Anubha-Mahajan

[amahajan68]

Week 1

1. Introduction to surrogate modeling
   • Operational Energy / Embodied Energy
2. Modeling Atlanta metro area - take upto half a year

Week 2

1. Set up the sub teams - Energy in Buildings (& emissions)
2. Geometry we will be working on - a script which only needs map co-ordinates
   • Have building data for English Avenue and East Midtown

Week 3

1. Literature Review - Surrogate modeling for sustainable building design – A review
2. Articulated the problem definition
3. Found a research gap

Week 4

1. Meeting with the meeting to discuss the specifics of the project (1/30)

   • Only in Atlanta vs other places
   • Whole city vs certain buildings - Urban building designs through the entire city
     • Creating a general surrogate model
     • Most common inputs/outputs to consider for our model
       • Inputs - Building geometry, internal loads, controls, weather, material of the walls, window to wall ratio, HVAC
       • Outputs - Total energy, electricity, gas, heating/cooling load, thermal comfort
   • Compartmentalize the city into residential, office buildings (commercial) for ease of model
   • Possibility of using an automated surrogate derivation model to model the different types of buildings

2. Meeting with Dr. Kastner about the focus of the project (2/1)

   • Depth of the model
     • Are we building from scratch?
     • Use baseline model on GitHub, make it better by:
       • Finding other data sources to integrate
       • Improving accuracy
     • Likely be trained for another climate zone (probably retrain, if not a comparable climate)
     • Use Python, with easily found inputs
     • Note: Hourly accuracy is challenging; aim for Mean Absolute Error around 10% for monthly evaluation, using RMSE as the metric
   • Are we looking at a model that shows us energy output based on inputs of projected building time?
     • Predictions to be for the present time
     • Use weather boundary condition later for the future for climate change considerations
   • Data scope
     • Start with residential, using the same variables (5-6) for all the buildings
       • Publicly available data for Atlanta from the Fulton County Board of Assessors
       • Dataset of smart heat and water meter data with accompanying building characteristics
     • Commercial buildings are difficult to predict
     • Outputs should be hourly
   • Focus area
     • We don't want to focus on buildings off the grid
     • Aim for as little data requirement as possible
   • Model that we want to base ours off - Urbano

Week 5

1. Chose papers A global model of hourly space heating and cooling demand at multiple spatial scales and High-Resolution Hourly Surrogate Modeling Framework for Physics-Based LargeScale Building Stock Modeling
   • Do a deeper dive to get an understanding of the GitHub models available
2. Emailed Prof. Marszal-Pomianowska to get access to the Aalborg Dataset

Week 6

1. Introduction and Demo of Grasshopper and Urbano by Silvia
2. Meeting with Dr. Kastner to get an understanding of the source code for BuildStock Batch and new directions for the project (2/15)
3. Set up and ran Optimization of building energy consumption
   • Multivariate Regression using design parameters as inputs and heating and cooling load as outputs
4. Got the build environment set-up for BuildStock Batch
   • Stuck on configuring AWS CLI and Athena Engine Version
5. Chose Rank 1 Solution for predicting energy consumption for commercial buildings
   • Stuck on configuring AWS EC2 and launching Deep Learning AMI
6. Found Atlanta Energy Benchmarking Map to compare our results to

[amahajan68]

Weekly Notebook Entry — Week 7

Overview

• Week Span: 2/19/24 to 2/25/24

Tasks for This Week

• [ ] Task 1: Dive into various data models with a focus on mastering AWS for our project's needs and exploring Rank 1 along with other model options for urban heat analysis.
• [ ] Task 2: Get hands-on with implementing and fine-tuning the BuildStock Batch and the highly acclaimed Rank 1 Solution from the ASHRAE competition.

Tasks

Task: Model and Data Collection - (ongoing)

• Current Status: I've been exploring how Kaggle models like Rank 1 work and trying to gather data from places like the Fulton County Board of Accessors and the insightful Building Data Genome Project 2.
• Expected Completion: I'm hoping to wrap up data collection soon so we can move on to the juicy part—analysis!

Task: Configure AWS - (completed)

• Description: Set up and fine-tuned AWS to support our sophisticated data models.
• Link to Work: AWS Setup Instructions
• Outcome & Reflection: AWS is up and running smoothly now, which is a huge relief and a big step forward for us.

Task: Test BuildStock and Rank 1 Models - (ongoing)

• Current Status: Currently, I'm testing these models to ensure they work as expected. It's quite the challenge but very exciting!
• Expected Completion: I'm aiming to have everything operating perfectly by next week.

Task: Operationalize Alternative Models - (In progress)

• Description: After some hiccups with the initial models, I'm now looking into alternatives to see how they might better meet our needs.
• Link to Work: Alternative Model Insights
• Outcome & Reflection: These alternatives are shedding some light on different ways we might approach our project, which is incredibly helpful.

Challenges & Solutions

• Navigating Technical Hurdles: Setting up AWS was a beast, but with some patience and a lot of Googling, I managed to get it done. Now, it's about making sure everything else falls into place.
• Data Wrangling: Trying to get all this data organized and usable is no small task. I’m learning a ton about data management out of necessity!

General Takeaway, Skills & Knowledge Acquired

• This week was a crash course in cloud computing and data analysis. I've learned more about AWS than I ever thought I would, and getting to grips with these complex models has been a real brain workout.

Team Meetings

• Meeting Highlights: We had some great discussions about our progress and how to tackle the challenges we’re facing. Getting access to new data sources was definitely the highlight of our week.
• Peer Feedback: The feedback was really encouraging—everyone seems impressed with how we're handling the AWS setup and are keen to see how the new models perform.

Personal Reflection

• Balancing the technical aspects with the overall project goals has been tough but rewarding. I’m really starting to see how all these pieces fit together, and I’m excited about the progress we’re making.

Additional Notes & Resources (optional)

• AWS Documentation - This has been my bible this week! A must-read for anyone diving into AWS.

---

Plans for Next Week

• [ ] Task 1: Put the final touches on our data setup and begin extensive testing of the BuildStock and Rank 1 models.
• [ ] Task 2: Keep evaluating alternative models based on our tests to ensure we're using the best tools for the job.

[amahajan68]

Weekly Notebook Entry — Week 8

Overview

• Week Span: 2/26/24 to 3/3/24

Tasks for This Week

• [ ] Task 1: Initiate the Rank 1 model on AWS with the goal to use it for analyzing heat usage data in Atlanta and also delve into finding the real data sources used by demand.ninja.
• [ ] Task 2: If challenges arise with Rank 1 or data availability, explore alternative data sources and possibly new models.

Tasks

Task: Run Rank 1 Solution on AWS - (ongoing)

• Current Status: I've started running the Rank 1 solution from the ASHRAE competition on AWS. This is my first real attempt at handling such a complex system on my own.
• Expected Completion: I'm aiming to have it up and running smoothly by the end of this week.

Task: Find demand.ninja datasets - (completed)

• Description: I looked into the data sources mentioned by the creators of renewables.ninja to understand how they manage their models.
• Link to Work: Renewables Ninja About Page
• Outcome & Reflection: It was a bit disappointing not to find the actual datasets I was hoping for; instead, I found some complex mathematical descriptions and processes which were quite challenging to decode.

Task: Search for Alternative Data Sources - (ongoing)

• Description: Exploring other potential data points like the local tax assessor's website and some resources Dr. Kastner shared. These don't directly provide energy usage data, but might help indirectly.
• Expected Completion: Continuing this search into next week.

Challenges & Solutions

• Technical Learning Curve: Tackling AWS and the intricacies of a high-ranking Kaggle model have been daunting but enlightening. I've had to upskill quickly, making use of numerous tutorials and forums.
• Data Sourcing: It's frustrating when data isn't readily available or is hidden behind complex technical processes. I'm keeping optimistic and persistent, exploring every possible avenue.

General Takeaway, Skills & Knowledge Acquired

• This week has been a crash course in AWS and the world of competitive data science through Kaggle. Despite the hurdles, I feel I'm making real progress in understanding these advanced tools and concepts.

Team Meetings

• Meeting Highlights: We didn't catch up with Dr. Kastner this week, but our team meetings were crucial in pooling our insights and strategies to tackle the model and data issues.
• Peer Feedback: Everyone is equally invested and facing similar challenges; the support has been incredible.

Personal Reflection

• Diving into AWS and data models head-first has been a challenge, but it's gratifying to see the pieces coming together. The more I learn, the more I realize how much there is to know—every little success feels like a big win.

Visual Documentation (optional)

• Week 8 Progress Slides

Additional Notes & Resources (optional)

---

Plans for Next Week

• [ ] Task 1: Continue refining the AWS setup and get the Rank 1 model fully functional.
• [ ] Task 2: Expand the hunt for usable data sources, perhaps even reaching out to building managers in Atlanta for firsthand data.

[amahajan68]

Weekly Notebook Entry — Week 9

Overview

• Week Span: 3/4/24 to 3/11/24

Tasks for This Week

• [ ] Task 1: Continue to run the Rank 1 solution on AWS, shifting focus to using GaTech’s PACE-ICE system due to resource limitations on AWS.
• [ ] Task 2: Delve into the BAIT article shared by Dr. Kastner to understand its implications for our Demand Ninja model project.
• [ ] Task 3: Implement Dr. Kastner's suggestion to cache results in a database and use a VPN to avoid having the IP blocked while gathering data from Fulton County's website for the 1000 buildings project.

Tasks

Task: Run Rank 1 Solution in AWS - (ongoing)

• Current Status: Set up AWS, encountered resource limitations especially with DL-AMI and necessary Python libraries. Transitioning to PACE-ICE for further implementation.
• Expected Completion: Planning to have it operational by end of next week using PACE.
• Link to Work: Rank 1 Solution Guide

Task: BAIT Article Review - (completed)

• Description: Reviewed how Building Adjusted Internal Temperature (BAIT) could affect our Demand Ninja model's accuracy by accounting for external climate effects.
• Link to Work: BAIT Nature Article
• Outcome & Reflection: Realized the need for extensive datasets on retrofitted buildings to enhance accuracy; such data is rarely public, raising concerns about our model’s reliability.

Task: Implement Caching and VPN for Data Collection - (ongoing)

• Description: Following Dr. Kastner's advice to optimize our data collection process for the Fulton County project by caching results and using a VPN.
• Current Status: Setting up the necessary infrastructure to ensure uninterrupted data access and storage.
• Expected Completion: End of the week.

Challenges & Solutions

• AWS Limitations: Struggled with AWS due to high costs and inadequate free resources which hampered our progress. Decided to switch to using PACE-ICE, which should provide the necessary capabilities without the financial overhead.
• Model Understanding and Data Access: Attempting to grasp complex statistical models and navigate data access issues has been challenging but enlightening. Implementing caching and VPN strategies as per Dr. Kastner's suggestion to mitigate data access issues.

General Takeaway, Skills & Knowledge Acquired

• Gained a better understanding of cloud computing platforms and statistical modeling. Learned crucial lessons in resource management and model accuracy considerations.

Team Meetings

• Meeting Highlights: Broad discussions on data sourcing from Assessors website and strategies to implement models efficiently using alternative resources.
• Peer Feedback: The team is pushing for a working model soon; feedback has been constructive and focused on overcoming technical barriers.

Personal Reflection

• This week was a rollercoaster with AWS frustrations leading us to seek alternative solutions. Also, diving into the complexities of the BAIT article was particularly challenging for a non-CS major like me, but it’s been a rewarding learning curve.

Visual Documentation (optional)

• Weekly Progress and Insights

Additional Notes & Resources (optional)

---

Plans for Next Week

• [ ] Task 1: Transition fully to PACE-ICE to run the Rank 1 solution.
• [ ] Task 2: Investigate Kaggle datasets to gather more insights into how we can refine the Demand Ninja model.

[amahajan68]

Weekly Notebook Entry — Week 10

Overview

• Week Span: 3/12/24 to 3/19/24

Tasks for This Week

• [ ] Task 1: Continue running the Rank 1 solution using GaTech's PACE and explore switching to Rank 5 due to efficiency concerns.
• [ ] Task 2: Secure and review the Kaggle dataset necessary for the Demand Ninja model implementation.

Tasks

Task: Run Rank 1 Solution Using PACE Cluster - (ongoing)

• Current Status: Utilized the PACE Cluster to operate the Rank 1 Kaggle solution. Encountered issues with long training times and limited session durations.
• Expected Completion: Reviewing alternatives for more efficient model execution.
• Link to Work: Instructional Cluster Environment - ICE

Task: Explore Rank 5 as an Alternative - (planned)

• Description: Given the operational challenges with Rank 1, considering pivoting to Rank 5 which promises similar outcomes with significantly less computational time.
• Outcome & Reflection: Decision pending based on team consensus and further testing.

Task: Data Acquisition for Demand Ninja - (completed)

• Description: Found and vetted potential datasets from the Building Data Genome Project 2 for the Demand Ninja model.
• Link to Work: Kaggle Data for Building Genome Project 2
• Outcome & Reflection: Identified useful datasets though understanding the data specifics remains challenging.

Challenges & Solutions

• PACE Limitations: The constraints of using PACE for high-demand models led us to consider more resource-efficient alternatives.
• Data Complexity: Navigating the complexities of the Kaggle datasets to ensure they are fit for our modeling needs.

General Takeaway, Skills & Knowledge Acquired

• Enhanced understanding of high-performance computing with PACE and initial explorations into Kaggle’s API and datasets. Learned to adapt project strategies based on technical and resource constraints.

Team Meetings on 3/14/24

• Meeting Highlights: Debated the feasibility of continuing with Rank 1 versus switching to Rank 5. Dr. Kastner emphasized efficiency and building on existing community work.
• Peer Feedback: There's a strong inclination within the team to optimize our approach to model running, with a focus on achieving results more efficiently.

Personal Reflection

• The complexities of running these high-level models on PACE have been a significant learning curve. Exploring Rank 5 as a faster alternative has opened up new avenues for our project.

Visual Documentation (optional)

• 

Additional Notes & Resources (optional)

---

Plans for Next Week

• [ ] Task 1: Implement and test the Rank 5 solution on PACE.
• [ ] Task 2: No work planned for Spring Break.

[amahajan68]

Weekly Notebook Entry — Week 11 - Spring Break

Overview

• Week Span: 3/20/24 to 3/27/24

Tasks for This Week

• [ ] Task 1: Utilize Spring Break to set up and initiate Rank 5 on GaTech PACE for faster processing.

Tasks

Task: Transition to Rank 5 - (completed)

• Description: Opted for Rank 5 instead of Rank 1 to expedite our modeling process.
• Outcome & Reflection: Switching to Rank 5 was a strategic move to save time while maintaining model effectiveness.

Task: Implement Rank 5 on PACE - (ongoing)

• Current Status: Setup began over Spring Break.
• Expected Completion: Aiming to have it fully functional by the end of next week.

General Takeaway, Skills & Knowledge Acquired

• Learned to optimize resources under time constraints.

Personal Reflection

• Pushing forward in our project underscored the importance of flexibility and efficiency in research.

Plans for Next Week

• [ ] Task 1: Finalize the setup of Rank 5 and begin testing its performance.

[amahajan68]

Weekly Notebook Entry — Week 12

Overview

• Week Span: 3/28/24 to 4/03/24

Tasks for This Week

• [ ] Task 1: Execute the Rank 5 Solution using PACE, aiming for a more efficient setup than previous models.
• [ ] Task 2: Troubleshoot and perfect the implementation of the web scraper for data collection, focusing on resolving input issues in the search bar.
• [ ] Task 3: Explore additional datasets and research papers for potential integration into our project models.

Tasks

Task: Rank 5 via PACE - (ongoing)

• Current Status: Running the Rank 5 solution on PACE which promises a quicker processing time. Planning to integrate the Atlanta dataset next to generate a functional heat map.
• Expected Completion: By the end of this week.

Task: Web Scraper Debugging - (ongoing)

• Current Status: Struggling with address input in the web scraper’s search bar; technical glitches are preventing data collection.
• Expected Completion: Ongoing debugging expected to resolve soon.
• Special Note: Dr. Kastner suggested using the inspect tool for debugging this issue, which I'm currently trying to use to fix the errors.

Task: Data Exploration - (ongoing)

• Description: Identified and began evaluating new data sources and research models that could enhance our understanding and implementation of energy management systems.
• Link to Work: Three-Year Building Dataset, Energy Data Imputation Models Paper
• Current Insights: The new dataset offers extensive support for research on building energy management, while the reviewed paper presents advanced but complex imputation models that require more features than we currently have access to.

Challenges & Solutions

• Kaggle API and PACE Setup Issues: Faced technical problems with the Kaggle API and PACE setup potentially affecting the Rank 5 execution.
• Web Scraper Input Errors: Encountering difficulties in entering data into the web scraper, impacting our data collection process.
• Action Taken: Utilizing the inspect tool as per Dr. Kastner's advice to identify and correct the underlying issues with the web scraper.

General Takeaway, Skills & Knowledge Acquired

• Learned to quickly pivot between different technologies and troubleshooting methods to enhance project efficiency and overcome technical barriers.

Team Meetings on Thursday this week

• Meeting Highlights: Updated Dr. Kastner on the Assessors website data collection efforts. Also discussed issues related to the Demand Ninja and Rank 5 setups.
• Peer Feedback: The team is eager to see a functioning model before the semester concludes, reinforcing the need to swiftly address the current technical challenges.

Personal Reflection

• This week was a test of my troubleshooting skills, especially with terminal and Python commands in different setups. These challenges, though frustrating, are molding me into a more proficient problem-solver and enhancing my technical versatility.

Visual Documentation

• Week 12 Project Slides

Plans for Next Week

• [ ] Task 1: Resolve all existing issues with Rank 5 on PACE to ensure it is fully operational.
• [ ] Task 2: Complete the debugging of the web scraper to start effective data collection without further delays.

[amahajan68]

Weekly Notebook Entry — Week 13

Overview

• Week Span: 04/04/24 to 04/11/24

Tasks for This Week

• [ ] Task 1: Focus on getting the Rank 5 solution fully operational using PACE.
• [ ] Task 2: Address issues with the web-scraper, particularly the challenges posed by Recaptcha.

Tasks

Task: Running Rank 5 with PACE - (ongoing)

• Current Status: Working to install necessary libraries for Rank 5, facing permissions issues on PACE as sudo access is restricted. Dr. Kastner recommended using the Kepler GeoJSON tool for our data visualization needs once we have the model running.
• Expected Completion: Aiming to have a working model by next week to meet our April 21st deadline for a preliminary GeoJSON presentation.

Task: Web Scraper Optimization - (ongoing)

• Description: Facing challenges with Recaptcha while running our web scraper. Silvia has suggested using an undetected Chrome driver to avoid detection. Additionally, I've implemented the use of a user agent and random timing tactics in scripts to simulate human interaction.
• Current Insights: Adjustments such as introducing random pauses (time.sleep()) within the script and using a sophisticated user agent setup are underway to better mimic human behavior and potentially bypass security measures like Recaptcha.

Challenges & Solutions

• Model Setup Challenges: Struggling with library installations and permission settings on PACE which are delaying our progress with Rank 5.
• Web Scraper Challenges: Recaptcha has been a significant barrier; utilizing a combination of undetected Chrome driver, user agents, and randomized script timing to overcome this.
• Navigational Steps: Exploring alternative installation methods without sudo rights and refining web scraper scripts to avoid detection and effectively collect data.

General Takeaway, Skills & Knowledge Acquired

• Learned the importance of persistence and creativity in troubleshooting—employing various strategies like user agents and Chrome drivers has broadened my technical skill set.

Personal Reflection

• The intensity of this week’s technical challenges has been both stressful and enlightening. Pushing through these issues is tough but rewarding, as each step forward boosts my confidence and prepares me better for future hurdles.

Plans for Next Week

• [ ] Task 1: Continue to refine and stabilize the Rank 5 setup on PACE to ensure full functionality.
• [ ] Task 2: Keep enhancing the web scraper with the latest techniques to ensure it operates smoothly and efficiently, avoiding any security flags while collecting data.

[amahajan68]

Weekly Notebook Entry — Week 14

Overview

• Week Span: 04/12/24 to 4/19/24

Tasks for This Week

• [ ] Task 1: Continue to troubleshoot and run the Rank 5 model on PACE, aiming to have a working model by semester's end.
• [ ] Task 2: Complete data extraction for both residential and commercial buildings from the Fulton County's website.

Tasks

Task: Running Rank 5 - (ongoing)

• Current Status: Consulting with Dr. Kastner about library installation issues on PACE; he advised against using sudo and recommended using pip or 'module load' for checking existing libraries. Transitioned to using files directly from Kaggle to simplify setup. Encountered issues with outdated libraries in the requirements.txt and problems with the Kaggle API which halted progress.
• Expected Completion: Hoping for resolution by next week, but contingency plans may need to be considered.

Task: Data Extraction for Building Analysis - (completed)

• Description: Extracted key features for 600 residential buildings initially. Expanded the scope to include commercial buildings and successfully compiled data for 800 addresses in total, enhancing our dataset significantly.
• Outcome & Reflection: Managed to document all findings comprehensively, which will support our model development and analysis efforts.

Challenges & Solutions

• Technical Issues with Rank 5: Multiple setbacks with installing and configuring necessary software on PACE. Utilized Anaconda to manage virtual environments more effectively after pip failed. Still struggling with library compatibility issues due to outdated dependencies.
• Data Collection: Initially only had partial data, but expanded efforts to include a broader range of building types, which required meticulous documentation and validation.

General Takeaway, Skills & Knowledge Acquired

• The importance of persistence and creative problem-solving in data science has never been clearer.

Personal Reflection

• This week has been incredibly challenging but also very instructive. The obstacles with Rank 5 have taught me a lot about the resilience needed in research. Discussing potential backup plans, like switching to Demand Ninja if necessary, in next week’s class.

Visual Documentation

• 

Plans for Next Week

• [ ] Task 1: Resolve the remaining issues with Rank 5 or begin setting up Demand Ninja as an alternative. Prepare to present our findings, regardless of the model, to jumpstart the next semester's efforts.

[amahajan68]

Weekly Notebook Entry — Week 15

Overview

• Week Span: 04/20/24 to 04/27/24

Tasks for This Week

List your specific objectives for this week, including learning targets and project milestones.

• [ ] Task 1: Debug issues within the BAIT function in the Demand model to ensure proper input handling and prevent NaN values.
• [ ] Task 2: Acquire and prepare weather data using the NREL API for the Demand model.
• [ ] Task 3: Create a GeoJSON map from the Demand model's predicted values for visual analysis.
• [ ] Task 4: Document the Demand model and the Random Forest multivariate regression model with detailed instructions.
• [ ] Task 5: Assemble the final project presentation for the project's culmination.
• [ ] Task 6: Integrate the web parser dataset with FEMA data and population estimates for comprehensive analysis.

Tasks

Task: Debug BAIT Function - (completed)

• Description: Addressed input issues in the BAIT function of the Demand model that were causing NaN outputs.
• Outcome & Reflection: By fixing the input issues, the model now properly processes data, ensuring accurate computational results.

Task: Data Preparation for Demand Model - (completed)

• Description: Used the NREL API to collect weather data for 1100 addresses and transformed EPW files into usable dataframes.
• Link to Work: NREL API, EPW DataFrames
• Outcome & Reflection: The prepared dataframes are now ready to be used as input for the Demand model, providing a solid foundation for accurate weather modeling.

Task: Create GeoJSON Map - (completed)

• Description: Transformed prediction outputs into GeoJSON format and used Kepler.gl for visualization.
• Link to Work: Kepler Map
• Outcome & Reflection: The map effectively illustrates the spatial distribution of data, enhancing our analytical capabilities.

Task: Documentation of Models - (completed)

• Description: Documented both the Demand and Random Forest models with comprehensive run instructions using Git LFS for large datasets.
• Outcome & Reflection: Both models are now well-documented, making them accessible and user-friendly for future researchers.

Task: Final Presentation - (completed)

• Description: Compiled and finalized the project presentation to summarize our findings and methodologies.
• Outcome & Reflection: The presentation is set to showcase our project effectively to peers and faculty, highlighting the comprehensive work done.

Task: Data Integration - (completed)

• Description: Combined the dataset from the web parser with the FEMA dataset and added population estimates for a holistic view.
• Outcome & Reflection: The integrated dataset now offers a richer, more detailed base for our models, increasing the robustness of our analyses.

Challenges & Solutions

• Encountered issues with the BAIT function not accepting inputs correctly, leading to NaN values in the model output. Adjustments to the function's code rectified these errors.

General Takeaway, Skills & Knowledge Acquired

• This week was intensive in terms of debugging and data preparation. I gained significant experience in handling complex datasets and enhancing model functionality through practical problem-solving.

Personal Reflection

• The tasks this week pushed me to refine my technical skills further and reinforced the importance of meticulous documentation and presentation in conveying complex information effectively.

Visual Documentation (optional)

• Final Presentation

Additional Notes & Resources (optional)

• Git LFS was essential for managing large data files efficiently on GitHub, proving invaluable for documentation.

---

Plans for Next Week

• [ ] Task 1: Review all project materials and ensure everything is ready for the final presentation.
• [ ] Task 2: Start preparing for the potential next steps and applications of our project for the following semester.