TOEFL                                                   0 days 06:51:59
Technische Mechanik                                     0 days 16:20:10
Theoretische Informatik                                 4 days 02:52:57
Name: Duration, dtype: timedelta64[ns]

Lecture Info

lectures = pd.DataFrame.from_dict({
    'Akustik': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Algorithms and Complexity': {'ECTS': 9, 'completed': True, 'University': 'EKU'},
    'Approximation Algorithms': {'ECTS': 5, 'completed': True, 'University': 'UH'},
    'BA Thesis': {'ECTS': 14, 'completed': True, 'University': 'RT'},
    'BWL/Rechnungswesen': {'ECTS': 2, 'completed': True, 'University': 'RT'},
    'Betriebssysteme und Echtzeit': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Betriebssysteme und Echtzeit Praktikum': {'ECTS': 1, 'completed': True, 'University': 'RT'},
    'Building AI': {'ECTS': 2, 'completed': True, 'University': 'UH'},
    'Business English': {'ECTS': 2, 'completed': True, 'University': 'RT'},
    'CS Colloquium': {'ECTS': 5, 'completed': False, 'University': 'UH'},
    'Combinatorial Optimization': {'ECTS': 5, 'completed': True, 'University': 'UH'},
    'Data Analysis with Python': {'ECTS': 5, 'completed': True, 'University': 'UH'},
    'Design and Analysis of Algorithms': {'ECTS': 5, 'completed': True, 'University': 'UH'},
    'Digitale Signalverarbeitung': {'ECTS': 5, 'completed': False, 'University': 'RT'},
    'Elektrische Antriebe': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Elektrische Antriebe Praktikum': {'ECTS': 2, 'completed': True, 'University': 'RT'},
    'Elements of AI': {'ECTS': 2, 'completed': True, 'University': 'UH'},
    'Finnish 1A': {'ECTS': 3, 'completed': True, 'University': 'UH'},
    'Finnish 1B': {'ECTS': 2, 'completed': True, 'University': 'UH'},
    'Finnish 2A': {'ECTS': 3, 'completed': True, 'University': 'UH'},
    'Finnish 2B': {'ECTS': 2, 'completed': False, 'University': 'UH'},
    'Introduction to Data Science': {'ECTS': 5, 'completed': True, 'University': 'UH'},
    'Introduction to the Programming of Quantum Computers': {'ECTS': 5, 'completed': False, 'University': 'UH'},
    'Kommunikationssysteme': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Kommunikationssysteme Praktikum': {'ECTS': 1, 'completed': True, 'University': 'RT'},
    'Kreativer Systementwurf': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Leistungselektronik': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Leitfaden Abschlussarbeit': {'ECTS': 0, 'completed': False, 'University': 'RT'},
    'Microcontroller': {'ECTS': 5, 'completed': False, 'University': 'RT'},
    'Microcontroller Praktikum': {'ECTS': 2, 'completed': False, 'University': 'RT'},
    'Modellierung und Simulation heterogener Systeme': {'ECTS': 4, 'completed': False, 'University': 'RT'},
    'Network Analysis': {'ECTS': 5, 'completed': True, 'University': 'UH'},
    'Praktikumsbewerbung': {'ECTS': 0, 'completed': False, 'University': 'RT'},
    'Praxissemester': {'ECTS': 0, 'completed': False, 'University': 'RT'},
    'Präsentation Stipendium': {'ECTS': 0, 'completed': False, 'University': 'Studienstiftung'},
    'Recht': {'ECTS': 2, 'completed': True, 'University': 'RT'},
    'Regelungstechnik': {'ECTS': 3, 'completed': False, 'University': 'RT'},
    'Regelungstechnik II': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Regelungstechnik II Praktikum': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Robotersysteme': {'ECTS': 5, 'completed': True, 'University': 'RT'},
    'Robotersysteme Praktikum': {'ECTS': 3, 'completed': True, 'University': 'RT'},
    'Software Engineering': {'ECTS': 3, 'completed': False, 'University': 'RT'},
    'Statistical Machine Learning': {'ECTS': 9, 'completed': True, 'University': 'EKU'},
    'Steuerungstechnik': {'ECTS': 5, 'completed': False, 'University': 'RT'},
    'Stipendium': {'ECTS': 0, 'completed': False, 'University': 'Studienstiftung'},
    'String Processing Algorithms': {'ECTS': 0, 'completed': False, 'University': 'UH'},
    'Studienbericht Studienstiftung': {'ECTS': 0, 'completed': False, 'University': 'Studienstiftung'},
    'TOEFL': {'ECTS': 0, 'completed': False, 'University': 'UH'},
    'Technische Mechanik': {'ECTS': 4, 'completed': False, 'University': 'UH'},
    'Theoretische Informatik': {'ECTS': 9, 'completed': True, 'University': 'EKU'},
}, orient='index')
online_courses = [
    'Elements of AI',
    'Building AI',
    'Data Analysis with Python',
]
lectures['online course'] = False
lectures['online course'][lectures.index.isin(online_courses)] = True
lectures.index = lectures.index.set_names('Lecture')
lectures['Lecture'] = lectures.index

# Rename universities
lectures['University'][lectures['University'] == 'RT'] = 'Reutlingen'
lectures['University'][lectures['University'] == 'EKU'] = 'Tübingen'
lectures['University'][lectures['University'] == 'UH'] = 'Helsinki'

hours_per_ects = 25
lectures['Time effort'] = grouped_data['Duration'].sum()
lectures['Required time'] = (lectures['ECTS'] * hours_per_ects).apply(lambda t: timedelta(hours=t))
lectures['Time fraction'] = lectures['Time effort'] / lectures['Required time']
lectures['Time fraction'] = lectures['Time fraction'].replace([np.inf, -np.inf], np.nan)

# Drop unfinished lectures
lectures = lectures[lectures['completed'] == True]

# Convert time to string for display
def conv(t):
    ts = t.total_seconds()
    hours, remainder = divmod(ts, 3600)
    minutes, _ = divmod(remainder, 60)
    return ('{} hours {} minutes').format(int(hours), int(minutes)) 

lectures['Time spent'] = lectures['Time effort'].apply(conv)

lectures

	ECTS	completed	University	online course	Lecture	Time effort	Required time	Time fraction	Time spent
Lecture
Akustik	3	True	Reutlingen	False	Akustik	0 days 16:07:56	3 days 03:00:00	0.215096	16 hours 7 minutes
Algorithms and Complexity	9	True	Tübingen	False	Algorithms and Complexity	4 days 00:06:28	9 days 09:00:00	0.427146	96 hours 6 minutes
Approximation Algorithms	5	True	Helsinki	False	Approximation Algorithms	3 days 02:51:25	5 days 05:00:00	0.598856	74 hours 51 minutes
BA Thesis	14	True	Reutlingen	False	BA Thesis	13 days 20:59:44	14 days 14:00:00	0.951416	332 hours 59 minutes
BWL/Rechnungswesen	2	True	Reutlingen	False	BWL/Rechnungswesen	0 days 20:06:05	2 days 02:00:00	0.402028	20 hours 6 minutes
Betriebssysteme und Echtzeit	3	True	Reutlingen	False	Betriebssysteme und Echtzeit	1 days 04:27:13	3 days 03:00:00	0.379381	28 hours 27 minutes
Betriebssysteme und Echtzeit Praktikum	1	True	Reutlingen	False	Betriebssysteme und Echtzeit Praktikum	0 days 13:14:22	1 days 01:00:00	0.529578	13 hours 14 minutes
Building AI	2	True	Helsinki	True	Building AI	0 days 04:26:13	2 days 02:00:00	0.088739	4 hours 26 minutes
Business English	2	True	Reutlingen	False	Business English	0 days 03:40:33	2 days 02:00:00	0.073517	3 hours 40 minutes
Combinatorial Optimization	5	True	Helsinki	False	Combinatorial Optimization	2 days 02:59:31	5 days 05:00:00	0.407936	50 hours 59 minutes
Data Analysis with Python	5	True	Helsinki	True	Data Analysis with Python	0 days 19:16:09	5 days 05:00:00	0.154153	19 hours 16 minutes
Design and Analysis of Algorithms	5	True	Helsinki	False	Design and Analysis of Algorithms	1 days 18:54:37	5 days 05:00:00	0.343282	42 hours 54 minutes
Elektrische Antriebe	3	True	Reutlingen	False	Elektrische Antriebe	1 days 06:03:16	3 days 03:00:00	0.400726	30 hours 3 minutes
Elektrische Antriebe Praktikum	2	True	Reutlingen	False	Elektrische Antriebe Praktikum	0 days 21:23:57	2 days 02:00:00	0.427983	21 hours 23 minutes
Elements of AI	2	True	Helsinki	True	Elements of AI	0 days 06:00:31	2 days 02:00:00	0.120172	6 hours 0 minutes
Finnish 1A	3	True	Helsinki	False	Finnish 1A	0 days 21:27:00	3 days 03:00:00	0.286000	21 hours 27 minutes
Finnish 1B	2	True	Helsinki	False	Finnish 1B	0 days 22:50:17	2 days 02:00:00	0.456761	22 hours 50 minutes
Finnish 2A	3	True	Helsinki	False	Finnish 2A	1 days 02:15:33	3 days 03:00:00	0.350122	26 hours 15 minutes
Introduction to Data Science	5	True	Helsinki	False	Introduction to Data Science	3 days 06:03:35	5 days 05:00:00	0.624478	78 hours 3 minutes
Kommunikationssysteme	3	True	Reutlingen	False	Kommunikationssysteme	1 days 02:23:43	3 days 03:00:00	0.351937	26 hours 23 minutes
Kommunikationssysteme Praktikum	1	True	Reutlingen	False	Kommunikationssysteme Praktikum	0 days 10:40:34	1 days 01:00:00	0.427044	10 hours 40 minutes
Kreativer Systementwurf	3	True	Reutlingen	False	Kreativer Systementwurf	1 days 19:02:09	3 days 03:00:00	0.573811	43 hours 2 minutes
Leistungselektronik	3	True	Reutlingen	False	Leistungselektronik	1 days 03:37:16	3 days 03:00:00	0.368281	27 hours 37 minutes
Network Analysis	5	True	Helsinki	False	Network Analysis	1 days 15:12:49	5 days 05:00:00	0.313709	39 hours 12 minutes
Recht	2	True	Reutlingen	False	Recht	0 days 19:42:00	2 days 02:00:00	0.394000	19 hours 42 minutes
Regelungstechnik II	3	True	Reutlingen	False	Regelungstechnik II	0 days 20:48:53	3 days 03:00:00	0.277530	20 hours 48 minutes
Regelungstechnik II Praktikum	3	True	Reutlingen	False	Regelungstechnik II Praktikum	0 days 19:45:18	3 days 03:00:00	0.263400	19 hours 45 minutes
Robotersysteme	5	True	Reutlingen	False	Robotersysteme	2 days 10:30:34	5 days 05:00:00	0.468076	58 hours 30 minutes
Robotersysteme Praktikum	3	True	Reutlingen	False	Robotersysteme Praktikum	0 days 22:06:53	3 days 03:00:00	0.294863	22 hours 6 minutes
Statistical Machine Learning	9	True	Tübingen	False	Statistical Machine Learning	5 days 09:41:23	9 days 09:00:00	0.576399	129 hours 41 minutes
Theoretische Informatik	9	True	Tübingen	False	Theoretische Informatik	4 days 02:52:57	9 days 09:00:00	0.439478	98 hours 52 minutes

Numerical Analysis

Mean of Time fraction

lectures['Time fraction'].mean()

0.386641850145076

lectures.groupby('University')['Time fraction'].mean()

University
Helsinki      0.340383
Reutlingen    0.399922
Tübingen      0.481007
Name: Time fraction, dtype: float64

lectures.groupby(['University', 'online course'])['Time fraction'].mean()

University  online course
Helsinki    False            0.422643
            True             0.121021
Reutlingen  False            0.399922
Tübingen    False            0.481007
Name: Time fraction, dtype: float64

Graphical Analysis

Time fraction Box Plot

sns.set_theme(style='ticks', palette='pastel')
# Draw a nested boxplot to show bills by day and time
sns.boxplot(x='University', y='Time fraction',
            data=lectures)
sns.despine(offset=10, trim=True)
plt.show()

Time fraction Strip Plot

sns.set_theme(style='whitegrid')
# Initialize the figure
f, ax = plt.subplots()
sns.despine(bottom=True, left=True)

# Show each observation with a scatterplot
sns.stripplot(x='University', y='Time fraction',
              data=lectures, dodge=True, alpha=.4, zorder=1)

# Show the conditional means
sns.pointplot(x='University', y='Time fraction',
              data=lectures, dodge=.532, join=False, palette="dark",
              markers='d', ci=None)

plt.show()

Interactive Plots

Time fraction Plot

fig = px.box(lectures, x='University', color='University', y='Time fraction', points='all', hover_name='Lecture', hover_data=['ECTS', 'Time spent'],
    title='Time fraction distribution per University')
fig.update(layout_showlegend=False)
fig.show(renderer='notebook')

Time fraction Plot Online Course

fig = px.box(lectures, x='University', color='online course', y='Time fraction', points='all', hover_name='Lecture', hover_data=['ECTS', 'Time spent'],
    title='Time fraction distribution per University')
fig.show(renderer='notebook')