Dataset:
Mongolian Air Pollution Dataset: https://1212.mn/en/statistic/statcate/573072/table-view/DT_NSO_2400_015V2
I got this dataset from 1212.mn and this is a statistic website for my country where I can request various datasets such as air pollution and I can pick which location/year/indicator I want. I requested PM2.5(measures air pollution) dataset in .csv format. I chose this dataset because I was interested how air pollution progressed in my city area from 2015 to 2024. To plot it, first I converted each PM2.5 value from mg/m^3 to µg/m^3 and then scaled it in a range of 1% to 10% and I smoothed the edges.
COVID-19 Recovered per 100 confirmed cases: https://www.kaggle.com/datasets/imdevskp/corona-virus-report?select=day_wise.csv
I got this dataset from Kaggle and the reason I chose this dataset is because I was interested how many people recovered per each confirmed 100 cases from 1/22/2020 to 7/27/2020. To plot it, I used 10th column(Recovered / 100 cases) and scaled the values from 1% to 20%.
Mongolian Density Dataset: https://1212.mn/en/statistic/statcate/573051/table-view/DT_NSO_0300_034V1
Similar to air pollution dataset, I got this dataset from 1212.mn. The reason I chose this dataset is because I wanted to visualize my city density from 2000 to 2023. To plot it, I scaled the values from 1% to 10%.
Design Process:
I used variables from the dataset such as time and certain category that I was interested in to shape and scale the 3D forms. In Mongolia air pollution, the values in the dataset are measured in mg/m^3 representing PM2.5 air pollution measurement. In COVID-19 dataset, the values are integers from 1 to 100 representing recovered cases per 100 confirmed cases. In Mongolian density dataset, the values are integers representing population per unit of land area.
I chose these forms because they represent trends and patterns in the dataset making it best for data physicalization. Each form has trends that a person might not be able to see just by looking at raw .csv file. For example, In air pollution, we can see that in the air pollution is hazardous every year in some months(if we examine further, we can see air pollution is hazardous in winter time because Mongolian winters can get to -30 celsius, thus people need to burn more coal or use more electricity to stay warm which makes air quality worse, however air quality each year gets a little better each winter because Mongolia progresses). In density, we can see that population density rises each year and for COVID we can see there is huge spike in 2-3rd month for recovered cases which we could assume that treatment might’ve gotten better.
Images of forms generated by Grasshopper/Rhino program and rendered in Rhino:
COVID-19 recovered per 100 cases:
Mongolian Density:
Mongolian Air Pollution:
Images of final printed artifacts:
COVID-19 recovered per 100 cases dataset:
Mongolian air pollution dataset:
Mongolian population density dataset:
A reflection on designing the data:
3D objects can help people understand data better by visualizing it. For example, when you turn data into 3D object you can see the patterns and trends. When you see raw .csv file, it might be harder to visualize it. The process of creating 3D objects from data allows us visualize and hold the object. These objects can have unique meanings because they make the data feel more real! It feels more easier to grasp the data!
Hey Bat!
Your artifacts look great! Did you have any trouble with supports? Did you use any? The covid and air pollution artifact seem like they absolutely need supports but I could be totally wrong so I’m curious!
I agree that looking at a raw .csv file is incredibly difficult to look at especially when you have 10+ categories. It all just seems like a jumbled mess. By making these into forms we can hold it’s much easier to deal with.
Thanks for sharing!
Thank you! And yes, I had trouble with supports on the air pollution dataset. I tried to use both tree and normal support, but the final artifact using tree support was breaking so I used normal but in final artifact, it’s hard to get rid of the support in between. For covid, I used normal support and it had no problems!
Hi Bat,
I like the fact you used COVID-19 as your dataset considering it was a recent event we can relate to. The shape of the object is really interesting. How did you configure the data in order for it to come out like that? Its almost like it came out reverse. I’m guessing in order to print out this object you printed it upside down? Its nice to see the spike in the object to know that people were getting better.
Hi, Ricardo!
Thank you for your comment! I printed it with support so I didn’t do it upside down but now that I think of it it might’ve been easier to do it. Also, I didn’t configure much except scaling and data cleaning! And yea, it’s nice to see that people got better exponentially.
Hi Bat,
I find it very funny that my first thought for a dataset was to use an air pollution one, the original one that I was going to use was for New York. I really enjoy the data representation for the COVID-19 Artifact, you can see how the trend ballooned towards the start with people recovering at a higher rate, then it dying down, and then finally increasing at a constant rate. I had the idea for the population in my head when I first read this post, but after seeing your Mongolian population density as a 3D print it gave me a greater sense of how populations can grow increasingly, very interesting to see there wasn’t a decline in the population.
Hi, Andrew!
Thank you for your comment! Air pollution got me interested because the city is popular for bad air quality. And for Mongolian population, I agree that populations can grow increasingly. It’s interesting to see it on a 3D object!