Geostatistics – II

By |

In many situations, we observe a process over a limited number of locations and are interested in estimating its values at nearby unsampled locations. Geostatistical methods provide a powerful tool to do this. In this session, we will build upon the material and intuitions covered in the first workshop. We will focus on the principles, application, and interpretation of results using R, and will cover topics such as variogram modeling, empirical variogram, estimation techniques, and prediction uncertainty.

You should be familiar with R and the basic concepts of GIS data, should have taken an introductory course in statistics, and know the basics of linear algebra.

Mapping tools for COVID-19 data – II

By |

This is the second workshop in this series and will build on the material and the techniques covered in the first workshop. We will use COVID-19 data available here https://github.com/nytimes/covid-19-data, combine it with other data such as from Census and will learn about creating maps, including web maps, that involve more than one variable.

We will use R and QGIS. You should know the basics of vector data GIS and introductory techniques of classification and clustering, be a regular R user, and should have some familiarity with QGIS.

R IV: Visualization

By |
This session will be held online, and presenters will be in touch with more information after you register.
Visualizing the results of research is a key aspect in facilitating scientific communication to a broad audience. The focus of this workshop will be on using common tools in R for visualization, general concerns in producing visual information, and demonstration of approaches that allow for interactivity.
Prereq: Some experience using R is required (R I, preferably R II workshops), as well as exposure to basic statistical analysis would be beneficial.

R by Example: Analyzing RECS using data.table

By |

In the R by Example series of workshops, we’ll discuss example analyses in R as a vehicle for learning  commonly used tools and programming patterns.  The “Analyzing RECS using data.table” workshop will focus on analyzing winter home temperatures in the US using data from the Residential Energy Consumption Survey (https://www.eia.gov/consumption/residential/).  We’ll use the data.table package for data manipulations and ggplot2 for plotting.  The workshop will be organized in a parallel fashion, with participants given time to build an analysis from scratch by adapting presented examples step by step. In the process, participants will become familiar with core data.table functionality including its pivot methods.  This workshop is geared towards beginner to intermediate R users or those new to data.table.

R by Example: Analyzing RECS using tidyverse

By |

In the R by Example series of workshops, we’ll discuss example analyses in R as a vehicle for learning  commonly used tools and programming patterns.  The “Analyzing RECS using tidyverse” workshop will focus on analyzing winter home temperatures in the US using data from the Residential Energy Consumption Survey (https://www.eia.gov/consumption/residential/).  We’ll use the tidyverse (tidyverse.org) throughout, relying on the dplyr package for data manipulations and ggplot2 for plotting.  The workshop will be organized in a parallel fashion, with participants given time to build an analysis from scratch by adapting presented examples step by step. In the process, participants will become familiar with core dplyr functions, pivoting using tidyr, and a basic ggplot2 example.  This workshop is geared towards beginner to intermediate R users.

Rcpp: Integrating C++ into R

By |

The Rcpp package for R provides “seamless R and C++ integration”.  In this workshop, we will discuss the use of Rcpp to speed up existing R code by rewriting slow functions in C++.  

The workshop will be centered around a couple of case studies with an opportunity provided for participants to implement a few of their own C++ functions, compile, and call them from R.  Participants should be comfortable programming in R, but need not have any prior exposure to C++.

Programming with R

By |

People using R for applied research are often not taught basic programming practices such as writing functions, efficient iterative processing, vectorization, and other practices that would make their research far more efficient and reproducible.  Understandably, focus is on basic data manipulation and getting model results.  Unfortunately, this can mean the data isn’t as explored as it should be, or other opportunities are lost (e.g. feature engineering), because of the presumed effort that would be required to deal with the data more fully.

This workshop will help you get more out of R so that you can take your efforts to the next level.
Prereq: Some basic experience using R is required.  You should know how to create and manipulate objects, run basic analyses, etc.  This could also be useful to anyone with programming experience in another language like Python.

Mediation Models: A demonstration using multiple packages

By |

Mediation models are commonly applied in a variety of modeling settings, and people will typically flock to tools specific to structural equation modeling like Mplus or Amos for analysis.  However, not only are such tools not necessary for the more common implementations of mediation, they are often limiting and have various drawbacks.

Fortunately there are a variety of packages in R that can do mediation analysis, often using straightforward code and familiar models or other tools.  This presentation will demonstrate a variety of ways in which to do a standard mediation model in R (and Python), and discuss the available complexities that can be handled with the tools, as well as their corresponding strengths and weaknesses.
Note that this is not an introduction to mediation analysis, but is a demonstration of tools.  Some familiarity with R and mediation models will be assumed.

Back to a Future: Asynchronous Computing with futures in R

By |

Asynchronous computing is an umbrella term encompassing parallel and concurrent computational programs in which some tasks can be executed without a strict sequential order.  future is a programming abstraction for a value that may be available at some future point in time and allows.  Like other forms of parallelism, futures are a powerful tool for writing programs that efficiently make use of available computing resources.  At the same time, futures can also be used to make interactive data analyses more time efficient. 

 In this workshop, we’ll discuss futures as implemented in the R package “future” and provide example use cases for both interactive analysis and batch processing.  

Open Source GIS

By |

This workshop will provide a gentle introduction to open source GIS tools in R and QGIS. We will cover introductory GIS concepts and will explore the functionalities of R and QGIS for manipulating and analyzing vector GIS data. Familiarity with R is required.