Work Log: FEATHERS-disAggregator-MATSim Tool Chain Analysis

Date : 9-Oct-2023
Activity Description : Watching Python tutorials on YouTube to learn programming
concepts and syntax
Objective : Gain proficiency in Python programming language for use in
transportation research and analysis
Steps Taken : 1. Watched tutorial video on basic Python syntax and
variables
2. Practiced coding along with the tutorial examples
3. Reviewed supplementary materials and documentation
on Python programming concepts
Result : Increased understanding of fundamental Python concepts and
syntax
Analysis/ : Recognized the importance of hands-on practice in reinforcing
Interpretation learning
Successes : Successfully completed coding exercise and understood the
concepts covered in the tutorial videos
Challenges : Found some topics, like object-oriented programming,
challenging to grasp initially
Lessons Learned : Consistent practice and repetition are crucial for mastering
programming concepts. Taking breaks and revisiting
challenging topics can improve understanding over time
Next Steps : - Continue watching the Python tutorial series to cover
more advanced topics
- Start working on small coding projects to apply learned
concepts in practice

Date : 17-Oct-2023
Activity Description : Meeting: Introduction session with supervisor to discuss the
concepts of the toolchain and the three patches related to
respective topics
Agenda : 1. Explanation of the toolchain concepts
2. Overview of the three research paths and their relevance
to respective topics
3. Clarification of any questions or concerns
Meeting Summary : - The supervisor provided an overview of the FEATHERS-
disAggregator-MATSim toolchain, emphasizing its role in
transportation planning and simulation.
- Each component of the tool chain was explained,
including FEATHERS for predictive modeling,
disAggregator for schedule disaggregation, and MATSim
for network-level adjustments.
- The supervisor outlined three research paths related to
respective topics:
1. Study of the 'schedule morphing' phenomenon:
Understanding the implications and dynamics of
schedule morphing within the tool chain.
2. Evaluation of research conclusions: Proposing
methods to evaluate the effect of schedule
morphing on research conclusions, depending on
the research questions
3. Identification of conflicting assumptions:
 Identifying assumptions in the three tools that
may conflict and proposing methods to resolve
conflicts
- The supervisor encouraged active participation and
collaboration among research team members and
emphasized the importance of regular communication
and progress tracking
Action Items : - Review the provided materials on the toolchain and
research paths
- Identify specific areas of interest within the research
paths for further exploration
Next Meeting : -

Date : 22-Oct-2023
Activity Description : Engaging with tutorials and interactive exercises on the
learnpython.org website to develop Python programming skills
Objective : Enhance proficiency in Python programming language for
application in transportation research and analysis
Steps Taken : 1. Completed interactive tutorials on basic Python syntax
and data structures.
2. Practiced coding exercises to reinforce learning and test
understanding.
Result : Improved understanding of Python programming concepts and
syntax through interactive learning.
Analysis/ : Found the interactive exercises effective for reinforcing
Interpretation theoretical concepts and gaining practical experience.
Successes : Successfully completed all exercises in the basic Python
tutorials and gained confidence in using Python syntax.
Challenges : Encountered difficulties in understanding certain advanced
concepts
Lessons Learned : Consistent practice and experimentation are essential for
mastering complex programming concepts. Seeking clarification
through additional resources or forums can help overcome
challenges.
Next Steps : - Continue exploring advanced Python topics
- Practice solving more complex coding challenges to
further solidify understanding
- Consider supplementing learning with additional
resources or tutorials on challenging topics

Date : 23-Oct-2023
Activity Description : Completing the warm-up tasks provided in the “warmUp.pdf”
attachment to become familiar with the data and Python3
Objective : Gain proficiency in using Python3 for data analysis and
manipulation, and become acquainted with the dataset to be
used in the research project
Steps Taken : 1. Downloaded and reviewed the warm-up tasks document
(warmUp.pdf) provided by the supervisor.
2. Accessed the schedules_baseYear_996.csv file attached
to the email.
3. Reviewed the description of the dataset and its variables
in the warm-up document to understand its structure
and contents.
 4. Installed Python 3 on the local development
environment if not already installed.
5. Loaded the schedules_baseYear_996.csv dataset into
Python
6. Completed warm-up tasks outlined in the warmUp.pdf
document
Result : Total Distance: 6770.0 km
Mode Distances:
Mode 2: 592.0 km
Mode -2: 0.0 km
Mode 5: 1386.0 km
Mode 4: 2654.0 km
Mode 7: 1728.0 km
Mode 1: 251.0 km
Mode 3: 159.0 km
Number of people making at least one trip but not using the car:
775
Average daily travel time per person: 54.07364341085271
minutes
Analysis/ :  The dataset captures a significant total distance
Interpretation traveled, indicating active mobility patterns within the
population.
 Mode 4 (presumably private vehicles) appears to be the
most dominant mode of transportation, followed by
Mode 7 (possibly public transit).
 The presence of a considerable number of individuals
not using the car suggests diversity in transportation
choices or accessibility to alternative modes.
 The average daily travel time per person provides
insight into the overall mobility demand and potential
congestion issues.
Successes :  Successfully loaded the dataset into Python
environment.
 Completed all warm-up tasks as outlined in the provided
document.
 Obtained insights into the structure and characteristics
of the transportation dataset.
Challenges : It’s my first time doing Python with real data, it takes a lot time
for me to do the code
Lessons Learned :  Regular practice and hands-on experience are essential
for developing proficiency in data analysis with Python.
 Effective documentation and organization of code
facilitate understanding and reproducibility of analysis
tasks.

Date : 29-Nov-2023
Activity Description : Meeting: Engaging in a discussion with colleagues or the
supervisor to review and discuss solutions to the warm-up
exercises completed earlier.
Objective : Collaborate with peers or supervisor to gain insights, clarify
doubts, and refine solutions to the warm-up exercises.
Steps Taken : 1. Gathered with colleagues or supervisor at the scheduled
time for the discussion session.
 2. Presented solutions to the warm-up exercises, including
calculations, code snippets, and interpretations.
3. Actively listened to feedback and insights shared by
peers or supervisor regarding the solutions presented.
4. Engaged in discussions to address any discrepancies,
clarify doubts, or explore alternative approaches to
solving the exercises.
5. Took notes during the discussion to capture key points,
suggestions, and areas for improvement.
Discussion Highlights : Presented findings and calculations for key metrics such as total
distance traveled, mode distances, and average daily travel time
per person.

Date : 13-Dec-2023
Activity Description : Meeting with supervisor to discuss various topics related to the
research project.
Agenda : - The supervisor reminded us about the upcoming agenda
item (4a. Schedule morphing)
- Declaration on Honour
- Completed Warm-up tasks and discuss any remaining
problems or difficulties encountered during completion
- Review decisions taken and progress made and reflect
on research questions
- Discuss the availability of project resources like code
and data
- Explain work log
Next Steps : - Prepare to explain the first steps of the work plan during
the next session

Date : 8-Jan-2024
Activity Description : Meeting with supervisor
Agenda : - Introduction to mEpiLister Scheme
- Preparation Document for the upcoming task
Action Items : - Review the preparation document in detail to familiarize
with the task objectives and requirements
- Prepare to initiate the tasks according the agreed
timeline

Date : 12-Feb-2024
Activity Description : Meeting with supervisor to discuss expected outputs and their
formats for the research project
Notes : - Research Plan (Expected Output 1)
- Detailed Work Log (Expected Output 2)
- List of Research Questions (Expected Output 3)
- Concise Detailed Report (Expected Output 4)
Action Items : - Review provided examples and guidelines for the
research plan and work log
- Formulate initial research questions and discuss them
with the supervisor for feedback and refinement

Date : 12-Feb-2024
Activity Description : mEpiLister Execution Results
Result : ---------------------------------------------------------
2024-02-12T08:34:38.686552 prsTotalCount=[730000]
personPrediction type=[<class 'person.Person_M'>]
personId=[p16932271] nEpisodes=5 age=[72]
personPrediction type=[<class 'person.Person_M'>]
personId=[p16932271] nEpisodes=5 age=[72]
personPrediction type=[<class 'person.Person_F'>]
personId=[16932271] nEpisodes=5 age=[4]
Frequency table for number of different/unique modes in a
schedule
nModesInSchedule=1 frequency=628787
nModesInSchedule=2 frequency=93210
nModesInSchedule=3 frequency=12795
nModesInSchedule=4 frequency=633
Total #schedules=735425
#Persons=735425 #Episodes=3092804
avg(episodes/person)=4.21
Analysis : - The execution of mEpiLister provided detailed statistics
on the population's travel behavior and characteristics.
- The majority of schedules contain 1 mode, indicating a
common pattern in transportation mode choice.
- The average number of episodes per person is 4.21,
suggesting frequent travel activities within the
population.

Date : 4-Mar-2024
Activity Description : Disaggregation Hypothesis Testing
Objective : To test the hypothesis formulated during the previous session
that disAggregation leads to lower total travel duration, with
consideration for the size of TAZs.
Step Taken : 1. Data Preparation
- Loaded the file "utn2zones.csv" containing TAZ
information.
- Extracted TAZ identifiers from the "subzone0" field and
regions from the "gebied10" field.
- Retained only TAZs for which the region belongs to
GEBIEDEN-1-5 (regions 1 to 5).
2. Analysis
- Ran the code to extract TAZ identifies and regions
- Reviewed the output, noting the regions and subzones
retained based on the specified criteria
- Observed the distribution of TAZs among regions 1 to 5
to ensure proper selection
Result : Step A:
“…………….
region=0 subZone=7776
region=0 subZone=7777
region=0 subZone=7778
region=0 subZone=7779
region=0 subZone=7780
region=0 subZone=7781
region=0 subZone=7782
region=0 subZone=7783
 region=0 subZone=7784
region=0 subZone=7785”
Analysis of Results :

Date : 30-Apr-2024
Activity Description : Analysis of Disaggregation Effect using mEpiLister.py
Objective : To observe the morphing effect of the disAggregator by using
mEpiLister.py with MATSim initial plans (file 1.plan.csv) and
FEATHERS schedules sorted.csv
Result : Initial Plans Data:
person_id gender age ... distance start_link end_link
0 p00000004 2 59 ... -2.000000 -2 -2
1 p00000004 2 59 ... 4193.415375 60176 50064
2 p00000004 2 59 ... 4193.415375 50064 60176
3 p00000037 1 78 ... -2.000000 -2 -2
4 p00000037 1 78 ... 7248.667843 60176 22808

[5 rows x 15 columns]

FEATHERS Schedules Data:

person_id activity_type ... new_mode_6 new_mode_7
0 1 1 ... 1 1
1 1 2 ... 1 1
2 1 1 ... 1 1
3 2 1 ... 0 0
4 2 2 ... 0 0

[5 rows x 35 columns]
Analysis/ :  The initial plans data provide details about individuals'
Interpretation activities and distances traveled.
 The FEATHERS schedules data contain information
about individuals' activity types and new modes
assigned.
 By comparing the initial plans and FEATHERS schedules,
the morphing effect of the disAggregator can be
observed, indicating changes in activity types and
modes.
Next Steps : - Analyze the differences between initial plans and
FEATHERS schedules to understand the impact of the
disAggregator on travel behavior.
- Explore additional metrics or visualizations to quantify
the morphing effect and its implications for
transportation planning.
- Consider conducting statistical analysis to assess the
significance of the observed changes.

Date : 30-Apr-2024
Activity Description : Analysis of Total Travel Duration
Objective : To determine the appropriate metrics for checking travel
duration and compare them between initial plans and
FEATHERS schedules.
Result : Total Travel Duration (Initial Plans):
person_id
 p00000004 -200:13:2900:06:40
p00000037 -200:06:4200:09:06
p00000040 -
200:20:4700:09:1200:13:4000:26:2400:16:4300:3...
p00000071 -200:18:0600:18:0600:13:5300:14:44
p00000084 -200:00:0000:18:1500:18:42
...
p17076988 -200:41:5100:42:54
p17076993 -200:22:5100:24:16
p17077035 -200:01:1200:00:49
p17077055 -200:00:0000:00:00
p17077057 -200:00:0300:00:07
Name: trav_time, Length: 735425, dtype: object

Total Travel Duration (FEATHERS Schedules):

person_id
1 21
2 63
3 90
4 49
5 19
...
17077067 126
17077068 245
17077069 126
17077070 126
17077071 126
Name: trip_duration, Length: 17077071, dtype: int64
Analysis/ :  The total travel duration values in the initial plans are in
Interpretation a different format compared to those in the FEATHERS
schedules.
 The negative values in the initial plans may indicate
incomplete or invalid data, requiring further
investigation and processing.
 Comparison of total travel duration between the two
datasets may provide insights into the effectiveness of
the disAggregator in morphing travel behavior.

Date : 30-Apr-2024
Activity Description : Analysis of Activity Types and Deviations
Objective : To analyze activity types and deviations between predicted and
typical activity durations in the MATSim plans
Result : The sample result:
Activity Type: home_1050, Deviation: -1052.0
Activity Type: home_0915, Deviation: -917.0
Activity Type: home_0390, Deviation: -392.0
Activity Type: home_0405, Deviation: -407.0
Activity Type: home_0195, Deviation: -197.0
Activity Type: home_1305, Deviation: -1307.0
Activity Type: home_0015, Deviation: -17.0
Activity Type: home_1155, Deviation: -1157.0
Activity Type: home_1410, Deviation: -1412.0
Activity Type: home_0675, Deviation: -677.0
Analysis/ :  The activity types follow a specific naming convention
Interpretation indicating the type of activity and its typical duration in
minutes.
 Deviations represent the difference between predicted
and typical durations, with negative values indicating
shorter durations and positive values indicating longer
durations.
 The purpose of generating activity types with typical
durations is to ensure consistency and reduce morphing
by MATSim.
Next Steps : -

Date : 30-Apr-2024
Activity Description : Analysis of Activity Types and Deviations
Objective : To analyze activity types and deviations between predicted and
typical activity durations in the MATSim plans
Result : The sample result:
Activity Type: home_1050, Deviation: -1052.0
Activity Type: home_0915, Deviation: -917.0
Activity Type: home_0390, Deviation: -392.0
Activity Type: home_0405, Deviation: -407.0
Activity Type: home_0195, Deviation: -197.0
Activity Type: home_1305, Deviation: -1307.0
Activity Type: home_0015, Deviation: -17.0
Activity Type: home_1155, Deviation: -1157.0
Activity Type: home_1410, Deviation: -1412.0
Activity Type: home_0675, Deviation: -677.0
Analysis/ :  The activity types follow a specific naming convention
Interpretation indicating the type of activity and its typical duration in
minutes.
 Deviations represent the difference between predicted
and typical durations, with negative values indicating
shorter durations and positive values indicating longer
durations.
 The purpose of generating activity types with typical
durations is to ensure consistency and reduce morphing
by MATSim.

Date : 6-May-2024
Activity Description : Analysis of Activity Types Deviations
Objective : To further analyze activity types and deviations including mean,
median, and standard deviation statistics
Result : The sample result:
Activity Type: home_1050, Mean Deviation: 0 days
00:13:28.931501366, Median Deviation: -1 days
+23:53:26.999999999, Standard Deviation: 0 days
01:33:55.054594114
Activity Type: work_0345, Mean Deviation: 0 days
00:36:16.330373764, Median Deviation: 0 days 00:09:16,
Standard Deviation: 0 days 01:58:24.611338678
Activity Type: home_0915, Mean Deviation: 0 days
00:16:38.395909565, Median Deviation: -1 days
+23:53:26.999999999, Standard Deviation: 0 days
 01:40:05.428695118
Analysis/ : - Mean deviation represents the average difference
Interpretation between predicted and typical durations for each
activity type
- Median deviation provides the middle value of the
deviations, indicating the central tendency of the data
- Standard deviation measures the dispersion or
variability of the deviations around the mean

Date : 6-May-2024
Activity Description : Code Review, Data Visualization, Analysis RQs, Summary
Findings
Result : Graphs and Report