Statistical computing for environmental science with R and RStudio – Online Edition

Home › Projects › Statistical computing for environmental science with R and RStudio – Online Edition

Statistical computing for environmental science with R and RStudio – Online Edition

This course cover some advanced issues in most statistical computing workflow for Environmental Science like data structure and workspace management, visualization techniques and statistical analysis using the free platforms and programming language R and Rstudio.

The course is given in four modules covering exploratory data analysis, univariate and multivariate statistical techniques and a final discussion session where students  work and discuss their projects

Image from Transmitting Science

Module I. Programming in R and Rstudio

  • A review of the most commonly used files in any statistical programming workflow like the scripts (.R), the allocated memory (.RData), and the novel Rstudio projects (.Rproj) were given for establishing dedicated working directories, workspace, history, and source documents.

Module II. Exploratory Data Analysis (EDA)

  • During any statistical programming workflow, almost half of the time must be given to data exploration. However, not every exploration is a valid exercise. In this module, we will review the most common assumption in a classical statistical analysis like normality, heterogeneity, and independence in the data.

Module III. Univariate Statistical Analysis (UniStat)

  • In module III, the classic univariate approach for statistics like the linear models and their extensions is reviewed. The most common analysis like ANOVA, ANCOVA, or Regression analysis is also covered. For those common cases in ecology, where the linear models fail (e.g. non-negative data in count/abundance data) we will present some extensions covering the Generalized Linear Models (GLM) and we will review some insights about Generalized Additive Models (GAM).

Module IV. Multivariate Statistical Analysis (MultiStat)

  • Along with this fourth module,  the analysis of multivariate data is covered. We will review two different approaches to understand community (multivariate) data based on different ordination techniques. Thus, two approaches from unconstrained ordination like Principal Component Analysis (PCA) and non-Metric Multidimensional Scaling (nMDS) will help us to reveal patterns along with our community data. Finally, we will use two approaches from constrained ordination like Redundancy Analysis (RDA) and Canonical Correspondence Analysis (CCA) to understand the role of some environmental (explanatory) variables over the community structure in our community data.

The first edition of this course was held in August 2020. It was taught online by Antonio Canepa (Ph.D., OMDS team, @CanepaOneto) and coordinated by  Dr. Soledad De Esteban-Trivigno (Transmitting Science).

 

+ CONTENTS

Module I. Programming in R and Rstudio

  • A review of the most commonly used files in any statistical programming workflow like the scripts (.R), the allocated memory (.RData), and the novel Rstudio projects (.Rproj) were given for establishing dedicated working directories, workspace, history, and source documents.

Module II. Exploratory Data Analysis (EDA)

  • During any statistical programming workflow, almost half of the time must be given to data exploration. However, not every exploration is a valid exercise. In this module, we will review the most common assumption in a classical statistical analysis like normality, heterogeneity, and independence in the data.

Module III. Univariate Statistical Analysis (UniStat)

  • In module III, the classic univariate approach for statistics like the linear models and their extensions is reviewed. The most common analysis like ANOVA, ANCOVA, or Regression analysis is also covered. For those common cases in ecology, where the linear models fail (e.g. non-negative data in count/abundance data) we will present some extensions covering the Generalized Linear Models (GLM) and we will review some insights about Generalized Additive Models (GAM).

Module IV. Multivariate Statistical Analysis (MultiStat)

  • Along with this fourth module,  the analysis of multivariate data is covered. We will review two different approaches to understand community (multivariate) data based on different ordination techniques. Thus, two approaches from unconstrained ordination like Principal Component Analysis (PCA) and non-Metric Multidimensional Scaling (nMDS) will help us to reveal patterns along with our community data. Finally, we will use two approaches from constrained ordination like Redundancy Analysis (RDA) and Canonical Correspondence Analysis (CCA) to understand the role of some environmental (explanatory) variables over the community structure in our community data.

+ Previous EDITIONS

The first edition of this course was held in August 2020. It was taught online by Antonio Canepa (Ph.D., OMDS team, @CanepaOneto) and coordinated by  Dr. Soledad De Esteban-Trivigno (Transmitting Science).

 

Do you need this, or a similar course?

Share: