Shivam Patel

[FlippyShivam]

Weekly Notebook Entry — Week []

Quick Overview

Top 3 tasks

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

• [ ] Task 1: Detailed description and expected outcome.
• [ ] Task 2:
• [ ] Task 3:

Weekly Accomplishments

Tasks completed

Tasks ongoing

Challenges & Learning

Biggest Challenge

Takeaway, Learning

Reflection & Planning

Progress

Main Focus next week

Resources

[FlippyShivam]

Weekly Notebook Entry — Week [1]

Quick Overview

Team onboarding, new people join in, and we all decided what projects we want to join.

Top 2 tasks

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

• [ ] Task 1: Prime myself on where we left off last semester
• [ ] Task 2: Fill form to see what team everyone is interested in

Weekly Accomplishments

None yet

Tasks completed

Catch up on reading

Tasks ongoing

Catching up

Challenges & Learning

We left on a slight weak note, so we need to push to create a solid program and model

Biggest Challenge

None yet

Takeaway, Learning

We split Energy in Buildings into two teams, commercial and residential. I am in Residential.

Reflection & Planning

NA

Progress

• NA

Main Focus next week

Meet new team, start onboarding for specific project

Resources

https://github.com/VIP-SMUR/24Fa-EnergyInBuildings-Res

[FlippyShivam]

Weekly Notebook Entry — Week [2]

Quick Overview

This week we all got our assigned teams, and started the onboarding process on what we all did last semester (Spring 24).

Top tasks

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

• [ ] Task 1: Have a call with new team members to onboard them and explain what the team and project goals are, and rundown on what was accomplished.

Weekly Accomplishments

Held multiple meetings to discuss updates on what research everyone has done. Last team was all CS, and now we have more diversity with architecture students and CS students.

Tasks completed

Held meetings, and created respective team group chats for people to get to know each other and communicate goals.

Tasks ongoing

Teach how computational modeling works, and overview on what we looked for.

Challenges & Learning

Most people do not come from a CS background, so it was important to teach them how this modeling project whilst also taking inputs to better understand architectural play as I do not have much experience in that.

Biggest Challenge

Onboarding and playing catch up

Takeaway, Learning

• We may have to spend 1-2 weeks to go over what we did last year and explain core modeling concepts.

Main Focus next week

Possibly have a workshop to discuss how the Multivariate model was synthesized.

Resources

https://github.com/VIP-SMUR/24Sp-EnergyInBuildings

[FlippyShivam]

Weekly Notebook Entry — Week [3]

Quick Overview

We continued research on modeling and explaining it more in-depth

Top tasks

• [ ] Task 1: Reviewed papers so I can catch up on the reading, and also looked into PACE again (need access to rerun)
• [ ] Task 2: Have in depth meeting for model explanation.

Weekly Accomplishments

-Read through some papers, and also held meetings to catch up on what exactly residential may look for over commercial. -Held long meeting to explain further detail how Machine learning works.

Tasks completed

Read resources and held meeting

Tasks ongoing

Onboarding, and attempting to still run ASHRAE

Challenges & Learning

ASHRAE is still taking a while, may have to communicate with Dr.Kastner how it may be a lost cause because it takes a long time anyway.

Biggest Challenge

• Now with the split of the team, finding out what specific task we should attack. While it is understood we will now look into heating/cooling outputs at a residential level, understanding the specifics of how we will reach is still undecided.

Takeaway, Learning

-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 & Planning

Everyone gained a better understanding of project.

Progress

-One final push on ASHRAE, if not the focus on models we have now, like Multivariate.

Main Focus next week

Now build an understanding on how we will tackle residential heating/cooling outputs.

Resources

none

[FlippyShivam]

Weekly Notebook Entry — Week [4]

Quick Overview

Attempt to create synthetic data, with some created via parametric grasshopper script or via https://www.mockaroo.com/

Top tasks

• [ ] Task 1: Gave final run on ASHRAE, then decided its bets to move forward with Multivariate. Attempting to run it and get an understanding on how it works.

Weekly Accomplishments

I am confused on why we use synthetic data, when we want to train model with real data.

Tasks completed

Built understanding on synthetic data

Biggest Challenge

• I'm not to sure on what we are trying to do or where we want to head now.

Takeaway, Learning

• Talked with Hang and Dr. Kastner. Built understanding that synthetic data would be good to create the pipeline for the project and then we can worry about collecting more robust data.

Reflection & Planning

• 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

Progress

• We may look into creating parametric data, so waiting on data validation to then see where to go further.

Main Focus next week

Collect data then run a base model to see what we can get from Multivariate

[FlippyShivam]

Weekly Notebook Entry — Week [5]

Quick Overview

Look into ResStock Data, and then understand which variable may be useful. (There are many inputs and outputs).

Top tasks

• [ ] Task 1: Grab potential data, and choose which ones we may want
• [ ] Task 2: Run data analysis on those columns, then see if I can run Multivariate on it.

Weekly Accomplishments

Ended up running Multivariate Model on 7 variables:

Tasks completed

Did analysis on variables that we thought maybe important to the collect total expected energy output.

Tasks ongoing

Fix Multivariate code (some parts had bugs) and continue with better variables.

Challenges & Learning

Understanding which columns in csv to use, and also better understanding why some outputs where not giving results that we would expect.

Biggest Challenge

• Choose which data portions to analyze, and what form of labeling to do (Label Encoding).

Takeaway, Learning

• Choosing columns is hard, Orientation was not showing expected output, may have to figure that out.

Reflection & Planning

Ask next week why orientation is not showing expected results, and choose better variables to then build better model.

Progress

• We are at a good stepping stone, but must move forward before we have a proper model.

Main Focus next week

Get better data, and possibly figure where to input dimensions and why orientation is not shown as significant (it SHOULD!!).

[FlippyShivam]

Weekly Notebook Entry — Week [6]

Quick Overview

Met team to discuss how to create a better dataset out of ResStock, understand what some variables could mean, and which inputs to run in Multivariate.

Top 3 tasks

• [ ] Task 1: Get another list of better columns to test for Multivariate, and then also see what columns we can manipulate and create (like getting IDF and dimensions) to then see what impact it has.

Weekly Accomplishments

Getting further in parametric in grasshopper, so hopefully we can run them and see if we get better results for outputs.

Tasks completed

Got a little busy with midterms, hopefully will try to do weekend.

Challenges & Learning

Now we must focus on the math and analysis of the data. ResStock does not give full dimensions, so we must combine our information and see of we can create simulated data and see if it has better correlation. (Big one is orientation).

Biggest Challenge

• Fixing orientational effects on energy output.

Takeaway, Learning

• We want more emphasis on heating and cooling over other outputs.

Reflection & Planning

Progress

Main Focus next week

Resources