QCBS R Workshop Series
Preface
0.1
Code of conduct
0.1.1
Expected behaviour
0.1.2
Unacceptable behaviour
0.2
Contributors
0.3
Contributing
Multivariate Analyses in
R
1
Learning objectives
2
Preparing for the workshop
PREAMBLE
3
Recap: Univariate analyses
4
Intro: Multivariate analyses
5
Setting up our goals
6
Matrix algebra, very briefly
6.1
Data sets
are
matrices
6.2
Association matrices
7
Exploring the dataset
7.1
Doubs river fish communities
7.2
Doubs river environmental data
(Dis)similarity measures
8
dist()
9
Types of distance coefficients
9.1
Metric distances
9.1.1
Euclidean distances
9.1.2
Challenge #1
9.1.3
Chord distances
9.1.4
Jaccard’s coefficient
9.2
Semimetric distances
9.2.1
Sørensen’s coefficient
9.2.2
Bray-Curtis’ coefficient
9.3
Nonmetric distances
9.3.1
Mahalanobis distance
9.4
Representing distance matrices
10
Transformations
10.1
Presence-absence transformation
10.2
Species profiles transformation
10.3
Hellinger transformation
10.4
Z-score standardization
10.4.1
Other association metrics
Clustering
11
Clustering
11.1
Single linkage agglomerative clustering
11.2
Complete linkage agglomerative clustering
11.3
Unweighted Pair Group Method with Arithmetic Mean (UPGMA)
11.4
Weighted Pair Group Method with Arithmetic Mean (WPGMA)
11.5
Ward’s minimum variance
11.6
Deciding on cut-off points
11.7
Playing with real data: the Doubs fish species data
Unconstrained Ordination
12
What does “unconstrained” mean?
13
Principal Component Analysis
13.1
Principal component analysis
not
in a nutshell
13.2
Principal component analysis using package functions
13.3
Principal component analysis on ecological data
13.4
Condensing data with principal component analysis
13.5
Scaling
13.6
Challenge #2
14
Correspondence Analysis
15
Principal Coordinates Analysis
16
Nonmetric MultiDimensional Scaling
Final considerations
17
Summary
18
Additional resources
19
References
QCBS R Workshop Series
Workshop 9: Multivariate Analyses in
R
Chapter 18
Additional resources