A Beginner's Guide to Data Exploration and Visualization with R
Published on 2. March 2015
Book
Paperback/Softback
174 pages
978-0-9571741-7-7 (ISBN)
Description
This book uses ecological datasets to discuss data exploration and visualisation tools. The authors also explain how to visualise the results of statistical models, an important aspect for publishing scientific papers. The book includes the R code needed to construct, visualise, and explore the main features of the data step by step.
More details
Series
Language
English
Place of publication
Newburgh
United Kingdom
Target group
College/higher education
Dimensions
Height: 233 mm
Width: 156 mm
Weight
400 gr
ISBN-13
978-0-9571741-7-7 (9780957174177)
Copyright in bibliographic data and cover images is held by Nielsen Book Services Limited or by the publishers or by their respective licensors: all rights reserved.
Schweitzer Classification
Persons
Elena N Ieno is senior marine biologist at Highland Statistics Ltd. She is a co-author of seven books on the analysis of ecological data. She teaches data analysis to ecologists and environmental scientists worldwide and is adept at bridging the gap between the two disciplines to dispel the fear of statistics. Alain F Zuur is senior statistician and director of Highland Statistics Ltd., a statistics consultancy based in the UK. He is the author of seven books on the analysis of ecological data. He has extensive experience teaching statistical methods to ecologists and environmental scientists in academic and non-academic courses worldwide.
Content
PREFACE V ACKNOWLEDGEMENTS V DATASETS USED IN THIS BOOK V 1 INTRODUCTION 1 1.1 SPEAKING THE SAME LANGUAGE 1 1.2. GENERAL POINTS 2 1.3 OUTLINE OF THIS BOOK 5 2 OUTLIERS 7 2.1 WHAT IS AN OUTLIER? 7 2.2 BOXPLOT TO IDENTIFY OUTLIERS IN ONE DIMENSION 8 2.2.1 Simple boxplot 8 2.2.2 Conditional boxplot 10 2.2.3 Multi-panel boxplots from the lattice package 13 2.3 CLEVELAND DOTPLOT TO IDENTIFY OUTLIERS 15 2.3.1 Simple Cleveland dotplots 15 2.3.2 Conditional Cleveland dotplots 17 2.3.3 Multi-panel Cleveland dotplots from the lattice package 18 2.4 BOXPLOTS OR CLEVELAND DOTPLOTS? 20 2.5 CAN WE APPLY A TEST FOR OUTLIERS? 21 2.5.1 Z-score 22 2.5.2 Grubbs' test 22 2.6 OUTLIERS IN THE TWO-DIMENSIONAL SPACE 24 2.7 INFLUENTIAL OBSERVATIONS IN REGRESSION MODELS 25 2.8 WHAT TO DO IF YOU DETECT POTENTIAL OUTLIERS 27 2.9 OUTLIERS AND MULTIVARIATE DATA 31 2.10 THE PROS AND CONS OF TRANSFORMATIONS 33 3 NORMALITY AND HOMOGENEITY 37 3.1 WHAT IS NORMALITY? 37 3.2 HISTOGRAMS AND CONDITIONAL HISTOGRAMS 38 3.2.1 Multipanel histograms from the lattice package 39 3.2.2 When is normality of the raw data considered? 41 3.3 KERNEL DENSITY PLOTS 42 3.4 QUANTILE - QUANTILE PLOTS 43 3.4.1 Quantile - quantile plots from the lattice package 44 3.5 USING TESTS TO CHECK FOR NORMALITY 45 3.6 HOMOGENEITY OF VARIANCE 47 3.6.1 Conditional boxplots 47 3.6.2 Scatterplots for continuous explanatory variables 49 3.7 USING TESTS TO CHECK FOR HOMOGENEITY 50 3.7.1 The Bartlett test 50 3.7.2 The F-ratio test 50 3.7.3 Levene's test 51 3.7.4 So which test would you choose? 51 3.7.5 R code 51 3.7.6 Using graphs? 52 4 RELATIONSHIPS 55 4.1 SIMPLE SCATTERPLOTS 55 4.1.1 Example: Clam data 55 4.1.2 Example: Rabbit data 57 4.1.3 Example: Blow fly data 58 4.2 MULTIPANEL SCATTERPLOTS 60 4.2.1 Example: Polychaeta data 60 4.2.2 Example: Bioluminescence data 61 4.3 PAIRPLOTS 62 4.3.1 Bioluminescence data 63 4.3.2 Cephalopod data 64 4.3.3 Zoobenthos data 65 4.4 CAN WE INCLUDE INTERACTIONS? 66 4.4.1 Irish pH data 66 4.4.2 Godwit data 68 4.4.3 Irish pH data revisited 70 4.4.4 Parasite data 71 4.5 DESIGN AND INTERACTION PLOTS 73 5 COLLINEARITY AND CONFOUNDING 77 5.1 WHAT IS COLLINEARITY? 77 5.2 THE SAMPLE CORRELATION COEFFICIENT 77 5.3 CORRELATION AND OUTLIERS 78 5.4 CORRELATION MATRICES 79 5.5 CORRELATION AND PAIRPLOTS 80 5.6 COLLINEARITY DUE TO INTERACTIONS 82 5.7 VISUALISING COLLINEARITY WITH CONDITIONAL BOXPLOTS 83 5.8 QUANTIFYING COLLINEARITY USING VIFS 85 5.8.1 Variance inflation factors 85 5.8.2 Geometric presentation of collinearity 86 5.8.3 Tolerance 88 5.8.4 What constitutes a high VIF value? 88 5.8.5 VIFs in action 89 5.9 GENERALISED VIF VALUES 91 5.10 VISUALISING COLLINEARITY USING PCA BIPLOT 93 5.11 CAUSES OF COLLINEARITY AND SOLUTIONS 94 5.12 BE STUBBORN AND KEEP COLLINEAR COVARIATES? 96 5.13 CONFOUNDING VARIABLES 97 5.13.1 Visualising confounding variables 99 5.13.2 Confounding factors in time series analysis 100 6 CASE STUDY: METHANE FLUXES 103 6.1 INTRODUCTION 103 6.2 DATA EXPLORATION 104 6.2.1 Where in the world are the sites? 104 6.2.2 Working with ggplot2 105 6.2.3 Outliers 108 6.2.4 Collinearity 111 6.2.5 Relationships 112 6.2.6 Interactions 114 6.2.7 Where in the world are the sites (continued)? 115 6.3 STATISTICAL ANALYSIS USING LINEAR REGRESSION 118 6.3.1 Model formulation 118 6.3.2 Fitting a linear regression model 118 6.3.3 Model validation of the linear regression model 120 6.3.4 Interpretation of the linear regression model 125 6.4 STATISTICAL ANALYSIS USING A MIXED EFFECTS MODEL 131 6.4.1 Model formulation 131 6.4.2 Fitting a mixed effects model 132 6.4.3 Model validation of the mixed effects model 132 6.4.4 Interpretation of the linear mixed effects model 132 6.5 CONCLUSIONS 134 6.6 WHAT TO PRESENT IN A PAPER 134 7 CASE STUDY: OYSTERCATCHER SHELL LENGTH 135 7.1 IMPORTING THE DATA 136 7.2 DATA EXPLORATION 136 7.3 APPLYING A LINEAR REGRESSION MODEL 138 7.4 UNDERSTANDING THE RESULTS 140 7.5 TROUBLE 143 7.6 CONCLUSIONS 146 8 CASE STUDY: HAWAIIAN BIRD TIME SERIES 147 8.1 IMPORTING THE DATA 147 8.2 CODING THE DATA 148 8.3 MULTI-PANEL GRAPH USING XYPLOT FROM LATTICE 148 8.3.1 Attempt 1 using xyplot 149 8.3.2 Attempt 2 using xyplot 150 8.3.3 Attempt 3 using xyplot 151 8.4 MULTI-PANEL GRAPH USING GGPLOT2 153 8.5 CONCLUSIONS 154 REFERENCES 155 INDEX 159 BOOKS BY HIGHLAND STATISTICS 161