Workshops Starting at 1:00 pm
Navigating Open Science: Pathways to transparency, replication, and access through scientific publishing: 1:00 pm
Presenter: Natalie Sopinka
Science is increasingly moving towards openness: open data, open peer review, and open access publication. This workshop presented by Canadian Science Publishing, a proud partner of the Canadian Society for Ecology & Evolution, aims to demystify what it means to practice open science in the modern era. Workshop participants will be invited to take an active part in breakout activities to better understand components of open science, including how to follow funder requirements, better understand compliance routes, discover avenues for data deposition, and discuss how open practices can make an impact beyond the scientific community and reach global audiences.
Developing a reproducible workflow in R using functions, {targets} and {renv}: 1:00 pm
Presenter: Alec Robitaille
This workshop will cover key elements of reproducible workflows targeted at users with familiarity with the R programming language. The workshop will start with project management and move to function writing in R that is flexible and testable. Next attendees will combine data and functions to create a reproducible workflow using the R packages {targets}, {renv}, and {conflicted}. Workshop attendees will gain fundamental project management and workflow skills that they can apply to their own projects and take back to their labs and collaborators. Project management that develops cohesive and well structured projects from the start promotes easy sharing, publishing and understanding by other users. Building and using functions allows for the creation of independent, logical steps which are easily tested and extended to different contexts. Ecology and evolution analyses in particular are often complicated with many interconnected steps and diverse data types. This combination makes good project management, function writing skills, and reproducible workflows all the more important. An exciting part of this workshop is the introduction to the R package {targets}. Many ecology and evolution analyses have modifications throughout the scientific process, which results in analytical steps that need to be rerun. {targets} tracks dependencies between input files, analysis steps and outputs and as you modify your project, it only reruns the relevant pieces. {targets} is an effective way to produce and manage a reproducible workflow, and is one of the first pipeline tools designed specifically for R, the main coding language of many CSEE attendees. As a relatively new tool, not many R users have taken advantage of it yet and this workshop provides an opportunity to get familiar.
From Absolute Zero to Coding Hero: An R Crash Course in Reproducible Research for Community Ecology: 1:00 pm
Presenter: Robert Colautti
The Absolute Zero to Coding Hero (AZ2CH) workshops are created by, and for, biologists. We assume no prior coding knowledge as participants are guided from the most fundamental syntax and coding concepts up to publication-ready analyses and data visualizations, with all tutorials provided free online for review during and after the workshop. In this workshop, we introduce coding in R with R Studio for the analysis of species communities, from large collaborative field surveys down to microbial communities reconstructed from high-throughput sequencing. This workshop is targeted at trainees and active researchers with little, if any, prior coding experience. Students working on original species community data are particularly encouraged to attend, but all are welcome and we will provide datasets to guide participants through a complete workflow for organizing, inspecting, restructuring, analyzing, and visualizing ecological data in R with R Studio. Based on the free and open textbook “R Crash Course for Biologists”, we will quickly introduce participants to key syntax and concepts for programming in R with R Studio, and using R markdown, with knitr, ggplot2, and dplyr for open and reproducible research with publication-ready graphics. After introducing these core skills, the second half of the course focuses on approaches for normalizing micro- and macro-ecology datasets, measuring species richness and evenness (alpha diversity), and constructing dissimilarity matrices to visualize and analyze community differences (beta diversity). This workshop will also emphasize reproducibility in ecological research, introducing the baRcodeR package for generating unique sample identifiers with scannable barcodes, and demonstrating how to build open and collaborative workflows in R Studio with Git and GitHub. By the end of the workshop, participants will gain conceptual and practical coding skills to analyze community ecology data and conduct open, reproducible, and data-driven ecological research.
Connecting theory to data: Introduction to Bayesian statistics for ecologists: 1:00 pm
Presenter: Andrew MacDonald
Have you ever found yourself with a detailed ecological dataset, an hypothesis you're excited to test -- but struggled to find a statistical procedure to connect the two? Bayesian statistics are exciting for ecologists because they offer us the ability to design a model that meets the full complexity of our dataset and our ecological ideas. An amazing array of models can be built, fit, and assessed using the same flexible toolkit. This workshop will cover the basics of applied Bayesian modelling for ecology. We will discuss some of the theory behind the Bayesian framework, and explore some common models using simulated. We'll learn how to fit, interpret, diagnose and visualize a Bayesian model using plots, and study a wide variaty of models useful to ecologists. These will include generalized linear models for count data and presence-absence data, as well as hierarchical, nonlinear, and zero-inflated models. All models will be illustrated using ecological examples and published datasets. We'll explore these model types using a combination of brms, rstanarm (two leading R packages) as well as some Stan code. Some prior experience in R is encouraged, but none with statistics nor with Bayesian approaches specifically.
This workshop is sponsored by the Canadian Institute of Ecology and Evolution - Institut canadien d’écologie et d’évolution.