Kiana-Layam

[kkvlayam]

Weekly Notebook Entry — Week [3]

Quick Overview

We all got assigned to teams and had an overview of the current project status.

Task 1: read through the github repository from spring

Weekly Accomplishments

What tasks did you complete this week?

• Introduction to Models meeting
• got an overview of the progress that was made in the spring as well as a quick introduction to the machine learning models they explored

Tasks are still ongoing? Current status?

• read through the github repository from spring

Challenges and Learning

What was the biggest challenge you faced this week? How did you address it?

• I am not familiar with machine learning, so it was daunting knowing that this project has a lot to do with programming. Luckily, the others on the team were open to answer any questions that we had about the project. Anubha helped answer my questions

What's one new thing you learned or skill you improved?

• I learned about many different types of models for machine learning

Did you attend any team meetings? Key takeaways?

• Energy in buildings Kickoff meeting
• Introduction to Models Meeting with entire Energy in Buildings team

Reflection and Planning

Your progress this week?

• Better understanding of the project

Main focus for next week?

• Reading through the github repository and get a brief understanding of the models we used and how we are going to use them

Any resources you're looking for?

• no

[kkvlayam]

Weekly Notebook Entry — Week [4]

Quick Overview

Continued Model Explanation

Task 1: Review Model Papers to prepare for an in-depth model explanation meeting Task 2: Review and familiarize myself with the progress from spring

Weekly Accomplishments

What tasks did you complete this week?

• Quickly read through the model papers to get a brief understanding of the model's we're using
• read through github repository and journal issues

Tasks are still ongoing? Current status?

• understand what goes into making a good model
• start finding the input variables needed for the dataset to train the model

Challenges and Learning

What was the biggest challenge you faced this week? How did you address it?

• Understanding my role in the project since I do not have much coding experience. The best thing I can do is understand what goes into a good model and try my best to help with the dataset.

What's one new thing you learned or skill you improved?

• I learned about the importance of a quality dataset to train the model. Helped emphasize that we need to create/find a good dataset in order for this model to be accurate.

Did you attend any team meetings? Key takeaways?

• In-Depth Model Meeting (09/12): reviewed models more in depth, understood that the ASHRAE model is most likely too involved for this project
• Meeting w/ Prof. Kastner (09/13): Start looking into a creating a grasshopper script to create a synthetic dataset to train the model

Reflection and Planning

Your progress this week?

• I definitely have a better understanding of the scope of the project. I also understand my role in the project more since we are attempting to create a parametric script using grasshopper. I have never used grasshopper before, so I will have to learn this on my own time.

Main focus for next week?

• Learning grasshopper and thinking about a parametric script to help create a dataset with lots of diversity.

Any resources you're looking for?

• grasshopper tutorials

[kkvlayam]

Weekly Notebook Entry — Week [5]

Quick Overview

Starting to create a parametric grasshopper script to create a synthetic dataset

Task 1: Learn how to use grasshopper Task 2: Start thinking about a parametric script to create a diverse dataset representative of buildings in Atlanta

Weekly Accomplishments

What tasks did you complete this week?

• Watched Grasshopper tutorials
• Found papers that use a parametric scripts for energy analysis and attempted to recreate their work

Tasks are still ongoing? Current status?

• continue learning grasshopper
• continue finding a way to create a parametric script with a lot of diversity

Challenges and Learning

What was the biggest challenge you faced this week? How did you address it?

• I have never used Grasshopper before and it seems like there is a learning curve. I am trying to watch youtube tutorials to help me learn. Also along with Grasshopper, I need to learn how to use honeybee.

What's one new thing you learned or skill you improved?

• Learning how to use Grasshopper. The basics of using it and getting it to do what I want.

Did you attend any team meetings? Key takeaways?

• VIP-Class (09/16): learned about the NREL building stock datasets
• First Group Meeting (09/17): Have a better understanding of what we need to do as a group, look into Honeybee tutorials
• Meeting w/ Prof. Kastner (09/20): Look at example parametric script from Prof. Kastner, look into ResStock

Reflection and Planning

Your progress this week?

• I think we had some good progress this week. We have better idea of the direction of our project. I need to dedicate a decent amount of time to learning how to use Grasshopper which will be good for the class, but also for the program.

Main focus for next week?

• Watch Grasshopper, Ladybug, and Honeybee tutorials
• Look into ResStock

Any resources you're looking for?

• Grasshopper, Honeybee Tutorials

[kkvlayam]

Weekly Notebook Entry — Week [6]

Quick Overview

Looking into ResStock and understanding how to use their dataset for our model

Task 1: Look at the inputs and outputs of ResStock datasets Task 2: Define goal and problem statement of project to help create team organization

Weekly Accomplishments

What tasks did you complete this week?

• Read into the ResStock dataset creation to understand where the info was coming from that way we know how to implement their dataset into our project effectively
• Found the data dictionary for the ResStock Datasets
• started to filter data columns to the input and outputs that we need for our model training dataset

Tasks are still ongoing? Current status?

• finding the right combination of inputs and outputs for our dataset

Challenges and Learning

What was the biggest challenge you faced this week? How did you address it?

• Now that we know we can use the ResStock dataset, we need to find out how to use it. There a lot of information that it provides, so a big task is to understand which variables we need/don't need. I found the data dictionary which should help with that

What's one new thing you learned or skill you improved?

• Continued watching grasshopper tutorials

Did you attend any team meetings? Key takeaways?

• Team Meeting (9/23): created problem statement and plan for the week

Reflection and Planning

Your progress this week?

• We finally got a good data source, so all we need to do is prepare the data for the model by filtering it to what we need before we pass it onto Shivam for the model. Even though Shivam ran a first test of a test dataset, there is still a long way before we will have a good model trained

Main focus for next week?

• Find the input and output variables for the dataset to train the model better
• Look into the IDF files

Any resources you're looking for?

• no

[kkvlayam]

Weekly Notebook Entry — Week [7]

Quick Overview

We all got assigned to teams and had an overview of the current project status.

Task 1: Filter ResStock Dataset and give to Shivam Task 2: Look into IDF files to extract the building geometry from ResStock

Weekly Accomplishments

What tasks did you complete this week?

• Finished filtering ResStock dataset so Shivam can try to run the model with a more input and output variables
• Found a way to convert OSM file to IDF

Tasks are still ongoing? Current status?

• need to find a way to extract building geometry from IDF file
• find what information we need from IDF file
• find a scalable workflow to process all 17k building files

Challenges and Learning

What was the biggest challenge you faced this week? How did you address it?

• We did not know how to extract the geometry from the given models on ResStock. Hang looked into it and was able to find a manual workflow. However, since we have so many geometry files we need a more scalable workflow to extract geometry information to inform our dataset. Both Residential and Commercial teams need to do this so we are working together to find a solution.

What's one new thing you learned or skill you improved?

• starting to learn how to use python to preprocess data

Did you attend any team meetings? Key takeaways?

• Meeting w/ Hang to discuss IDF files (10/3): Learned that we are provided with OSM files not IDF files. Now we need to find a way to extract geometry from the OSM files efficiently.

Reflection and Planning

Your progress this week?

• Attempted to run the model with a larger and more inclusive dataset to see how that will impact the correlation of different variables
• Found a manual workflow for getting geometry from OSM files

Main focus for next week?

• find a way to automate extracting geometry from OSM files to inform our dataset. If we are able to get the geometry our dataset will be more robust.

Any resources you're looking for?

• python script or grasshopper script for reading a large amount of IDF files

[kkvlayam]

Weekly Notebook Entry — Week [8]

Quick Overview

We all got assigned to teams and had an overview of the current project status.

Task 1: Filter ResStock Dataset and give to Shivam Task 2: Look into IDF files to extract the building geometry from ResStock

Weekly Accomplishments

What tasks did you complete this week?

• Finished filtering ResStock dataset so Shivam can try to run the model with a more input and output variables
• Found a way to convert OSM file to IDF

Tasks are still ongoing? Current status?

• need to find a way to extract building geometry from IDF file
• find what information we need from IDF file
• find a scalable workflow to process all 17k building files

Challenges and Learning

What was the biggest challenge you faced this week? How did you address it?

• We did not know how to extract the geometry from the given models on ResStock. Hang looked into it and was able to find a manual workflow. However, since we have so many geometry files we need a more scalable workflow to extract geometry information to inform our dataset. Both Residential and Commercial teams need to do this so we are working together to find a solution.

What's one new thing you learned or skill you improved?

• starting to learn how to use python to preprocess data

Did you attend any team meetings? Key takeaways?

• Meeting w/ Hang to discuss IDF files (10/3): Learned that we are provided with OSM files not IDF files. Now we need to find a way to extract geometry from the OSM files efficiently.

Reflection and Planning

Your progress this week?

• Attempted to run the model with a larger and more inclusive dataset to see how that will impact the correlation of different variables
• Found a manual workflow for getting geometry from OSM files

Main focus for next week?

• find a way to automate extracting geometry from OSM files to inform our dataset. If we are able to get the geometry our dataset will be more robust.

Any resources you're looking for?

• python script or grasshopper script for reading a large amount of IDF files