engsci s2 recap #vibes?

My sincerest apologies for being MIA over the past couple of months. All I can really say is I was goin thru it ! As in like crying in the architecture toilets .. crying in the kate edgar toilets .. u name it I’ve cried there probably #mentalbreakdowns and can’t lie, this semester was the worst ever (but also best ??? weird) since I started uni but it’s over now so like … its k 💯

Anyway, here is my semester 2 recap! I did 5 papers this sem, 3 of which were Engineering and 2 were Arts (for my conjoint). While I was really really mf stressed this semester, it honestly was fun, and exam season was the best ever bc we had one exam for engsci and it was on October 28th!! Life has been good for the past 2 weeks … maybe suffering is worth it sometimes ??

 

ENGGEN 204

Not sure if I’m the only one but 204 was my worst paper this sem lol- I’m not going to go into the frustration this paper caused me but it’s the general group project paper that all part 2s take. The advice that I’d like to give you about this paper probably is not in my best interest to put on a public forum (digital footprint and all) so all I’ll say is please contribute to group work because everyone always appreciates a good group member that actually gets work done.

 

ENGSCI 263

263 was the engsci group project paper, but admittedly more interesting- the first section was computational mechanics (CM) where we worked in groups of 4-5 to create a proposal for a client. There were 5 different projects, 4 of which were related to a pressure ODE, and the last project (which was mine) was related to a temperature ODE. Essentially, we had to create a model for the system (basically coding up an ODE in Python), create benchmarks and unit tests, find the model fit parameters, and perform uncertainty analysis. After that whole process, we presented our findings and recommendations to the other teams. Finally, we individually wrote a report detailing all the steps we took and what our findings were, and made a recommendation to the client. In between the project stuff, we also had weekly labs that taught us some coding skills in Python that helped with our projects. It was a lot of work, but I had the best group ever so it honestly was just a great time

The second section was operations research (OR), where we worked in groups of 4 to generate a minimal cost truck logistics plan for Foodstuffs supermarkets. Basically, we were given the number of pallets delivered to the 48 different supermarkets over 4 weeks, the durations and distances between them and the warehouse, and constraints (like costs and capacity of trucks). Using this information, we had to estimate demands at each supermarket, create a set of feasible routes, then use PuLP (a linear program in Python) to solve for the optimal solution. Part 2 of the project involved simulating demands to ensure our program worked, and we came up with a cost and a solution. All of this was then summarised in our individual reports, where we wrote a recommendation to Foodstuffs, as well as considered wider implications.

The exam overall was okay, quite long and stressful in my opinion, but it could’ve been worse. All in all, I actually enjoyed this paper and I’d recommend it (but like it’s literally a core paper and you have to do it to graduate so not rly sure why I’m recommending it but yes 👍 )

 

ENGSCI 205 

(Bit of a note: engsci has electives in second semester, normally you get 2 but bc I’m a conjoint, I only got 1- don’t rly know much about the other electives but the main others that I know people did were MECHENG 270 and STATS 210. Would def recommend doing some research before picking your elective(s) because they can either be hell or be really fun and useful)

Very elusive- when I was trying to find what elective to do, all the information I could find about this paper was that it was a ‘special topic’, and that we’d do geospatial modelling and machine learning? I think both the lecturers of this paper aren’t even at uni next year so this paper won’t be taught in 2023, but like if anyone is reading from the future maybe idk this could be helpful

The first section was geospatial modelling, which was coded mostly in R. We used R packages like tmap, sf, and shiny, and learnt to read in data, convert data to the right coordinate system, and plot maps. This paper was assessed mostly in labs, which spanned around 3-4 weeks each- the first lab was made of 3 parts, of which part B was using NZ statistical area data and cropping it to Auckland, then plotting different maps of the region based on population and train stations. Part C was mostly rasters, we made and plotted contour lines, downloaded our own satellite data of somewhere around the world, made rasters of that, and predicted land usage using a classification tree. The second lab was also lots of rasters, this time using kriging and gstat/variogram functions for plotting interpolated rainfall. Then, we did some pretty invasive stuff I must say, using addresses in Auckland and the OSRM library to create maps that showed the shortest distance between 2 houses- and then you guessed it, more rasters, this time for the distance travelled by an EV. Finally, the end of geospatial modelling labs, we created our own shiny apps/dashboards for NZ electricity demand. There was also an in-person test that was insanely difficult for me (cause tbh idk what I was doing half the time), but overall this section was super helpful, and I definitely got way better at coding in R.

The second section was machine learning in Python, which was like the whole last half of the semester- if I’m being honest I’ve kind of forgotten what we did because the last few weeks of the semester were ROUGH but i’ll try my best. Essentially, we were given housing data with prices, and a bunch of other feature data like number of bedrooms, average area of living room, etc. Using this data, we used scikit learn and regression functions to create a house price prediction, which we submitted to kaggle (which ranked us all based on our prediction). There was also an in-person test, but our lecturer was very lovely and made it quite a bit easier than the first one, so that was just beautiful

205 was a really good choice in my opinion, cause got pretty good at coding in R and got to do a bit of machine learning which was quite fun- even though it’s not offered next year, hope it will be again soon

 

I’ll do a post on conjoints n stuff next, but yes hope this helps even a little bit – have a good break!

Leave a Reply

Your email address will not be published. Required fields are marked *