This morning most of the DSSG students had software carpentry tutorials again. Kivan wasn't here today because he is on vacation. I started my morning by downloading the files we are going to need in the afternoon for our Graph Lab tutorial, which is the second part and the hands-on period of the two session tutorial series. Eventually, my team regrouped after getting out of the software carpentry and Frank pushed new code. The important part of his new code was the it grouped the R files into a project, which should make the future R work more organized and easy to work on in a team environment. I had just finally finished getting the right syntax for keeping track of the different leg times when I realized that the new code Frank pushed included a field that also would count leg times. He also had created a column to keep track of the number of passengers on board, so now we should have everything we need to find the ugly rides. I spent the rest of my day trying to figure out the syntax and how we're going to code that up. I just get tripped up trying to figure out which variables I will want to iterate over first to make it most efficient.
     Our day ended up the Graph Lab tutorial. It was cool to actually see an example of what it does and to hear more about the vocabulary used with machine learning, but I was really disheartened because my iPython notebook wouldn't open even though I had everything downloaded and everything! Since I didn't actually get to play with the code, the tutorial wasn't as cool for me, plus I was really tired and it was the end of the day on a Friday. Overall, it could have been better.

     This morning we all decided to meet with Anat and have her give us an introduction into the concept of dynamic programming, and how that fits in with our rerouting problem. I have not had much experience talking about dynamic programming, and I have not done much with routing problems that are as complex as ours, so it was helpful for me to get some more background on just how to even talk about and discuss the problem formally. We're planning on continuing our discussion of dynamic programming on Tuesday, which I think will be helpful because I'm still not as clear as I would like to be on how to apply it to our situation and what constraints we need to define and which we could, should, or won't need to define. 
     So everyone in the DSSG program was offered to the chance to sign up for these Software Carpentry classes today and tomorrow, but I was not put on the email because I'm not technically a DSSG intern. Software Carpentry is a series of classes and tutorials that are basically meant to take you from beginner to intermediate level on basic programming skills (R, Python, bash commands, etc.) very quickly. Turns out that everyone in my group except me had signed up and said they would do the classes, so after the dynamic programming intro I was mainly working on my own as the only one really focused on the project. Thankfully, my team members would come out and check in with me and help clear up questions I had whenever they had breaks from the Software Carpentry sessions, and Rohan actually skipped the second half of the afternoon session because he felt that it was too much of an introductory class and he was not really getting much out of it. Also, Valentina didn't sign up for the classes, so she was working with me on fixing an issue* that came up when transferring the full data set to the SQL server. Along with this, I was also asking Rohan to help me while I tried to remember the syntax I would need to create a new column for our dataframe in R that would track the time (based on ETAs) for each leg of a run, which would in turn be used to help us find the ugly rides. I haven't programmed in R in quite a while, so today was pretty frustrating with just trying to remember simple syntax problems. It helped that I started by reviewing Frank's code to see what his does and how I might modify it further. 



*For those curious, the issue was that when transferring from Access to csv to the SQL server, we were getting quotation marks in every string field in the SQL server database, which is incredibly annoying for querying because then you have to do the specific single quote then double quote syntax whenever you want to search or filter on a specific string. To get rid of this, we had to actually start again from the Access db file and separate on semicolons instead of commas because one of our fields had cells that had text occasionally separated by commas. 

     This morning started off with a two hour long session focused on team building and collaboratively working with a group. I was pretty excited for this session because I love team building and bonding exercises, but the session ended up feeling very...unnecessary. It was just a very general talk about what it meant to work in a team and we did different brainstorming activities that I often felt were unhelpful since they were just forcing us to think about things we already knew. It was cool to get to spend more time with my team though.
    After lunch we had a paratransit group meeting where we Skyped in Anat (she was home sick today). We talked generally again about our project and updated our goals to reflect everything that has happened and that we've learned since our meeting with Matthew on Friday. We decided to break the project up into two main objectives: working on the optimization for rerouting, and identifying and analyzing the ugly rides. After the meeting we tried to come up with a Google Doc that outlined all the will be entailed in these two projects, but we still have yet to say who is working on what. I feel like I could be a lot of help on both projects, but I like how much more structured the ugly ride project is, so I might try to work on that part if given the choice.
    Our last thing on the agenda today was a very basic python intro tutorial session. It was really aimed at people who have just about no coding experience and no python experience whatsoever. I really don't have any formal experience using python; I've only modified scripts that I was given by other students on the same projects I've worked on. Therefore, I thought it could still be pretty helpful to me. Actually, I do think it was pretty helpful to get some slow and very basic intro into just the syntax for python and using ipython notebooks, since I've also never had the opportunity to use those before. However, at some point in the tutorial I started to feel that it was going a bit too slowly over concepts I was already intimately familiar with (for loops, if else statements, etc) and since the syntax was very similar to what I already know I started writing this post instead. :)

    I started off the morning finally catching up on some of these blog posts. Whoops! It just gets very busy when I'm working in the eScience center, which is good, but a challenge to writing daily blog updates. After catching up on downloading GraphLab and ODL, I went back over the the CSE building for the first time in about a week to meet with Anat and Zhitao about our projects. Anat also announced to us that she wants us to help her make these engineering start up kits that she's going to use in workshops with the visiting high-schoolers. I'm really excited about it because it sounds cool, and honestly I know just about nothing about engineering since I have always stayed on the programming side of computer science. With that, at our meeting next week we are supposed to have a tutorial with Anat about 3D printing, which we'll use to make the different parts and pieces. I'm looking forward to it.
    Back at eScience, we had a GraphLab tutorial that was pretty cool because it outlined the importance and gave an overview of machine learning. I don't have much experience dealing with machine learning, but I'm excited to learn more about it.
    After lunch we met with the ALVA students for the first time! That was fun. We did a scavenger hunt in our project group, and we were the first ones to make it back so we won! Nic is the ALVA student assigned to our group, and he seems like a nice guy. We tried to catch him up to speed, but I'm worried he might start feeling a little overwhelmed soon since there are just so many complexities in our problem and project. 
     Our last workshop of the day went over how to create pages and blog pages from github. I had never heard of Jekyll before, so it was very informative, but I don't think I'll be using any of it for this project since I already have this blog here. 
     At the end of the day, I worked with my team.