Deep dive into clustering packages, R Shiny meets software dev & Posit::conf 23

The next Wellington R User group meeting is finally here!
We are happy to announce talks from Louise McMillan (Victoria University of Wellington) and Joann Zhang (Epi-interactive). There will be pizza, and soft drinks after the presentations; so stay on for some networking.
Everybody welcome.

WHEN:
Thursday, 16th November 2023, 5.00pm – 6.30pm

WHERE:
RH105 (1st floor – two floors up from the ground floor with the mezzanine floor is in between),
Rutherford House, Victoria University of Wellington,
33 Bunny Street, Pipitea

IMPORTANT:
Spaces are limited due to room restrictions. Please RSVP if you plan to attend the event; including any guest you might bring. If you register but you can't attend, please remove yourself from the event in Meetup (click "edit RSVP" – choose "Not going"). If you want to bring guests, please ask them to register as well as we need to provide an attendee list to the University who kindly offered to host the event.
Thank you!

Any questions? Just email the Epi-interactive team at events@epi-interactive.com

AGENDA:
5:00 – 5:10pm: Introductions and posit::conf 2023 highlights – Dr. Uli Muellner, Epi-interactive
5:10 – 5:35pm: clustglm and clustord: R packages for clustering with covariates for binary, count, and ordinal data – Louise McMillan, Victoria University of Wellington
5:35 – 6:00pm: My first year as full-time R Shiny developer – Experiences, learnings and pitfalls – Joann Zhang, Epi-interactive
6:00 – 6:30pm: Networking and pizzas shouted by Epi-interactive

ABSTRACTS:

clustglm and clustord: R packages for clustering with covariates for binary, count, and ordinal data
Louise McMillan, Victoria University of Wellington
We present two R packages for model-based clustering with covariates. Both packages can perform clustering and biclustering (clustering survey participants and survey questions simultaneously, for example). Both use likelihood-based methods for clustering, which enables users to compare models using AIC and BIC as measures of relative goodness of fit.
clustglm implements techniques from Pledger and Arnold (2014) for handling binary and count data, or data from other single-parameter exponential family distributions, such as normal distributions with constant variance. It leverages glm and can fit pattern detection models that include individual-level effects alongside cluster effects. For example, when applied to data about which questions participants answered in a survey, you can cluster participants and questions while also taking into account any single-participant effects, and any additional covariates such as demographic information about the participants. clustglm can also be applied to datasets from many other fields including ecological capture-recapture data or presence-absence data.
clustord handles ordinal categorical data, using techniques outlined in Fernández et al. (2016). It builds on the ordered stereotype model, which accommodates flexibility in the ordinal scale used. The clustering results can reveal when two ordinal categories are effectively equivalent and can be combined to simplify the model. It can also handle similar models to clustglm. If you have survey responses where each question is answered on a scale from 1 to 5, you can cluster participants and/or questions according to the pattern of the answers, and can include additional covariates such as demographic information about the participants.

My first year as full-time R Shiny developer – Experiences, learnings and pitfalls
Joann Zhang, Epi-interactive
Are you wondering how to transition from university into the work-force? In this presentation, I'll share my journey from a statistics undergraduate to computer science graduate and then to a Junior Software Engineer at Epi-interactive. I'll provide insights into my transition from uni to industry, the challenges I encountered while learning R Shiny, and the role R Shiny plays in bridging data analysis and software development. I'll also offer a glimpse into my experiences working within the Epi-interactive team, highlighting the collaborative environment and innovation that define my professional journey.