Joint, marginal, conditional distribution visualizer
Contact: R Pruim
Main idea: visulize a joint distibution of two random variables along with the conditional (slices) and marginal (“shadows”) distributions.
Possible resource: Use d3-3d or some other 3d visualizaiton add-on to D3?
Where in the world are Calvin students? (Interim Travel)
Craig Hanson idea based on seeing this YouTube video.
Data source: Cynthia Slagter has provided itineraries for each interim group.
Plaster Creek
Contact: Deanna Geelhoed/Julie Wild
Main idea: We think Plaster Creek Stewards has some potentially interesting data for a project. We have a series of pressure transducers that record ongoing measurements of the flow in Plaster Creek at various locations in the watershed.
Solar Panel Data
Contact: Herb Fynewever
“I put solar panels on my house in April 2019 and have production (watts) vs. time of day for every 15 minutes since then. Out of curiosity, I’d love to get a rolling average and a rolling maximum plot of production vs. time of day. We should be able to see the summer come and go as the days get longer and shorter, etc. Should also be able to estimate effect of my inverter maxing out on very sunny days.”
As far as visualization goes, I’ve seen some solar data visualized as a playable time lapse with a globe getting brighter and dimmer. That would work – light intensity should be a direct function of watts.
I’m used to graphs though, and I’d really love something like a 3-D graph like a tunnel you could walk through. Each day is a step forward (one axis), left to right is time of day (another axis), and the ceiling on the tunnel (third axis) is the wattage. So the tunnel is spacious in mid-summer and is a tiny crawlspace this time of year. It’d be cool to have two versions of the tunnel: one with a rolling average wattage (to smooth out “noise”) and one with a rolling maximum wattage (would simulate what ideal solar weather production would be).
A couple notes about the data:
I get power on cloudy days much more than on cloudy + rainy days. If the panels are wet, the production drops to zero until they are dry.
Production also drops to zero if there’s a power outage on the grid. This is frustrating, but I’m told it’s for the safety of line workers who could be hurt by solar customers sending power to the grid when they’re trying to get power back up.
Production also drops to zero if there’s a software error. This happened a couple times in November, each outage lasted for days.
Data: Excel spreadsheets in hand.
Cherwell Ticket Data
Contact: Lucas Moore
We have a lot of statistics from our IT work order system (Cherwell) and have started pulling it together in Tableau. Sarah Greenfield was helping me put together data/calculations I needed as well as design a dashboard we could display in the CIT office. Sarah is really busy with Workday at the moment so this project is on hold. It would be great to have someone continue her work and get the data I need available to me in Tableau via reports and dashboards. I would be the primary consumer of this data but would also like to “tell a story” via a main dashboard that would be accessible by everyone. We have a general mockup of what the dashboard could look like, as well as the types of data calculations I am looking for (Mean Time To Resolution, tickets closed by teams/individuals, etc.)
International Student Deadlines
Contact: Brent Wilkinson
Potential starting point/inspiration (with needed improvements): https://global.arizona.edu/international-students/optional-practical-training-opt
Explaining the date ranges and deadlines for international students on OPT is always a challenge since there are so many moving pieces. An effective infographic or calculator would be a useful tool in helping students make their graduation plans. Some graphics exist, but they leave out important information and do not show the timeline for applying to graduate school after OPT.
Church Volunteers
Contact: Paul Bylsma
“My church recently collected a bunch of information from the congregation, and we would like to make it available in a way that is easier to understand than a spreadsheet. The rows are each person’s responses, and the columns are everything from contact information to preferred methods of communication to various interests. Right now, I have filters on the columns to find out who (for example) can do a car repair and is reachable by text, but it would be nice to have easier ways of slicing this data and spitting out names who meet the requirements.”
Tornado simulations
Contact: Fred Haan
I have velocity and pressure data from a building model experiment in a tornado simulator. The data consists of time series of velocity from a sensor just above the roof of the building and time series of pressure from about 90 pressure taps on the surface of the building. Some of the data are for a stationary tornado and some are for a tornado that is moving past the building. What I would like is a visualization that shows a time-dependent velocity vector and time-dependent colors on the building (where colors indicate pressure magnitude) as the tornado moves past. Ideally, this would include some interactive tool where the user could move the tornado back and forth and see the effect on the building pressures. The intended use is for researchers browsing large data sets or for non-researchers who would enjoy an illustration of what this research is all about.
Race and Gender in Movies
Contact: Elisha Maar
I have a data set on the top 100 grossing films from 2016 listing the top five individuals’/characters’ perceived race, gender, age, etc. and whether they were a main or supporting character. We also have information about the film such as genre and amount earned (and the ranking of earnings). I haven’t had a chance to do much with it. Some of the initial findings were that when women are the leading characters they were younger, but middle aged and older women were more likely to be supporting characters. There were few Hispanic and Asian characters, and few people of color in Westerns. Black women and White women had similar patterns except that there were fewer Black women in dramas.
Interactive Syllabi
Contact: Krista Carter
First Year Seminar courses try to help students transition to college. Right now, they act as a bridge between orientation and academic coursework (and maybe life on campus in general). One idea I’ve been toying with (I co-direct FYS currently) is to try to help students gain comfort in working with a syllabus: understanding its purpose, gauging the importance of class policies, and seeing it as a pedagogical tool. Could your students take the FYS syllabus and make elements (or all of it) represented in a more user-friendly way that exchanges alphabetic text for something that could speak to students differently? If so, this might lead to something that could be shared with other professors on campus seeking to make their syllabi more accommodating and accessible.
Camera Settings
Contact: Jennifer Steensma Hoag
I would love to have a visualization to use in my photography classes. Ideally the visualization would be interactive and could help students understand exposure by showing the relationship between light, ISO, aperture and shutter speed. Changing one has an impact on all the others and students frequently have a very difficult time understanding how they relate. I would use this for both digital and analogue cameras.
Job Projections:
The US Bureau of Labor Statistics has released their latest employment projects, which try to predict the future job demand in the US.
Forecasting the Future of Work in Michigan
Contact: Niel Carlson
Perhaps could be combined with the previous one.
Church Buildings
Contact: Taylor Hartson
We have a set of data on converted church buildings (church buildings that have been re-adapted into restaurants, businesses, offices, etc; non-religious buildings that have been re-adapted into worship spaces) that could be visualized primarily geographically. We have addresses for converted structures in 21 cities across the U.S., as well as the denomination of the churches that reside in non-religious buildings. While we don’t have any concrete ideas for visualizations quite yet, the overall goal would be to discover where clusters or instances of each type of converted structure are located, visualize which denominations from our data set are most likely to adapt non-religious space for worship, and uncover any other meaningful spatial patterns.
Calvin Student Demographics
possible contact: Pennylyn Dykstra Pruim
idea: visualize changes in the Calvin student population over time.