https://www.kaggle.com/rishpande/avocado-prices-data-visualization-beginner
We had selected some datasets from Kaggle to choose from. Maarten suggested a dataset on Avocados, Maja one on world happiness. We went with the one on avocados.
This dataset includes information on the sale and production of avocados in various regions of the US. We decided to compare changes and differences over time and per region.
- todo: Maarten, can you comment on this?
We were still figuring out some things about matplotlib. The work resulted in some fairly basic plots. We discussed what kind of plots would be meaningful, but had a difference in opinion on whether this was needed. On one side the idea that it is better to choose meaningful plots that tell something more about the data, on the other hand the idea that "We are learning, the meaning is secondary."
Most of the basic work was done on October 27, 2020. Future additions pending.
A fair bit of talk went into discovering just how closely to work together. This resulted in a first phase of independent exploration. As a team we have the advantage that we can spontaneously exchange the findings from the explorations phase and this made for a fluent transition into more coöperatively coördinating ideas and selecting output to use in the end.
- Make some basic plots about average price and total volume over time and per region.
- Figure out how to combine multiple columns
- Make a 3D plot
- Add text sections between the graphic plots to introduce and clarify them
- Update this readme file with a "why?" section
- Going over more theory while peeking at the datasets on tuesday.
- Settling on the avocado data and going over it together while on the discord chat.
- Maarten made a repo and shared with Maja. Maja branching off of master.
- Toying around with the data and both coming up with some unusable first plots.
- First usable plot made by Maja while Maarten following along.
- Distribution of plots among each other.
- Working on plots separate, while helping with solutions on discord.
- Combining work and streamlining files on github.
In my opinion a good exercise. I was glad Maja was there to shoot questions at. It was 33 procent teamwork 33 procent individual work and 33 procent problem solving each others work. Which for me is the perfect ratio. Which leaves only 1 percent unaccounted for. ;)
Good excercise, much learned. At some point I felt like I was left to Google for help, but later on we were successfully co-working on each others code issues. We combined our ideas, explored them in chat. There was space to communicate about workflow and applications. Even now we're still exchanging on how we can make the code and the graphs work better. On the whole it was an interesting and enjoyable collaboration. One point of feedback is I would have wanted to make our choices more data-driven, instead of the rather random selection now.
Our results are visible in the single output notebook. We made in total four line plots, a bar chart and a pie chart, choosing data and visualization methods that may reveal something meaningful about the chosen dataset.