SOIL SPECTRAL INFERENCE WITH R analysing digital soil spectra.

Foreword
Preface
Acknowledgements
Endorsements
Contents
About the Authors
1 Introduction
	1.1 Spectroscopy in Soil Science
	1.2 Populating a Soil Database
	1.3 Objectives of This Book
	References
2 Getting Started with R
	2.1 Use of R and RStudio
	2.2 Simple Manipulations
	2.3 Data Structure
	2.4 Programming Tools
	2.5 Plotting
	2.6 Documentation and Help
	References
3 Materials
	3.1 Datasets
		3.1.1 The Geeves Dataset
		3.1.2 Soil Mineral Reference Spectra
		3.1.3 Soil Spectra and Colour
		3.1.4 Spectra for Standardization
		3.1.5 Spectra with Moisture
	3.2 R Packages
		3.2.1 Soil Science and Pedometrics
		3.2.2 Spectroscopy
		3.2.3 Modelling
		3.2.4 Plotting
	3.3 The soilspec Book Package
		3.3.1 Installing the Package
		3.3.2 Functions
		3.3.3 Datasets
	References
4 Data Handling of Spectra
	4.1 Importing Data
	4.2 Loading ASD Data
	4.3 Plotting the Spectra
	4.4 Averaging the Replicates
	4.5 Converting Units of Measurement
	4.6 Exporting the Spectra
	References
5 Pre-processing of Spectra
	5.1 Noise Removal
		5.1.1 Spectral Trimming
		5.1.2 Moving Window Average
		5.1.3 Savitzky-Golay Filtering
	5.2 Scatter Correction
		5.2.1 Standard Normal Variate
		5.2.2 Multiplicative Scatter Correction
		5.2.3 Detrending
	5.3 Derivatives
		5.3.1 First- and Second-Order Derivatives
	5.4 Centring and Standardizing
	5.5 Spectral or Dimension Reduction
		5.5.1 Resampling
		5.5.2 Wavelets
	5.6 Other Specific Transformations
		5.6.1 Splice Correction
		5.6.2 Continuum Removal
	References
6 Exploratory Soil Spectral Analysis
	6.1 Feature Selection
		6.1.1 Identifying Secondary Clay Minerals and Iron Oxides
		6.1.2 Comparing Soil Spectra with Spectra of Reference Materials
	6.2 Principal Component Analysis
	6.3 Spectral Prediction Domain
	6.4 Soil Colour
	References
7 Similarity Between Spectra and the Detection of Outliers
	7.1 Similarity/Dissimilarity Measures
		7.1.1 Euclidean Distance
		7.1.2 Mahalanobis Distance
		7.1.3 Correlation Similarity
		7.1.4 Spectral Angle Mapper
		7.1.5 Spectral Information Divergence
		7.1.6 A Practical Example
	7.2 Detecting Outlier Spectra
		7.2.1 Distance from the Average Spectrum
			The Mahalanobis Distance
			H Distance
		7.2.2 Multivariate Outliers
	References
8 Selection of the Samples for Laboratory Analysis
	8.1 Sampling Design
		8.1.1 Simple Random Sampling
		8.1.2 Kennard-Stone
		8.1.3 K-means Clustering
		8.1.4 Conditioned Latin Hypercube Sampling
		8.1.5 Presence of Outliers in the Data
	8.2 Sample Size
		8.2.1 Assessing the Representativeness of the Sample
		8.2.2 Optimal Sample Size Based on the Spectra
	References
9 Estimating Soil Properties and Classes from Spectra
	9.1 Goodness of Fit Measures
	9.2 Models for Quantitative Variables
		9.2.1 Principal Component Regression
		9.2.2 Partial Least Squares Regression
		9.2.3 Cubist
		9.2.4 Random Forest
		9.2.5 Memory-Based Learning
	9.3 Models for Categorical Variables
		9.3.1 Partial Least Squares Discriminant Analysis
		9.3.2 Random Forest
	9.4 Soil Spectral Inference Systems
	References
10 Spectral Transfer and Transformation
	10.1 Spectral Transfer Between Instruments  Using a Standard Sample
	10.2 Direct Standardization
	10.3 Piecewise Direct Standardization
	10.4 Removing External Effects, such as Soil Moisture (EPO)
	References