07/28/18 (+Final, Beyond): Jiin

Last day of REU! On Wednesday and Thursday, I spent time finishing formatting our protocol in LaText with Overleaf, cleaning and uploading my code to Git, and cleaning my desktop for future users. I also attended workshops on grad school application and industry job search, as well as the Stauffer Series on the commonality of Shakespeare and Python.

Today, I prepared for our final REU in Data Science presentation, went to a final CS/DS lunch, did a team debrief of the REU program, and presented for the last time! In addition, I improved my resident contact script so that Green Crew could edit the e-mail/text content more easily through Google Docs. I added a testing sheet and function so that it would be easier for Kanami to edit the content and test-run it on herself.

Final:

This summer was a summer of learning, fun, and coding experience for me! I especially loved all the workshops and lectures, where I heard talks ranging from machine learning to operations research! I know that I am taking back transferrable skills such as problem solving, presentation, and teamwork, and I really enjoyed the experience of doing community outreach and applying technology to social justice issues.

Beyond:

(8/1~8/2)¬†Of course, you are never done with debugging/developing! I was in contact with Kanami, because she was running into some errors with the contact tool. I couldn’t find any error with my code, so I looked through her Google Sheets, and found that there were some issues with the content (missing resident name, values in the wrong columns). Hence, the e-mail was not sent for the rows following the first error discovered.

While I would like to make error messages easier to Kanami to understand and fix herself, that would take a longer time to develop. So what I did was make it easier for her to revisit her problems by adding two new menus, resendEmail and resentText that will allow Green Crew to resend e-mail or text for a desired row range that they enter. It also allows them to choose which content they want to send (weekly/custom). This will prove useful in case a function failed due to some error in the content and some residents did not get the message.

It’s exciting to have my tool put into real world application, because now I get to see what additional features would be nice to have, and some problem areas the user runs into!

Advertisements

7/26/18: Heather

Notable issues from today: I found out that our final benefit analysis calculator didn’t work. Which is terrible. However, I went back through some of our old versions and found one that did work. After copying the code over, tweaking it, and re-commenting it, it consistently works (hooray!). I even tested its limits and made notes about it in the protocols. I’m very happy with how this has turned out. Kanami has been in communication with us and is also happy about this. It is such an odd feeling to create something, then have it be actually and immediately used by a client.

Done today:

  • Improved explanation about summing results
  • Fixed structure of protocol
  • Fixed issues with calculation errors (this was a big issue!)
  • Tested the limits of the benefits analysis
  • Improved the Green Crew protocols
  • Commented code in general

To do:

  • Continue improving the Green Crew protocols
  • Work a bit more with explaining the website
  • Create a better explanation about adding the benefits
  • Put protocol into overleaf

7/25/18: Heather

I continued working on the protocols today. I also had the opportunity to go to a workshop about job applications and grad school. I got to learn a little more about both, which has been very good for my decision process. I think that I’ve mostly made up my mind about not going to grad school. I’m very glad that this program provides all of these workshops and info sessions so that we can learn about things you never have the time for during the school year.

Done today

  • Note about the lack of individual benefit calculations
  • Comment out individual benefits in script and improved comments in script
  • Made a version of the protocol for the Green Crew
  • Made notes about lack of flowcharts for individual benefits

To do

  • Improve the Green Crew protocols
  • Talk in more detail about how you have to manually add results
  • Comment code in general
  • Work a bit more with the website

7/24/18: Heather

Today I worked on finishing up the paper.

Things done today:

  • Finished overview
  • Updated protocol format
  • Created flowcharts for protocol
  • Created a key-terms list
  • Put in some sort of explanation about Individual vs. Overall benefits in the protocol
  • Explain the units for each of the benefits –> make a section of the protocol dedicated to the benefits themselves
  • Began transferring to Overleaf (Jiin)

Things to do tomorrow:

  • Put in an explanation + improvement ideas for individual benefits
  • Explain the multiple versions of the google sheet
  • Finish transferring to Overleaf
  • Upload to git

We’re also going to dinner tonight! I hope it goes well!

07/24/18: Jiin

Today, I finished cleaning our Python codes for UTC and uploaded it to Git. I also worked on transferring our protocol in Google Docs to LaTeX. I am amazed at how beautiful LaTeX allows you to format papers, although making tables in LaTeX is a nuisance. Today’s CS Chat was on abnormal crowd behavior detection, which was really interesting!

Tomorrow, I will finish moving our protocol to LaTex in Overleaf. I will also finish cleaning up my remaining code that uses Twilio.

7/23/18: Heather

Today I worked on finishing the protocol for my part of the project. It has been somewhat difficult to sort through my script, which has many different parts and is not organized particularly well. It is in part Apps Scripts’s fault, as there isn’t a good way to view/work with/organize scripts with multiple features. I’ve been commenting the code along the way as I write the protocol, and it is certainly neater. I am technically done with the protocol as of now, but I still want to read it over and make it clearer. I also want to write an overview for the entire project and write some research questions for next year’s group. Tomorrow I will work on those things, plus I should begin uploading my code to git.

07/23/2018: Carla

Today, I was able to work on the protocol that we needed for our Complete Streets project. I am halfway done, and was able to work on a part of the introduction, segment analysis, and automating data. I have never written a protocol before for anything so this is something that is really interesting to learn from. At Pitzer, when I write I usually write for research papers in sociology or philosophy, but today I had to write from a scientific approach. To be quite honest, it was quite refreshing. The past week, I was working on finalizing the protocols for automating data, which took some time because I had not necessarily written or reported things down. I was trying to follow Moe’s protocol as well, but for the entire City of Claremont, which helped us get results for our REU gathering at USC on Friday.

unnamed-1.png

At our presentation at USC, it was quite quick, but I felt confident when I was presenting to people. Over the summer, I was glad to improve my presenting skills, as I really had not done that in a while or in a CS setting. I really liked watching my teammates present and hearing each other improve over time. I believe I had a lot of growth in my presentation skills within the last few weeks, which is really helpful for my college career. Hopefully I can apply the same skills.

Tomorrow, I plan to finish the protocol for Complete Streets. If I am able to finish early, then hopefully I can work on editing it as well.