- Solving the Model Representation Problem With broom
Alex Hayes, an active contributor to broom (a tidyverse package), will be giving a presentation on how to use broom. Presentation abstract: The R objects used to represent model fits are notoriously inconsistent, making data analysis inconvenient and frustrating. The broom package resolves this issue by defining a consistent way to represent model fits. By summarizing essential information about fits in tidy tibbles, broom makes it easy to programmatically work with model objects. Combining broom with list-columns results in an especially powerful way to work with many model fits at once. This talk will feature several case studies demonstrating how broom resolves common problems in data analysis. I'll conclude with some thoughts on the future of broom and how broom might fit into an overall grammar of statistical modeling.
- Topic TBD
The topic is yet to be determined. Please fill out the survey to help determine what the topic(s) will be: https://www.surveymonkey.com/r/9R9CN7L Check the Today In The Union (TITU) listings the day of for the room (either at the Union or online at http://union.ems.wisc.edu/VirtualEMS/CustomBrowseEvents.aspx?data=MeN0rP%2fRPNoJS2WjBImM5NV%2b64hv0ZxW)
- Let's have an informal gathering/brainstorming over drinks
Hui - it's Maori for a meeting/exchange of ideas. Let's have an R hui where we can informally decide on a schedule / plan for future events. I'm thinking we could do workshops or kaggle hack nights in the new year, but at the very least, it's an excuse for drinks and merriment with good people.
- Introduction to Geoknife and USGS Vizlab
geoknife (Jordan Read) Downloading huge datasets for desktop processing can eat network bandwidth and also pose a challenge for local data storage and analysis. We created the geoknife R package to provide a reproducible way to subset or summarize large datasets to your area(s) of interest before they ever make it into a local processing environment. geoknife creates geo-processing requests that are carried out on a remote server, and lets you download the result or load it into R. In this presentation, I will demo some features of geoknife that may be of interest to users of climate or land-use data, including sampling gridded data relative to overlap with irregular features (e.g., research plots, watersheds, and states/ecoregions). These tasks can be computationally intensive and error-prone when it comes to dealing with multiple datasets in different coordinate reference systems, and geoknife handles this complexity for you. Jordan Read is the Chief of Data Science at the USGS's Office of Water Information, which develops frameworks for reproducible research and data visualization, as well as participating in efforts to integrate and centralize disparate datasets. Jordan approaches research questions that involve lake and stream simulations and sensor data analysis, including the challenges of modeling over ten thousand lakes across the US. Jordan received his PhD from the University of Wisconsin-Madison in 2012 in environmental fluid mechanics. USGS Vizlab (Jordan Walker) The USGS Vizlab team rapidly develops open-source, customized, web-based interactive data visualizations with USGS and external partners. We develop visualizations on timely, societally relevant topics, e.g. visualizations to accompany publication of journal articles or reports. Our team will access, analyze, and visualize data in plots, maps, and other visual products, and work with partners to generate accompanying text derived from underlying literature and authoritative scientific sources. In this presentation, I will share some of our visualizations to date and briefly cover the use of R in creating these visualizations. I will cover the basics of the package and discuss visualizations more generally. Discussed will be the techniques used in our visualization products, including SVG images, semantic web concepts, and reproducible workflows. I will also cover some of the challenges in producing high quality rapid visualizations for the web. Jordan Walker is a Data Scientist with the US Geological Survey. In his 7 years with the USGS, Jordan has helped develop numerous web applications and services for disseminating environmental data. With the Data Science team Jordan focuses on reproducibility and scalability in research and visualizations. Jordan graduated from the University of Wisconsin-Madison in 2010 with a Master's Degree in Computer Science.
- Analyzing Music with tuneR and seewave
In this meetup Yeng Chang will cover how we can use R to generate spectrograms for musical excerpts, what spectrograms can tell us about music, how music theory supports observations that we can find from spectrograms, and topics that should be explored for working with musical data. Even if you are not a musician this should be an interesting presentation with wider applications. Yeng is a data scientist at the Wisconsin Department of Public Instruction and a musician. He has a Bachelor's in Mathematics and is working on a Master's in Statistics from Iowa State University.
- R Programming Hack Night!
By popular demand we are going to have another hack night! At this one it might be fun to look at publicly available data and try to create a useful analysis or a shiny app from it. For example, the City of Madison has data at https://data.cityofmadison.com/. Of course, you can also work on your own project, help someone else with their project, or just hang out. People who have never tried R but would like to learn are welcome. Be sure to bring a laptop if you will need one. The library will have tables and chairs set up. Bring something to drink and eat if you want to.
- uRprogramming - Write an R Book Using Software Development Tools by Zekai Otles
Zekai Otles works for the State of Wisconsin at Employee Trust Funds as a Web Application Specialist. He has a Ph.D in Atmospheric Science from UW-Madison and worked on weather research models at Iowa State University for 5 years. Zekai has worked academic, private, not for profit and state government at different capacities. Started scientific programming with Fortan, unix script tools, java and perl programming for websites. He later involved with SAS and R programming for clinical trials data analysis. The extensive breadth of programming experience and fundamental understanding of code maintenance, automations and verifications are triggering him to compile his knowledge in a book. Zekai has created and presented DataFax (commercial software for clinical trials data management) reporting tools (dmRTools) a R package specific to Datafax in User group meeting. He has also contributed to the odfweave package. He occasionally reviews book for Manning publications, the noteworthy the book reviewed by him is "R in action" His goal to finish this book in 2017 and electronically publish it.
- R Programming Hack Night!
Let's get together for a hack night/socializing! At the hack night you can work your own project, help someone with their project, or just hang out. People who have never tried R but would like to learn are welcome. Be sure to bring a laptop if you will need one. The library will have tables and chairs set up. Bring something to drink and eat if you want to.
- SynergyScreen, an R Package for Design and Analysis of Compound Synergy Screens
Yury Bukhman, a scientist at the University of Wisconsin - Madison, will cover the development of the R package SynergyScreen as well as the use of S4 classes. About the SynergyScreen R package: Screening sets of chemical compounds for potential synergy or antagonism has a wide range of applications, from medicine to bioenergy research. In order to facilitate design and analysis of synergy screens, we have developed an R package, SynergyScreen. Given a set of compounds and dose ranges, SynergyScreen can produce a design of a screen testing their pairwise combinations. The package generates layouts for a set of 96-well microtiter plates. It includes titration series for each compound and each combination in a fixed ratio, comprising a set of single-ray synergy experiments. Once the experiments have been carried out, SynergyScreen can analyze the data to detect synergistic and antagonistic compound pairs. The analysis includes normalization to remove potential plate bias, modeling each individual dose-response curve, and computing interaction index values for a set of effect sizes. The package can produce tabular outputs and visualizations.