Math Geosci (2021) 53:823–858

https://doi.org/10.1007/s11004-020-09895-w

Classical and Robust Regression Analysis with

Compositional Data

K. G. van den Boogaart1 · P. Filzmoser2 ·

K. Hron3 · M. Templ4 · R. Tolosana-Delgado1

Received: 15 November 2019 / Accepted: 14 September 2020 / Published online: 6 October 2020
© The Author(s) 2020

Abstract Compositional data carry their relevant information in the relationships

(logratios) between the compositional parts. It is shown how this source of information
can be used in regression modeling, where the composition could either form the
response, or the explanatory part, or even both. An essential step to set up a regression
model is the way how the composition(s) enter the model. Here, balance coordinates
will be constructed that support an interpretation of the regression coefficients and
allow for testing hypotheses of subcompositional independence. Both classical least-
squares regression and robust MM regression are treated, and they are compared within
different regression models at a real data set from a geochemical mapping project.

B K. G. van den Boogaart

[email protected]
P. Filzmoser
[email protected]
K. Hron
[email protected]
M. Templ
[email protected]
R. Tolosana-Delgado
[email protected]

1 Helmholtz Institut Freiberg for Resources Technology, Freiberg, Germany

2 Institute of Statistics and Mathematical Methods in Economics, Vienna University of
Technology, Vienna, Austria
3 Department of Mathematical Analysis and Applications of Mathematics, Faculty of Science,
Palacký University, Olomouc, Czech Republic
4 Institute for Data Analysis and Process Design, Zurich University of Applied Sciences, Winterthur,
Switzerland

123
824 Math Geosci (2021) 53:823–858

Keywords Balances · Robust regression · GEMAS project · Hypothesis testing ·

Robust bootstrap

1 Introduction

Although regression analysis belongs to the most developed statistical procedures,

there is not too much literature available if compositional data are involved (see,
e.g., Aitchison 1986; Daunis-i-Estadella et al. 2002; Tolosana-Delgado and van den
Boogaart 2011; van den Boogaart and Tolosana-Delgado 2013; Egozcue et al. 2013;
Pawlowsky-Glahn et al. 2015; Fišerová et al. 2016; Coenders et al. 2017; Filzmoser
et al. 2018; Greenacre 2019). Regression with compositional data presents certain
particularities because of the special statistical scale of compositions. Compositions
have been typically (and restrictively) defined as vectors of positive components sum-
ming up to a constant, with the simplex as their sampling space (Aitchison 1986).
However, since the beginning of the twenty-first century, it has become clearer that
data may not abide to the constant sum condition and nevertheless be sensibly con-
sidered as compositional, the determining point being the relativity of the information
provided (Aitchison 1997; Barceló-Vidal et al. 2001). That means, when dealing with
compositional data the crucial quantities are formed by the ratios between the compo-
sitional parts (i.e., the variables) rather than by the reported data values directly. There
are different proposals to extract this relative information, frequently referred to as
transformations or coordinates, with the additive logratio (alr), the centered logratio
(clr), and the isometric logratio (ilr) transformation as the most well-known represen-
tatives (van den Boogaart and Tolosana-Delgado 2013; Pawlowsky-Glahn et al. 2015;
Filzmoser et al. 2018). The scores obtained with such transformations can be referred
to as “coordinates” thanks to the Euclidean space structure of the simplex (Billheimer
et al. 2001; Pawlowsky-Glahn and Egozcue 2001a), described later on in Sect. 2.
Due to its affine equivariance, linear regression is known to provide exactly the same
result whichever logratio transformation is used (van den Boogaart and Tolosana-
Delgado 2013), although the use of the clr transformation may entail unnecessary
numerical complications (the generalised inversion of a singular covariance matrix)
whenever the composition plays an explanatory role. Even more, regression models
with compositional data can be established without using any logratio transformation
at all. Indeed, this fact allows to consider the objects occurring in the regression
analysis (slopes, intercepts, gradients, residuals, predictions, etc) as fully intrinsically
compositional objects, the use of one or another logratio transformation being a choice
of representation. This paper just works with an isometric logratio representation
because of its intimate connection with tests for exclusion of single components or
subcompositions.
Generally, one must distinguish different types of regression models involving com-
positional data. The composition could form the response part in the regression model,
and the explanatory part is consisting of one or more non-compositional features. This
leads to a multivariate regression model, later on denoted as Type 1 model. The reverse
problem, composition as explanatory part, non-compositional response, is denoted as
Type 2 model in this contribution; if the response is only univariate, we end up with a

123
Math Geosci (2021) 53:823–858 825

multiple linear regression model. The Type 3 model is characterized by a composition

as explanatory part and another composition as response part, thus becoming a multi-
variate multiple linear regression model. For each considered type, we can construct
the regression model exploiting the Euclidean structure of the simplex, or else in terms
of ilr coordinates or clr transformed scores. However, since there are infinitely many
possibilities to construct ilr coordinates (Egozcue et al. 2003), one can focus on those
coordinates that allow for an interpretation of the model and the corresponding regres-
sion coefficients. Even more, if particular hypotheses shall be tested, it is essential to
construct such coordinates that support the testing problem. This contribution focuses
particularly in so called tests of subcompositional independence, or rather uncorrela-
tion. Here the goal is to establish subcompositions (or single components) that do not
depend on or do not influence the covariables considered.
A further issue which is also relevant for traditional non-compositional regression
is parameter estimation. The most widely used method is least-squares regression,
minimizing the sum of squared residuals. In presence of outliers in the explanatory
or/and response variable(s) it is known that this estimator can be heavily biased, and
robust counterparts should be preferred (Maronna et al. 2006). Here, the highly robust
MM estimator for regression is considered, for both the multiple and the multivariate
linear regression model (Van Aelst and Willems 2013). There is also technique avail-
able for robust statistical inference (Salibian-Barrera et al. 2008). Such techniques
are of the utmost importance in compositional data, because here small values (near
or below determination limits, rounded zeros, counting zeros) can otherwise become
highly influential after the logratio transformations.
In this contribution we thoroughly define the different types of regression models
involving compositional data, with geometric interpretations, and explain their use in
a case study. In Sect. 2 we provide more detailed information about compositional data
analysis, and define the three considered types of models. Section 3 refers to least-
squares estimation, and depending on the model it is shown how parameter estimation
and hypothesis testing can be carried out. Estimation and hypothesis testing using
robust MM regression is treated in Sect. 4. All three linear models and classical as
well as robust estimation are illustrated in Sect. 5 with a data set originating from the
GEMAS project, a continental scale soil geochemistry survey in Europe. The final
Sect. 6 summarizes and concludes.

2 Compositional Data Analysis

2.1 Compositional Geometry

A (column-)vector x = [x1 , x2 , . . . , x D ] is considered a D-part composition if its

components do inform on the relative importance of a set of parts forming a total.
Because of this, compositions are often identified with vectors of positive components
and constant or irrelevant sum, i.e. vectors that can be reclosed to sum up to 1 (or to
any other constant) by C [x] = (1t · x)−1 x, without loss of relevant information. In
the eighties of the twentieth century Aitchison (1986) proposed to statistically treat
this kind of data after some sort of one-to-one multivariate logratio transformation,

123
826 Math Geosci (2021) 53:823–858

generalising the logit, and he defined several of them, most notably the centered
logratio (clr) transformation (Aitchison 1986)

x  
clr(x) = ln 
D
 , inverse: clr −1 (x∗ ) = C [exp(x∗ )] ,
xi

with logs and exponentials applied component-wise. The sample space of a random
composition, the simplex S D , has been since the end of the nineties recognized
to have a vector space structure, induced by the operations of perturbation x ⊕ y =
C [x1 y1 , x2 y2 , . . . , x D y D ] (Aitchison 1986) and powering λx = C [x1λ , x2λ , . . . , x D
λ]

(Aitchison 1997). The neutral element with respect to this structure is proportional to a
vector of D ones, n = C [1 D ]. This structure can be extended (Pawlowsky-Glahn and
Egozcue 2001b) to a Euclidean space structure (Aitchison et al. 2002) by the scalar
product

1   xi
D D
yi
x, y A = ln ln = clr(x), clr(y) , (1)
2D xj yj
i=1 j=1

which induces a compositional distance (Aitchison 1986)

D D  
1  xi yi 2
d 2A (x, y) = ln − ln , (2)
2D xj yj
i=1 j=1

both fully compliant with the concept of relative importance conveyed in the modern
definition of compositions.
To take this relative character into account in an easy fashion when statistically
analyzing compositional data sets, the principle of working on coordinates is recom-
mended (Pawlowsky-Glahn 2003). This states that the statistical analysis should be
applied to the coordinates of the compositional observations, preferably in an orthonor-
mal basis of the Euclidean structure {S D , ⊕, , ·, · A }, and that results might be
expressed back as compositions by applying them to the basis used. A simple and
easy way to compute orthonormal coordinates is provided by the isometric log-ratio
(ilr) transformation (Egozcue et al. 2003)

ilr(x) =: x∗ = Vt · ln x = Vt · clr(x), (3)

where V = (v1 , v2 , . . . , v D−1 ) is a matrix with orthonormal columns, each orthogonal

to the vector of ones, i.e.

vit · v j = δi j and 1tD · vi = 0. (4)

Each of these columns provide the clr coefficients of each of the compositional vectors
forming the orthonormal basis used, so that the orthonormal basis on the simplex (with
respect to the Aitchison geometry) is wi = clr −1 (vi ). Conversely, given a vector of

123
Math Geosci (2021) 53:823–858 827

coordinates x∗ = [x1∗ , x2∗ , . . . , x D−1

∗ ], the inverse isometric transformation provides a
convenient way to apply it to the basis

D−1
xi∗  wi =: ilr −1 (x∗ ) = C [exp(V · x∗ )], (5)
i=1

and the clr coefficients can be recovered as well from the ilr coordinates with

clr(x) = V · ilr(x). (6)

There are infinitely many ilr transformations, actually as many as matrices satisfying
the conditions of Eq. (4). Each represents a rotation of the coordinate system (see, e.g.,
Egozcue et al. 2003; Filzmoser et al. 2018). For the purposes of this contribution, it is
relevant to consider those matrices linked to particular subcompositions, i.e. subsets
of components. If one wants to split the parts in x into two groups, say the first s
against the last r = D − s, there is a vector that identifies the balance between the
two groups,
⎡ ⎤
1 ⎣r, r, . . . , r , −s, −s, . . . , −s ⎦ .
v= √ (7)
r s + sr 2
2    
s r

Given that one can always split the resulting two subcompositions again into two
groups, this sort of structure can be reproduced D − 1 times, generating D − 1 vectors
of this kind with only three values every time (a positive value for one group, a
negative value for the other group and zero for those parts not involved in that particular
splitting). This is called a sequential binary partition, and the interested reader may
find the details of the procedure in Egozcue and Pawlowsky-Glahn (2005, 2011).
A particular case occurs when the splitting of interest places one individual variable
against all other (a sort of “one component subcomposition”). Considering the first
component as the one that is desired to be isolated, the balancing vector is then obtained
with Eq. (7) taking r = 1 and s = D − 1. Balances isolating any other part can be
obtained permuting the resulting components. This sort of balances, which can be
identified with so-called pivot coordinates (Filzmoser et al. 2018), are thus useful also
to check the possibility to eliminate one single part from the regression model.

2.2 Compositional Linear Models

Three kinds of linear models involving compositions have been defined (van den
Boogaart and Tolosana-Delgado 2013; Pawlowsky-Glahn et al. 2015; Filzmoser et al.
2018): models with compositional response (Aitchison 1986; Daunis-i-Estadella et al.
2002), models with compositional explanatory variable (Aitchison 1986; Tolosana-
Delgado and van den Boogaart 2011), and models with compositions as both
explanatory variable and response (Egozcue et al. 2013). The next sections systemati-
cally build each regression model solely by means of geometric operations, and show

123
828 Math Geosci (2021) 53:823–858

how these models are then represented in an arbitrary isometric logratio transforma-
tion.

2.2.1 Linear Model with Compositional Response (Type 1)

A model with compositional response assumes that a random composition Y is a

linear function (in the sense of the Aitchison geometry) of several explanatory real
random variables X 0 , X 1 , . . . , X P , which gives the expected value of some normally
distributed composition,

P
Ŷ = X i  bi , Y ∼ NS D (Ŷ, Σ ε ), (8)
i=0

where NS D (Ŷ, Σ ε ) stands for the normal distribution on the simplex of Y (Mateu-
Figueras and Pawlowsky-Glahn 2008), parametrized in terms of a compositional mean
vector and a covariance matrix of the random composition in some ilr representation.
This reflects the fact that the normal distribution on the simplex of a random com-
position corresponds to the (usual) normal distribution of its ilr representation. This
regression model is useful for explanatory variables of type quantitative (regression),
categorical (ANOVA) or a combination of both (ANCOVA). Note that one can establish
this regression model for compositional data in a least-square sense (Mood et al. 1974,
Chapter X), free of the normality assumption, by using the Aitchison distance [Eq.
(2)] as Daunis-i-Estadella et al. (2002) proposed. However, the normality assumption
is needed in the context of hypotheses testing which is one of the main contributions
of this paper. Specifically, it serves for deriving the distribution of the test statistics in
the classical (least squares) regression case and serves also as the reference model for
robust regression.
If a logratio transformation is applied to this model, this yields a conventional
multiple, multivariate linear regression model on coordinates


P
Ŷ∗ = X i · bi∗ , Y∗ ∼ N D−1
(Ŷ∗ , Σ ε ). (9)
i=0

The model parameters are thus the slopes b∗0 , b∗1 , . . . , b∗P , and the residual covariance
matrix Σ ε . Note that it is common to take X 0 ≡ 1 and then b∗0 rather represents
the intercept of the model in the logratio coordinate system chosen. The specification
in Eq. (9) has the advantage to be tractable with conventional software and solving
methods. Once estimates of the vector coefficients are available, they can be back-
transformed to compositional coefficients, e.g. b̂i = ilr −1 (b̂i∗ ) if calculations are done
in ilr coordinates. Alternatively, ilr coordinates can also be converted to clr coefficients
with b̂iclr = V · b̂i∗ .
It is important to emphasise that the predictions provided by this regression model
are unbiased both in terms of any logratio representation [Eq. (9)], and in terms of
the original composition [Eq. (8)] with respect to the Aitchison geometry discussed in

123
Math Geosci (2021) 53:823–858 829

Sect. 2.1. This follows directly from the isometry of the ilr or clr mappings (Egozcue
et al. 2012; Pawlowsky-Glahn et al. 2015; Fišerová et al. 2016). If interest lies on
understanding the unbiasedness properties of predictions (8) with respect to the con-
ventional Euclidean geometry of the real multivariate space R D , i.e. on the nature
of the expected value of Ŷ − Y, then one can resort to numerical integration of the
model explicated by Eq. (8), which provides the conditional distribution of Y given
Ŷ (Aitchison 1986).

2.2.2 Regression with Compositional Explanatory Variable (Type 2)

A model with a compositional explanatory variable X and one explained real variable
Y is (both in composition and coordinates)

Ŷ = b0 + b, X A = b0 + [b∗ ]t · X∗ = b0 + clr(b), clr(X) , Y ∼ N (Ŷ , σε2 ).

The model parameters are thus the intercept b0 , the gradient b∗ and the residual vari-
ance σε2 , which again can be estimated with any conventional statistical toolbox. The
gradient, once estimated in coordinates, can be back-transformed to a compositional
gradient as b̂ = ilr −1 (b̂∗ ), or to its clr representation by b̂iclr = V · b̂i∗ . Note that
solving a Type 2 model directly in clr would require the use of generalised inversion
of the covariance matrix of clr(X), which provides the same results but at a higher
computational cost.

2.2.3 Compositional to Compositional Regression (Type 3)

A model with both an explanatory X ∈ S Dx and an explained Y ∈ S D y composi-

tional variables can be expressed in several ways. This time, the easiest is directly in
ilr coordinates,

Ŷ∗ = b∗0 + B∗ · X∗ , Y∗ ∼ N D−1

(Ŷ∗ , Σ ε ), (10)

with model parameters a (D y −1)-component vector intercept b∗0 , a (D y −1)×(D y −1)-

element residual covariance matrix Σ ε , and a (D y − 1) × (Dx − 1)-matrix of slope
coefficients B∗ = [bi∗j ]. Note that in this composition-to-composition model it is
necessary to distinguish between the number of components Dx versus D y and the
transformations ilr x and ilr y used for each of the two compositions. In this repre-
sentation, again, the model parameters can be estimated with any available tool, and
then we can interpret them in compositional terms. The intercept b0 = ilr −1 ∗
y (b0 ) is
the expected response composition when the explanatory composition has a neutral
value X = nx . The covariance matrix Σ ε has the same interpretation as in the model
with compositional response (Eq. 9), as the variance of the response conditional on
the explanatory variable. Finally, the slope matrix can be seen in three different ways.
First, each row represents the gradient bi· = ilr −1 ∗ ∗ ∗
x [bi1 , bi2 , . . . , bi(Dx −1) ] ∈ S
Dx

along which one particular ilr response coordinate increases fastest,

Ŷi∗ = (b0∗ )i + bi· , X A = (b0∗ )i + [bi1

∗ ∗
, bi2 ∗
, . . . , bi(D x −1)
]t · X ∗ .

123
830 Math Geosci (2021) 53:823–858

Second, each column represents the slope b·i = ilr −1 ∗ ∗ ∗

y [b1i , b2i , . . . , b(D y −1)i ] ∈ S
Dy

along which each explanatory ilr coordinate can modify the response,

P
Ŷ∗ = b0 ⊕ X i∗  b·i .
i=1

Third, one can interpret B∗ as the matrix representation (on the chosen bases of the
two simplexes) of a linear application B: S Dx → S D y , which is nothing else than the
combination of a rotation on S Dx and a rotation on S D y , together with a univariate
linear regression of each of the pairs of rotated axes:


R
∗
B = ui · di · vit , R ≤ min(Dx , D y ) . (11)
i=1

Here, the matrix U = [ui ] of left vectors is the rotation on the image simplex S D y , and
that of the right vectors V = [vi ] the rotation on the origin simplex S Dx . The coefficient
di is then the slope of the regression between the pair of rotated directions. Note that
this representation coincides with a singular value decomposition of the matrix B∗ ,
and is reminiscent of methods such as canonical correlation analysis or redundancy
analysis (Graffelman and van Eeuwijk 2005). To recover clr representations of the
matrix of coefficients, or of these singular vectors, one just needs the respective basis
matrices Vx and V y ,

Bclr = V y · B∗ · Vtx , uiclr = Vx · ui∗ , viclr = V y · vi∗ (12)

These expressions apply to the model coefficients (B∗ ) and to their estimates (B̂∗ )
given later on in Sects. 3 and 4.
Note that the same issues about the unbiasedness of predictions raised in Sect. 2.2.1
apply to predictions obtained with Eq. (10).

2.3 Subcompositional Independence

One of the most common tasks of regression is the validation of a particular model
against data, in particular models of (linear) independence, partial or complete. In
a non-compositional framework, independence is identified with a slope or gradient
matrix/vector identically null (complete independence), or just with some null coef-
ficients (partial independence). Complete independence for compositional models is
also identified with a null slope, null gradient vectors, or null matrices of the model
established for coordinates (each slope bi∗ , the gradient b∗ resp. the matrix B∗ ). But
partially nullifying one single coefficient of these vectors or matrices just forces inde-
pendence of the covariable(s) with a certain logratio, not with the components this
logratio involves. The necessary concept in this context is thus rather one of subcom-
positional independence, i.e. that a whole subset of components has not influence in

123
Math Geosci (2021) 53:823–858 831

resp. is not influenced by a covariable. One must further distinguish two cases, namely
internal and external subcompositional independence.
Consider the first s components of the composition as independent of a given
covariable. One can then construct a basis of S D with three blocks: an arbitrary basis
of s − 1 vectors comparing the first s components (independent subcomposition), the
balancing vector between the two subcompositions (Eq. 7), and an arbitrary basis of
r − 1 vectors comparing the last r = D − 1 components (dependent subcomposition).
In a Type 1 regression model (compositional response), internal independence of a
certain subcomposition with respect to the ith explanatory covariable X i means that
this covariable is unable to modify the relations between the components of the inde-
∗ = b∗ = · · · = b∗
pendent subcomposition, i.e. b1i 2i (s−1)i = 0. External independence
further assumes that the balance coordinate is independent of the covariable, bsi ∗ = 0.

In a Type 2 regression model (compositional input), internal independence means

that the explained covariable Y cannot change due to variations within the independent
subcomposition, i.e. that the coordinate gradient satisfies b1∗ = b2∗ = · · · = b(s−1)
∗ = 0.
External independence further assumes that the explained variable only depends on
the relationships within the dependent subcomposition, thus additionally bs∗ = 0.
Subcompositional independence for a Type 3 regression model (composition-to-
composition) inherits from the concepts mentioned before. The response subcompo-
sition formed by its first s y parts is internally independent of the input subcomposition
formed by its first sx parts if no modification within the latter can induce any change
within the former, i.e. bi∗j = 0 for i = 1, 2, . . . , (s y − 1) and j = 1, 2, . . . , (sx − 1).
Similarly, external independence further assumes bi∗j = 0 for i = 1, 2, . . . , s y and
j = 1, 2, . . . , sx , including regression coefficients involving the balancing elements
as well.
If the case study or question at hand does not suggest a subcomposition or subcom-
positions to test for such subcompositional independence hypotheses, candidates can
be found with the following heuristic, inspired by the Q-mode clustering of composi-
tional parts (van den Boogaart and Tolosana-Delgado 2013; Filzmoser et al. 2018). For
a certain vector of regression coefficients bi∗ (a slope, a gradient, a row or a column
of a Type 3 coefficient matrix B∗ ), one can obtain the vector of clr coefficients by
biclr = V · bi∗ . The key idea now is to realize that if biclrj − bik
clr = 0 then log(X / X )
j k
does not influence Yi , resp. X i does not influence log(Y j /Yk ). Hence we can compute
the matrix of interdistances between the clr variables d 2jk = (clr j (bi )−clr k (bi ))2 , and
apply any hierarchical clustering technique, naturally defining an ilr basis that isolates
those subcompositions which clr coefficients are most similar, and in consequence
more probably are not influenced by resp. do not influence the ith covariable.

3 Classical Least Squares (LS) Estimation

3.1 LS Estimation in Regression Type 2

We denote the n observations of the response by y1 , . . . , yn , and those of the explana-

tory logratio-transformed compositions by the vectors x1∗ , . . . , xn∗ of D components
(i.e., D − 1 logratio coordinates plus a one in the first component to account for the

123
832 Math Geosci (2021) 53:823–858

intercept, if that is included in the model). The regression parameters are denoted by
the vector b∗ = [b0∗ , b1∗ , . . . , b∗D−1 ]t , and the scale of the residuals is σε . The residuals
are denoted as ri (b∗ ) = yi − [b∗ ]t xi∗ , for i = 1, 2, . . . , n.
Considering the vector of all responses y = [y1 , . . . , yn ]t and the matrix of all
explanatory variables X∗ = [x1∗ , . . . , xn∗ ] (each row is an individual, the first column
is the constant 1, and each subsequent column an ilr coordinate), the least squares
estimators of the model parameters are


b∗ = [(X∗ )t · X∗ ]−1 · (X∗ )t · y

and

1  2 ∗
n
σ̂ε2 = ri (b ).
N−D
i=1

Finally, the covariance matrix of 

b∗ can be estimated as

 b = σ̂ε2 · [(X∗ )t · X∗ ]−1 .

Σ

3.2 LS Estimation in Regression Types 1 and 3

When the response is compositional, we obtain observed logratio score vectors

y1∗ , . . . , yn∗ of dimension D − 1, and the regression coefficients are collected in the
(D − 1) × (P + 1) matrix B∗ . The first column of this matrix represents the intercept
coordinate vector b∗0 . The remaining columns can be linked to P explanatory real
covariables (Type 1 regression) or to the P = Dx − 1 ilr-transformed explanatory
composition (Type 3 regression). The residual vectors are ri∗ (B∗ ) = yi∗ − B∗ xi , for
i = 1, . . . , n.
Considering the matrices of explanatory and response variables X∗ = [x1∗ , . . . , xn∗ ]
and Y∗ = [y1∗ , . . . , yn∗ ] (each row is an individual, compositions are ilr-transformed),
the least squares estimators of the model parameters are


B∗ = [(X∗ )t · X∗ ]−1 · (X∗ )t · Y∗

and

1  ∗ ∗ t ∗ ∗
n
ε =
Σ ri (B ) · ri (B ).
N−P
i=1

Finally, the covariance matrix of 

b∗ can be estimated as

b = Σ
Σ  ε ⊗ [(X∗ )t · X∗ ]−1 ,

where ⊗ is the Kronecker product of the two matrices, and 

B∗ is vectorized stacked
by columns.

123
Math Geosci (2021) 53:823–858 833

3.3 LS Testing of the Subcompositional Independence

The classical theory of linear regression modeling provides a wide range of tests on
regression parameters, both in univariate regression (Type 2) and multivariate regres-
sion models (Types 1 and 3) (Johnson and Wichern 2007). Among them, we are
particularly interested in those that are able to cope with subcompositional indepen-
dence (in its external and internal forms, respectively), as introduced in Sect. 2.3.
For the model of Type 2 and the internal subcompositional independence, the cor-
responding hypothesis on the regression parameters can be expressed as Ab∗ = 0
with A = (0, I), where I is the (s − 1) × (s − 1) identity matrix and 0 stands for an
(s − 1) × (D − s + 1) matrix with all its elements zero. In the alternative hypothesis the
former equality does not hold. Note that for the case of external subcompositional inde-
pendence, the sizes of the matrices I and 0 would just change to s × s and s × (D − s),
respectively. In the following, only the internal subcompositional independence will
be considered, the case of the external independence could be derived analogously.
Under the model assumptions including normality on the simplex and the above null
hypothesis, the test statistic

(S R − S)/(s − 1)
T = , (13)
S/(n − D)

where


n 
n
S= ri (
b∗ ), S R = ri (
b∗R ),
i=1 i=1

follows an F distribution with s−1 and n− D degrees of freedom. Here,  b∗R denotes the
LS estimates under the null hypothesis (i.e. just the submodel is taken for the estimation
of regression parameters). The hypothesis on internal subcompositional independence
is rejected if t ≥ Fs−1,n−D (1 − α), the (1 − α)-quantile of that distribution. This test
statistic coincides with the likelihood ratio test on the same hypothesis, that can be
easily generalized for the case of multivariate regression. The statistic can be written
also in the form
S ∗R − S ∗
T =
s−1

for
n  ∗ 
 
 ri (b ) 
n
ri (
b∗R )
∗
S = , S ∗R = . (14)
σ̂ε σ̂ε
i=1 i=1

Finally, note that frequently the fact is used that the distribution of (s − 1)T converges
in law to a χ 2 distribution with s − 1 degrees of freedom for n → ∞.
Similarly, it might be of interest if some of the (non-compositional) explana-
tory variables do not influence the compositional response (Type 1); in case of a

123
834 Math Geosci (2021) 53:823–858

Type 3 model we could consider the problem in terms of subcompositional indepen-

dence again. Possible testing of subcompositional independence in the compositional
response thus should be performed directly with the involved subcomposition. For the
(D − 1) × (P + 1) matrix of regression coefficients B∗ the null hypothesis can now
be expressed as AB∗ = 0 with the alternative that this equality does not hold. Matrix
A has the same structure as before, it is just adapted to the new notation, thus hav-
ing s − 1 (internal subcompositional independence) or s (external subcompositional
independence) columns, respectively, and D − 1 rows. The usual strategy is to employ
the likelihood ratio test with the statistic (for the case of internal subcompositional
independence)
 b) n/2
det(Σ
Λ= ,
 bR )
det(Σ

used in the modified form

    bR ) 
1 det(Σ
TM = n − P − 1 − (D − (s − 1)) ln , (15)
2 b)
det(Σ

where Σ  b R denotes the estimated covariance matrix of the estimated matrix of regres-
sion parameters in the submodel, formed under the null hypothesis. For n → ∞ the
statistic TM converges to a χ 2 distribution with (D − 1)(s − 1) degrees of freedom
(Johnson and Wichern 2007).

4 Robust MM Estimation

Many proposals for robust regression are available in the literature (see Maronna et al.
2006). The choice of an appropriate estimator depends on different criteria. First of
all, the estimator should have desired robustness properties, i.e. robustness against
a high level of contamination, and at the same time high statistical efficiency. MM
estimators for regression possess the maximum breakdown point of 50% (i.e. at least
50% of contaminated samples are necessary in order to make the estimator useless),
and they have a tunable efficiency. Although other regression estimators also achieve
a high breakdown point, like the LTS regression estimator, their efficiency can be
quite low (Maronna et al. 2006). Another criterion for the choice is the availability of
an appropriate implementation in software packages. MM estimators for regression
are available in the software environment R (R Development Core Team 2019). For
univariate response (Type 2 regression) we refer to the function lmrob of the R
package robustbase (Maechler et al. 2018), for multivariate response (Types 1
and 3) there is an implementation in the package FRB, which also provides inference
statistics using the fast robust bootstrap (Van Aelst and Willems 2013).

123
Math Geosci (2021) 53:823–858 835

4.1 MM Estimation in Regression Type 2

In the following we provide a brief description of MM estimators for regression, with

the same notation as in Sect. 3.1, considering first the case of Type 2 regression. A
regression M estimator is defined as


n  
 ri (b)
b∗ = argmin ρ , (16)
b σ̂ε
i=1

where σ̂ε is a robust scale estimator of the residuals (Maronna et al. 2006). The function
ρ(·) should be bounded in order to achieve good robustness properties of the estimator
(for details, see Maronna et al. 2006). An example is the bisquare family, with
    2  4 
r 2
3 − 3 rk + rk for |r | ≤ k
ρ (r ) = k . (17)
1 else

The constant k is a tuning parameter which gives a tradeoff between robustness and
efficiency. When k gets bigger, the resulting estimate tends to LS, thus being more
efficient but less robust. A choice of k = 0.9 leads to a good compromise with a given
efficiency.
The crucial point is to robustly estimate the residual scale which is needed for
the minimization problem (Eq. 16). This can be done with an M-estimator of scale,
defined as the solution of the implicit equation
 
1
n
ri (b)
ρ1 = d. (18)
n σ̂ε
i=1

Here, ρ1 is a bounded function and d is a constant. With this choice, regression S-

estimators are defined as

b̂ = arg min σ̂ε (r1 (b) , . . . , rn (b)) . (19)

b

Regression S-estimators are highly robust but inefficient. However, one can compute
an S-estimator  b∗(0) as a first approach to b∗ , and then compute σ̂ε as an M-estimator
of scale using the residuals from  b∗(0) (see Maronna et al. 2006). Yohai (1987) has
shown that the resulting MM estimator  b∗ inherits the breakdown point of 
b∗(0) , but
its efficiency under normal distribution is determined by tuning constants. The default
implementation of the R function lmrob attains a breakdown point of 50% and an
asymptotic efficiency of 95% (Maechler et al. 2018).

4.2 MM Estimation in Regression Types 1 and 3

For the case of regression Types 1 and 3, multivariate regression MM estimators

can be used as robust counterparts to LS estimators. With notation as in Sect. 3.2,

123
836 Math Geosci (2021) 53:823–858

compositional MM estimators are defined as

 

n
ri∗ (B)t C−1 ri∗ (B)
(B , 
∗ C) = argmin ρ , (20)
B σ̂
i=1

with the scale estimator σ̂ := det(Σ  S is obtained from a mul-

 S )1/(2D−2) , where Σ
tivariate regression S-estimator (see Van Aelst and Willems 2013, for details). The
estimated residual covariance matrix is then given by Σ ε = σ̂ 2 
C.

4.3 MM Testing of Subcompositional Independence

Robust hypothesis tests in linear regression are not straightforward, because they have
to involve robust residuals, and some tests also rely on a robust estimation of the
covariance matrix of the regression coefficients. In the following we will focus on
tests which can cope with subcompositional independence.
For the univariate case (Type 2) a robust equivalent to the test mentioned in Sect. 3.3
is available. It is a likelihood ratio-type test which, unlike a Wald-type test, does not
require the estimation of the covariance matrix of  b∗ . The hypothesis to be tested is
∗
the same as stated in Sect. 3.3, namely Ab = 0, with A = (0, I) and I an identity
matrix of order s − 1. For the alternative hypothesis Ab∗ = 0. In analogy to the terms
in (14), the test is based on
 ∗   

n
ri (b ) 
n
ri (
b∗R )
∗ ∗
S = ρ , SR = ρ , (21)
σ̂ε σ̂ε
i=1 i=1

where ρ(·) is a bounded function and σ̂ε is a robust scale estimator of the residuals,
see also Eq. (16). With the choice
n   
 ∗
i=1 ψ ri (b )/σ̂ε
ξ= n  2  ,
i=1 ψ ri (
b∗R )/σ̂ε

where ψ = ρ  , the test statistic

T = ξ(S ∗R − S ∗ ) (22)

approximates a χ 2 distribution with s − 1 degrees of freedom, χs−12 (see Hampel et al.

1986). The null hypothesis is rejected at the significance level α if the value of the test
2 (1 − α).
statistic t > χs−1
For regression Type 1 and 3 we can use the robust equivalent of the likelihood ratio
test mentioned in Sect. 3.3. According to Eq. (15), the covariance matrix of the esti-
mated matrix of regression parameters is needed. This can be obtained by bootstrap
as follows. In their R package FRB, Van Aelst and Willems (2013) provide function-
ality for inference statistics in multivariate MM regression by using the idea of fast

123
Math Geosci (2021) 53:823–858 837

and robust bootstrap (Salibian-Barrera and Zamar 2002). A usual bootstrap procedure
would not be appropriate for robust estimators, since it could happen that a bootstrap
data set contains more outliers than the original one due to an over-representation of
outlying observations, thus causing breakdown of the estimator. Moreover, recalculat-
ing the robust estimates for every sample would be very time consuming. The idea of
fast and robust bootstrap (FRB) is to estimate the parameters only for the original data.
Let θ contains all estimates   ε in vectorized form, and denote by ΩΘ the set of
B and Σ
possible values of this vectorized model parameter. MM-estimators can be written in
form of a system of fixed-point equations, i.e. thanks to a function g: ΩΘ → ΩΘ such
that 
θ = g(θ ). Indeed, if the function g is known, one can estimate θ as the fixed point
of the equation. The function g depends on the sample, hence for a bootstrap sample
we obtain a different function gb . The idea is thus to use the original estimate and
the fixed-point equation for the bootstrap sample, obtaining  θ b := gb (
1
θ ). This results
in an approximation of the bootstrap estimates  θ b which would be obtained directly
from the bootstrap sample, i.e. solving  θ b = gb (
θ b ). Applying a Taylor expansion, an

improved estimate θ b can be derived, estimating the same limiting distribution as 
I
θ b,
and being consistent for  θ . For more details concerning fast and robust bootstrap for
the MM-estimator of regression see Salibian-Barrera et al. (2008).

5 Case Study: the GEMAS Data Set

5.1 General Information

The GEMAS (“Geochemical Mapping of Agricultural and grazing land Soil”) soil
survey geochemical campaign was conducted at European level, coordinated by
EuroGeoSurveys, the association of European Geological Surveys (Reimann et al.
2014a, b). It covered 33 countries, and it focuses on those land uses that are vital for
food production. The area was sampled at a density of around 1 site per 2,500 km2 .
Samples were taken from agricultural soils (0 to 20 cm) and grazing land soils (0 to
10 cm). At each site, 5 samples at the corners and in the center of a square with 10 by
10 m were collected, and the composite sample was analyzed. Around 60 chemical
elements were obtained in samples of both kinds of soil. Soil textural composition
was also analyzed, i.e. the weight % of sand, silt and clay. Some parameters describ-
ing the climate (climate type, mean temperature or average annual precipitation) and
the background geology (rock type) are also available. More specifically, the average
annual precipitation and the annual mean temperature at the sample locations are taken
from Reimann et al. (2014a) and originate from www.worldclim.org. The subdivision
of the GEMAS project area into climate zones goes back to Baritz et al. (2005).
From the several variables available, we focus on the effects between the soil com-
position (either its chemistry or its sand-silt-clay texture) and the covariables: annual
average precipitation, soil pH (both as continuous variables) and climate zones [as cat-
egorical variable, with the respective sample sizes; the categories are Mediterranean
(Medi, 438), Temperate (Temp, 1,102), Boreal–Temperate (BoTe, 352) and Suprabo-
real (Spbo, 203) ]. Figure 1 shows a set of descriptive diagrams of these variables
and compositions. A total of n = 2095 samples of the GEMAS data set were used,

123
838 Math Geosci (2021) 53:823–858

A D E

1000
clay
600
Frequency

Frequency

600
200

0 200
0

500 1000 1500 2000 2500 Medi Temp BoTe Spbo

sand silt
Ann. prec. climate

B F
600
Frequency

400
200

. ..
Na2O .. ..
.. .. .. .... ...
0

.
. . .. .... ................ .
500 1000 2000
.
. . . . . ....................... . ... ..
. .. ....... . ......... .... . .
. . .. ........................................................ ...... .. .
Ann. prec. (log−scale) . . ... ................................................. ........ .... . . . . . . .. .
. . . . . . . .... . .. .
. . .. ....... .......................................................................... .... .MgO ...... ... . . . .
.
. . . .
. .. ... . .................................................................. ....... . ....... . ... . . ..... .
. .. . .. .......................................... .... ... ............ ..... ..... . . . . .. . CaO
............................................. ....................... ........ .. .. ............. .... ...
. ...Al2O3
.
C . .K2O .. . . . . . . . . . . . . . . .. . . . . .
...................... ......... ... .. . . ...... . .. ... .. . .. ....... .
. . ..............................................Fe2O3
. .
........................................... ..... .. .. .. . .... . ... .. . ... . . . .
. . . .........TiO2
. . ...........MnO .. . .. .. ... . . . .. .. ... . . . ...
. .... ........................P2O5 .... ... .... . . .. . .. . . . . . .. . . . .
2000

. ............................... ............ ... . . ... . .. . ..... . ... ... . . . . .

Ann. prec. (log−scale)

..SiO2
. .... .... .......... ............. .... .... . . . LOI .. . . . .
. . .. . . . ...... . . ...
. . .
. . ... .... ........ .......... . ... . . . . . . . . . . .
. . . . . . . ... ..
. .. . . .
.. . . ....... ............. .. . . .. ... . . .
1000

. . ... . . . . .. ... ...... . . . . . . .

.. . . ...... . . .. . . . ... . .. . .
. . . . . . .. . .
. . . .. . .. . ... . .... . . .
500

. .. .. . . . .
. . . .
. . . .
.. . .
.
Medi Temp BoTe Spbo
.
.
climate

Fig. 1 Descriptive diagrams of the sets of variables used. Displayed are histograms of Annual Precipitation
in the original (a) and logarithmic (b) scales. Boxplots of log-scaled Annual Precipitation (c) and histogram
of sample sizes (d) according to climate groups follow. The climate groups are used to color sand-silt-clay
compositions in the ternary diagram (e). Finally, the multivariate data structure of chemical compositions
is captured using the compositional biplot (f)

covering almost all Europe, excepting Romania, Moldova, Belarus, Eastern Ukraine
and Russia (Fig. 2). From a comparison between panels A and B (Fig. 1), one can
conclude that the logarithm of Annual Precipitation is required for further treatment.
Though symmetry or normality are not attained, even with a logarithm [both p values
of the Anderson–Darling test for normality (Anderson and Darling 1952) were zero],
at least a view by the four climatic groups suggest that departures from symmetry are
moderate to mild (Fig. 1c), not going to affect negatively further regression results.
As indicated above, the data present a rather unbalanced design with respect to cli-
matic areas (Fig. 1d), particularly due to the dominance of temperate climate, which
accounts for more than 50% of the samples, see also Fig. 2.
The sand-silt-clay textural composition is represented in Fig. 1e as a ternary dia-
gram, with colors after the four climatic zones: these show a certain control on the

123
Math Geosci (2021) 53:823–858 839

70
65
60
55
latitude
50
45
40
35

−10 0 10 20 30
longitude

Fig. 2 Sample location. Colors after climatic zones: red for Mediterranean, blue for Temperate, green for
Boreal–Temperate and violet for Supraboreal

amount of clay, and this will be explored later. With regard to the major oxide com-
position including SO3 and LOI (loss on ignition), this is represented in Fig. 1f as
a centered logratio covariance biplot, as conventional in compositional data analysis
(Aitchison 1997; Aitchison and Greenacre 2002). This shows a quite homogeneous
data set without any strong grouping that could negatively affect the quality of the
next regression steps.

5.2 Grain Size Composition Versus Precipitation (Type 1 Regression)

The first task is to express the sand-silt-clay composition (response) as a function of

precipitation, in log-scale (explanatory variable), using the regression model from Sect.
2.2.1. Figure 3 displays this dependence, by plotting each of the three possible logratios
on the vertical axis against the logarithm of annual precipitation on the horizontal
axis. The available data were transformed after the following matrix resp. these ilr
coordinates

 −1 −1 +2
 2
yclay
√ √ √ y1∗ = √1 ln ysilt ·ysand
V =
t
−1
6
+1
6 6 , 6 , (23)
√
2
√
2
0 y2∗ = √1 ln ysilt
ysand
2

123
840 Math Geosci (2021) 53:823–858

6.0 6.5 7.0 7.5 6.0 6.5 7.0 7.5

4
2
sand

0
−2
4

4
grain size composition
2

2
0

0
silt
−2

−2
−4

−4
2
0

clay
−2
−4

6.0 6.5 7.0 7.5 6.0 6.5 7.0 7.5

log(Annual Precipitation)

Fig. 3 Grain size composition as a function of (log) annual precipitation; observed GEMAS data (dots) and
fitted models (red: classical; blue: robust). Symbol size (in the lower left panels) is inversely proportional
to the weights computed in robust regression

and a model of the form of Eq. (9) was fitted by the LS method. Table 1 shows the
logratio coefficients, as well as their values once back-transformed. Note that the back-
transformed values would be exactly the same, whatever other logratio transformation
would have been used for the calculations.
Table 1 reports the coefficients for the ilr coordinates defined in Eq. (23), the
corresponding p values, and the back-transformed coefficients, using LS and MM
estimators.The LS estimates show that the ratio silt-to-sand is not affected by annual
precipitation, while their relation to clay does depend on this covariable. In contrast,
for the MM estimators both coordinates are affected by annual precipitation. Figure 3
shows both the LS and MM models, re-expressed in each of the possible pairwise
logratios. Note that the slope and intercept given for the coordinate y2∗ in Table 1
correspond to panel (2,1) of this Fig. 3. The intercepts and slopes for each of the other
panels can be obtained by transforming the coefficients (ysand , ysilt , yclay ) accordingly.

123
Math Geosci (2021) 53:823–858 841

Table 1 Regression models of grain size composition against (log) annual precipitation, using LS and MM
regression

y1∗ y2∗ p value.y1∗ p value.y2∗ Sand Silt Clay

LS: Intercept 0.931 − 0.402 0.023 0.234 0.255 0.144 0.600

LS: log(AnnPrec) − 0.298 0.019 0.000 0.718 0.366 0.376 0.258
MM: Intercept 1.813 0.656 0.000 0.014 0.055 0.139 0.806
MM: log(AnnPrec) − 0.405 − 0.125 0.000 0.004 0.418 0.350 0.233

The columns refer to the estimated parameters for the ilr coordinates, the corresponding p values, and the
back-transformed regression coefficients
2

clay
1
0
y*2
−1
−2
−3

sand silt

−3 −2 −1 0 1

y*1

Fig. 4 Scatter diagrams (left: ilr plane; right: ternary diagram) of the data and of the predictions of both
the LS (red) and MM (blue) models. The lines show the models extrapolated beyond the range of observed
annual precipitation. Symbol size is inversely proportional to the weights computed by robust regression

Figure 4 shows the model predictions for the classical (red) and the robust (blue)
model. The left plot presents the predictions in the ilr coordinates, as they are used
in the regression models, and the right plot shows the predictions for the original
composition. The symbol sizes are inversely proportional to the weights from robust
MM regression, and here it gets obvious that due to very small values of clay (rounded
values), data artifacts are produced in the ilr coordinates, but these observations are
downweighted by MM regression. This is the main reason for the difference between
the LS and the MM model.

5.3 Grain Size Composition Versus Climate (Type 1 Regression)

A regression of the grain size composition (response) against climate zones (explana-
tory variables) should take into account that the climate zones are ordered in a clear
sequence from Mediterranean (Medi), Temperate (Temp), Boreal–Temperate (BoTe)
to Supraboreal (Spbo), ordered from South to North. This is clearly seen in Fig. 5,
showing a relatively constant average sand/silt ratio across climatic zones, but a clear

123
842 Math Geosci (2021) 53:823–858

4
4

4
3

2
log(sand/clay)
log(sand/silt)

log(silt/clay)
2

0
1
0

−2
−1
−2

Medi Temp BoTe Spbo −2 Medi Temp BoTe Spbo Medi Temp BoTe Spbo

climate climate climate

Fig. 5 Logratios of the grain size composition for the different climate zones

Table 2 Compositional centers

Sand Silt Clay
in each climate category,
expressed in percentages Medi 45.7117 33.2898 20.9985
Temp 49.5057 34.2782 14.0220
BoTe 58.8528 36.5830 7.6181
SpBo 65.0806 36.4010 3.7676

monotonous trend of average sand/clay and silt/clay ratios northwards. Such a trend
is followed also by the compositional centers (Pawlowsky-Glahn et al. 2015) for the
respective climate categories, see Table 2.
Thus the following hypothesis of uncorrelation appear as sensible:
1. the balance of sand to silt is uncorrelated with climate (i.e. the sand-silt subcom-
position is internally uncorrelated with climate)
2. the balance of clay to the other two depends on climate only in so-called linear
terms, as explained in the next paragraph.
Given these hypotheses, the same ilr coordinates as in the preceding section (Eq. 23)
will be used here.
In R—package stats; (R Development Core Team 2019)—, a regression model
with an ordered factor of 4 levels requires building an accessory (n×3)-element design
or contrast matrix X, where each row is taken as the corresponding row of Table 3.
The labels L—“Linear”, Q—“Quadratic” and C—“Cubic” stand for the kind of trend
between the four categories fitting the data, L implying that the differences between
two consecutive categories are constant (Simonoff 2003).
Table 4 summarizes the numerical output of this regression model, including esti-
mated coefficients (intercept, and effects L, Q and C) for each of the two balances,
the p values of the hypotheses of null coefficient, and the back-transformed coeffi-
cients. These results are given for both classical (LS) and robust (MM) regression.
Classical LS regression shows that C and Q effects can be discarded for y2∗ but not
L effects, i.e. the first hypothesis (inner uncorrelation of the sand-silt subcomposition
with climate) must be rejected. With regard to the second hypothesis, nullifying the

123
Math Geosci (2021) 53:823–858 843

Table 3 Row vectors to construct the design matrix associated with the categorical variable climate

L Q C

Medi − 0.6708 0.5000 − 0.2236

Temp − 0.2236 − 0.5000 0.6708
BoTe 0.2236 − 0.5000 − 0.6708
SpBo 0.6708 0.5000 0.2236

These numbers result from applying the R function contrasts on the ordered factor variable climate. L
stands for linear effect, Q for quadratic effect and C for cubic effect

Table 4 Fitted coefficients and p values of the regression models of grain size composition versus climate

y1∗ y2∗ p value.y1∗ p value.y2∗ Sand Silt Clay

LS: intercept − 1.238 − 0.308 0.000 0.000 0.548 0.355 0.097

LS: L − 1.195 − 0.142 0.000 0.001 0.493 0.403 0.103
LS: Q − 0.120 − 0.019 0.002 0.594 0.354 0.344 0.302
LS: C 0.045 0.009 0.165 0.765 0.325 0.329 0.346
MM: intercept − 1.143 − 0.223 0.000 0.000 0.515 0.376 0.109
MM: L − 1.202 − 0.157 0.000 0.000 0.498 0.399 0.102
MM: Q − 0.270 − 0.085 0.000 0.006 0.390 0.346 0.264
MM: C 0.017 0.054 0.728 0.022 0.319 0.344 0.338

coefficients for L and Q effects on y1∗ are significantly different from zero (p values
smaller that 0.05 critical level), which implies that the second hypothesis is false as
well. Nevertheless, C effects can be discarded. A global test in the fashion of what
was explained in Sect. 3.3 gives a zero p value for the hypothesis of absence of Q
or C effects, thus supporting these conclusions. Robust regression delivers a similar
picture, except that here all effects are significant for y2 .
Of course, other contrasts could be used for this analysis, depending on the nature
of the hypotheses of dependence that we are interested on testing. If, for instance, one
would want to check whether soils from different climatic zones have on average the
same soil texture, one could have used the constr.treatment function of R to
force this sort of comparison.
One way or another, in a categorical regression model like this, the intercept can
be interpreted as a sort of global average value compensating for the lack of balance
between the four categories. While the conventional compositional center is [sand; silt;
clay] = [52.39%; 35.27%; 12.34%] the least squares regression delivers an estimate
[54.83%; 35.48%; 9.69%] and the robust regression [51.54%; 37.60%; 10.86%], both
downweighting the importance of clay. Note that this intercept does not depend on
which contrast set is chosen for capturing the categorical variable.

123
844 Math Geosci (2021) 53:823–858

LS regression MM regression

0.2 0.2
clr slope coefficient

0.0
0.0

−0.2
−0.2

−0.4
−0.4

−0.6

−0.6
SiO2

TiO2

Al2O3

MnO

MgO

CaO

Na2O

K2O

P2O5

LOI

SiO2

TiO2

Al2O3

MnO

MgO

CaO

Na2O

K2O

P2O5

LOI
Fe2O3

Fe2O3
Fig. 6 Least squares (left) and MM (right) regression coefficients of the clr transformed major oxide
composition versus log-annual precipitation

5.4 Major Oxides Versus Precipitation (Type 1 Regression)

Much richer hypotheses can be tested if the composition used has many parts. To illus-
trate this, a regression of the major oxides against (log) annual precipitation follows. A
natural initial question is whether the soil geochemistry is influenced by precipitation.
For this purpose, Fig. 6 shows the clr coefficients estimated with classical least squares
and robust regression: clr coefficients should never be interpreted individually; rather,
differences between them can be understood as the influence on a particular pairwise
logratio. Thus, we are seeking for the smallest differences between coefficients as
they identify pairs of variables whose balance is not influenced by the explanatory
variable. As a subset of pairwise logratios identifies a (sub)composition, this gives
information about subcompositions that might be potentially internally independent
of the covariable, such as:
– TiO2 –Fe2 O3 –MnO
– Al2 O3 -LOI (with Na2 O in least squares regression, or MgO in robust regression)
– SiO2 –K2 O
A set of ilr coordinates is selected accordingly to contain balances between these
subcompositions. The matrix of signs to build these balances is given in Table 5.
Remember that in a sign table, + 1 indicates variables that appear in the numerator
of the balance, − 1 variables in the denominator, and 0 variables are not involved in
that particular balance. For instance, the balance between the subcompositions TiO2 –
Fe2 O3 –MnO and Al2 O3 –Na2 O–LOI is y4∗ , and the balances (y7∗ , y8∗ ) describe the
internal variability in the subcomposition TiO2 –Fe2 O3 –MnO.
Using this set of balances, a regression model with explanatory variable (log) annual
precipitation is fit, with LS and MM regression. Results are reported in Table 6.
Paying attention to the p values of the slopes of the two models, we conclude that
the subcomposition Al2 O3 –Na2 O–LOI (y7∗ , y8∗ ) is internally independent of annual
precipitation (both classical and robust methods agree in that). Loosely speaking, the

123
Math Geosci (2021) 53:823–858 845

Table 5 Table of signs to

y1∗ y2∗ y3∗ y4∗ y5∗ y6∗ y7∗ y8∗ y9∗ ∗
y10
construct the balances for major
oxide composition in a SiO2 +1 −1 0 0 0 0 0 0 +1 +1
regression problem against (log)
annual precipitation TiO2 +1 +1 +1 +1 +1 +1 0 0 0 0
Al2 O3 + 1 +1 +1 −1 0 0 +1 +1 0 0
Fe2 O3 + 1 +1 +1 +1 +1 −1 0 0 0 0
MnO +1 +1 +1 +1 −1 0 0 0 0 0
MgO −1 0 0 0 0 0 0 0 0 0
CaO +1 −1 0 0 0 0 0 0 −1 0
Na2 O +1 +1 +1 −1 0 0 +1 −1 0 0
K2 O +1 −1 0 0 0 0 0 0 +1 −1
P2 O5 +1 +1 −1 0 0 0 0 0 0 0
LOI +1 +1 +1 −1 0 0 −1 0 0 0

Table 6 Intercept (int) and slope (slp) estimated coefficients and p values (.p), for least squares (LS) and
robust (MM) regression

LS.int MM.int LS.slp MM.slp LS.int.p MM.int.p LS.slp.p MM.slp.p

y1∗ 0.596 1.229 0.013 − 0.086 0.053 0.000 0.776 0.053

y2∗ − 7.722 − 7.469 0.822 0.789 0.000 0.000 0.000 0.000
y3∗ 3.301 3.454 − 0.218 − 0.233 0.000 0.000 0.000 0.000
y4∗ − 3.450 − 3.937 0.140 0.218 0.000 0.000 0.000 0.000
y5∗ 2.876 3.192 − 0.082 − 0.130 0.000 0.000 0.003 0.000
y6∗ − 1.011 − 0.654 − 0.035 − 0.086 0.000 0.000 0.085 0.000
y7∗ − 1.151 − 0.339 0.024 − 0.076 0.002 0.361 0.662 0.157
y8∗ 1.752 1.176 − 0.004 0.074 0.000 0.000 0.911 0.089
y9∗ − 1.243 − 0.242 0.451 0.314 0.036 0.000 0.761 0.000
∗
y10 2.825 2.791 − 0.044 − 0.042 0.000 0.000 0.111 0.073

same applies to the balances SiO2 /K2 O (y10∗ ) and MgO against all other components
∗ ∗
(y1 ). Finally the balance TiO2 /Fe2 O3 (y6 ) appears to be uncorrelated with annual
precipitation only from a least-squares perspective.
Now, global tests of internal and external independence of Al2 O3 –Na2 O–LOI with
respect to annual precipitation were performed after the methodology of Sect. 3.3, and
delivered p values of 0.884 and 0,respectively.These results are somehow at odds with
the common understanding of weathering as a process of enrichment in Al2 O3 (and
perhaps LOI) at the expenses of Na2 O (and CaO). Annual precipitation, one of the
factors of chemical weathering, is not showing any significant effect on the logratio
Al2 O3 /Na2 O. The robust global test, on the other hand, results in significant effects:
in both cases, the p values are zero.

123
846 Math Geosci (2021) 53:823–858

LS regression MM regression
0.6

0.4 0.4
clr slope coefficient

0.2 0.2

0.0 0.0

−0.2 −0.2

−0.4 −0.4
SiO2

TiO2

Al2O3

MnO

MgO

CaO

Na2O

K2O

P2O5

LOI

SiO2

TiO2

Al2O3

MnO

MgO

CaO

Na2O

K2O

P2O5

LOI
Fe2O3

Fe2O3
Fig. 7 Regression gradient of pH against the major oxide composition, expressed in clr coefficients: least
squares estimates (LS, left) and robust estimates (MM, right)

5.5 Compositional Input: pH Versus Major Oxides (Type 2 Regression)

A similar approach can be followed if the goal is to predict an external quantitative

variable as a function of a composition. This section presents the regression of soil pH
against the major oxide composition, the rationale being that the mineral composition
of a soil may be influenced—for example through accelerating dissolution of unstable
minerals under certain pH values at atmospheric conditions—resp. that the mineralogy
may be one of the factors controlling availability of free H3 O+ ions—for example
soils on karstic landscapes are strongly buffered, while those on top of felsic rocks
are strongly acidified. Following the same steps as in the preceding subsection, it is
convenient to start having a look at the regression model coefficients expressed in clr.
In that case, though, it is necessary to start the analysis with a black-box ilr coordinate
set, and convert the estimated ilr coefficients to clr coefficients, in order to avoid the
singularity of the clr variance matrix. Similar as in Fig. 6, one should pay attention to
bars with similar length, in order to identify pairs of variables which balance has no
influence on the explained variable. Figure 7 suggests the following pairs, which are
taken into account by constructing balances according to Table 7: Al2 O3 /P2 O5 (y6∗ )
and Ti2 O5 /MgO (y10 ∗ ). Moreover, by taking the variables with the largest positive and

negative weights we also find balances which could concentrate most of the predicting
power for pH: these are CaO/Na2 O (y2∗ ) and K2 O/LOI (y5∗ ).
The regression results are presented in Table 8 for LS and MM regression. The
table shows the estimated regression coefficients and the corresponding p values.
Both methods reveal that the coefficients for balances y2∗ and y5∗ are significant, while
those for balances y6∗ and y10 ∗ are not. Using the methods from Sect. 3.3 we can

further test subcompositional independence of pH with respect to subcompositions

such as MnO–Fe2 O3 –MgO–TiO2 (p value = 0.291, subcompositionally independent),
or P2 O5 –Al2 O3 –K2 O–LOI (p value = 6.75e−43, not independent).
As an example, Fig. 8 investigates more deeply the relationship between pH and
balance y2∗ , which relates CaO and Na2 O, where the color information is for high (red)

123
Math Geosci (2021) 53:823–858 847

Table 7 Table of signs to

y1∗ y2∗ y3∗ y4∗ y5∗ y6∗ y7∗ y8∗ y9∗ ∗
y10
construct the balances for major
oxide composition in a SiO2 −1 0 −1 0 0 0 −1 0 0 0
regression problem with pH as
the response variable TiO2 −1 0 −1 0 0 0 1 1 1 1
Al2 O3 − 1 0 1 −1 0 1 0 0 0 0
Fe2 O3 − 1 0 −1 0 0 0 1 1 −1 0
MnO −1 0 −1 0 0 0 1 −1 0 0
MgO −1 0 −1 0 0 0 1 1 1 −1
CaO 1 1 0 0 0 0 0 0 0 0
Na2 O 1 −1 0 0 0 0 0 0 0 0
K2 O −1 0 1 1 1 0 0 0 0 0
P2 O5 −1 0 1 −1 0 −1 0 0 0 0
LOI −1 0 1 1 −1 0 0 0 0 0

Table 8 LS and MM regression coefficients and p values for a regression of pH on the major oxide balances

LS.coeff MM.coeff LS.p.value MM.p.value

Intercept 7.099 7.092 0.000 0.000

y1∗ 0.044 0.057 0.077 0.037
y2∗ 0.800 0.805 0.000 0.000
y3∗ − 0.349 − 0.396 0.000 0.000
y4∗ 0.223 0.178 0.020 0.077
y5∗ 0.565 0.575 0.000 0.000
y6∗ − 0.046 − 0.057 0.549 0.458
y7∗ 0.111 0.095 0.011 0.038
y8∗ − 0.091 − 0.068 0.058 0.165
y9∗ 0.045 0.025 0.608 0.778
∗
y10 − 0.002 − 0.022 0.966 0.685

and low (blue) values of pH. One can see that the ratio of CaO and Na2 O increases
strongly with higher values of pH, leading to a non-linear relationship. The reason is
seen in the middle panel of Fig. 8 using the same coloring, where high pH values are
connected to high concentrations of CaO. These pH rich samples are indicated in the
project area in the right panel with red color. This supports the starting hypothesis of
a strong control on pH of the buffering ability of carbonate soils. This trend can be
explained as the contrast between silicic–clastic plus crystalline rocks with significant
contributions of Na-rich silicates versus carbonate karstic landscapes, dominated by
CaCO3 with its very strong buffering effect at slightly basic pH values. Such a complex
trend could be better captured either with a non-linear regression method, or stepwise
linear regression to be carried out only for the samples which behave similarly (blue
or red).

123
848 Math Geosci (2021) 53:823–858

70
4

65
7

60
3

latitude
55
6

Na2O
pH

50
2
5

45
1

40
4

35
−2 0 2 4 0 0 10 20 30 40 −10 0 10 20 30

y*2 CaO longitute

Fig. 8 pH versus balance y2∗ (left), Na2 O versus CaO (middle), and map of the area (right), with color
indicating pH values lower or higher than 7

Nevertheless, it is also interesting to inspect regression diagnostic plots. For this

purpose, a robust regression diagnostic plot based on standardized regression residu-
als and robust Mahalanobis distances was employed to unmask outlier observations.
This plot enables distinguishing between vertical outliers (outliers in y-direction) and
leverage points (Rousseeuw and van Zomeren 1990). The latter are observations with
unusual values in the explanatory variables, which can either strengthen (good lever-
age points) or break (bad leverage points) the overall regression trend. The cut-off
lines for vertical outliers are represented by the quantiles of the standard normal dis-
tribution and the red curve represents a kernel fit of the points. The MM regression
diagnostic plot (Fig. 9, left) reveals only few vertical outliers and bad leverage points
which indicates that model assumptions such as homoscedasticity (see Fig. 10) are
fulfilled. However, in line with previous findings, observations with high pH values
(higher than 7) form a specific pattern. This is further confirmed when observed versus
fitted values are plotted (Fig. 9, right) which indicates a certain bias resulting from an
underlying non-linearity in the relationships between the response and the explanatory
variables, and this was already assumed previously.

5.6 Grain Size Versus Chemistry (Type 3 Regression)

In a final example we want to investigate if the grain size composition is affected

by the major oxides. Similar as in the previous example, the modelling hypothesis
here is that the grain size is controlled by mineralogy, with certain minerals (and
their constituting elements) being enriched in certain fractions: coarse quartz (SiO2 )
and feldspars ((K,Na,Ca)Al(Al,Si)2 O8 ) in sand; highly stable heavy minerals as rutile
(TiO2 ) or apatite (Ca5 (PO4 )3 (F,Cl,OH)) in silts; and alteration and clay minerals with
highly complex chemistry (but typically enriched in K, Fe, Mg, Al, OH and water) in
clay-sized fractions. These relations, combined with the fact that size information is
of relatively bad analytical quality (see Table 9) one could consider the possibility to
support the quantification of size particles with chemistry.

123
Math Geosci (2021) 53:823–858 849

Standardized residuals vs. Robust Distances Response vs. Fitted Values

3
Robust Standardized residuals
2

8
1

7
Response
0

6
−1

5
−2

724

4
1108
−3

1108
724 488
488

3
−4

0 5 10 15 3 4 5 6 7 8
Robust Distances Fitted Values

Fig. 9 Robust regression diagnostic plot (left), observed versus fitted pH values using the major oxides as
predictors (right). Observations with pH values higher than 7 are marked in red
1
MM residuals
0
−1
−2

0 500 1000 1500 2000

Index

Fig. 10 MM regression residuals. Observations with pH values higher than 7 are marked in red

Table 9 Descriptive statistics of the variability of the total sum of %sand + %silt + %clay, which should
be 100%

Min. 1st Qu. Median Mean 3rd Qu. Max.

59.00 95.00 99.00 98.97 103.00 141.00

To start with, we follow the same approach as in the previous cases and plot the
coefficients resulting from LS or robust regression in terms of clr coefficients, Fig. 7.
We can then look at similar contributions of the several oxides on each grain size
fraction to formulate hypotheses. Note that this leads to a matrix of regression coeffi-
cients, linking the grain size distribution as responses to the major oxide composition
as explanatory variables. Table 10 reports the coefficients of an LS Type 3 regression
model, albeit with all coefficients expressed in clr representation.
For establishing subcompositional independence, though, it is more convenient
to work in isometric logratios. Hence, and given the results obtained until now it
appears sensible to study the single component independence of clay (vs. sand and
silt in balance y2∗ ) on the one hand, and on the other the internal subcompositional

123
850 Math Geosci (2021) 53:823–858

Table 10 Least-square
Sand Silt Clay y1∗ y2∗
coefficients of a Type 3
regression model, represented in Intercept − 3.5756 3.3797 0.1958 4.9181 0.2398
clr coefficients (first 3 columns)
and in ilr coordinates (last 2 SiO2 0.7096 − 0.4581 − 0.2515 − 0.8257 − 0.3080
columns) TiO2 − 0.8453 0.5952 0.2501 1.0186 0.3063
Al2 O3 0.4882 − 0.5858 0.0977 − 0.7594 0.1196
Fe2 O3 − 0.0012 0.0434 − 0.0422 0.0315 − 0.0516
MnO − 0.1958 0.0029 0.1928 0.1405 0.2362
MgO − 0.0402 − 0.1334 0.1737 − 0.0659 0.2127
CaO − 0.0794 − 0.0214 0.1008 0.0410 0.1235
Na2 O 0.1614 0.2218 − 0.3832 0.0428 − 0.4693
K2 O − 0.5873 0.1671 0.4203 0.5335 0.5147
P2 O5 0.3422 0.0191 − 0.3613 − 0.2284 − 0.4425
LOI 0.0480 0.1493 − 0.1973 0.0716 − 0.2417

independence of the subcomposition {sand, silt} in balance y1∗ . The corresponding

model coefficients are also reported in Tables 10 and 12 respectively for y1∗ and y2∗ .
To simplify the model, we now seek major oxide subcompositions that are non-
influential on each of these two logratios of grain size composition. A subcomposition
will not be influential on a certain response Yk∗ if the gradient (isometric) logratio
coefficients b∗k associated to that subcomposition are zero, which is equivalent to
saying that the clr coefficients of the gradient in that subcomposition show the same
value (not necessarily zero!).
For each of the grain size ilr coefficients the heuristic to basis selection of Sect. 2.3
was applied (using the centroid clustering method as hierarchical agglomeration cri-
terion), and an ilr basis was obtained. The balance of clay to sand-silt showed no
subcompositional independence worth exploring. Better results were obtained for the
balance of silt to sand. Table 11 contains the sign table defining the ilr balances, and
the associated gradient coefficients with their corresponding p value of the hypothesis
of null coefficient. At the global 5% level of confidence and applying an incremental
Bonferroni correction, we accept the null hypothesis for all balances between x1∗ and
x5∗ (this last one because 0.05/5 < 0.0110), hence the following subcompositions
have no internal influence on the sand-to-silt balance: SiO2 –Al2 O3 and CaO–Fe2 O3 –
MgO–Na2 O–LOI. The hypotheses of external independence of both subcompositions
produce p values below 2 × 10−16 resp. 5 × 10−6 and hence external independence
must be disregarded. These results suggest that minerals carrying most of Al2 O3 and
SiO2 (essentially, feldspars vs. quartz) may not preferentially concentrate between
sand and silt, and the same happens with dominant minerals composed of elements
of the subcomposition CaO–Fe2 O3 –MgO–Na2 O–LOI (plagioclase, calcite, dolomite,
silicates of smaller typical crystal size). On the contrary, elements such as TiO2 –P2 O5 –
K2 O (captured by balances x7∗ to x9∗ ) and x10 ∗ , the balance of SiO –Al O versus the
2 2 3
other, do have a strong influence on silt/sand. This supports the previous understanding
that stable heavy minerals such as rutile or apatite should be (relatively) enriched on
silt sized soils, while sandy soils should have more quartz and feldspars. Note that the

123
 Table 11 Least squares regression model coefficients and null p values (last three columns) for a tailored ilr representation of the major oxide composition (represented in
the first ten columns) to explain the logratio sand to silt (note: intercept not included)
Math Geosci (2021) 53:823–858

SiO2 TiO2 Al2 O3 Fe2 O3 MnO MgO CaO Na2 O K2 O P2 O5 LOI Estim. Stderr. p value

x∗1 0 0 0 0 0 0 −1 1 0 0 0 0.001 0.017 0.943

x∗2 0 0 0 −1 0 0 1 1 0 0 0 0.009 0.046 0.855
x∗3 0 0 0 1 0 0 1 1 0 0 −1 − 0.029 0.033 0.376
x∗4 −1 0 1 0 0 0 0 0 0 0 0 0.047 0.055 0.398
x∗5 0 0 0 1 −1 0 1 1 0 0 1 − 0.084 0.033 0.011
x∗6 0 0 0 1 1 −1 1 1 0 0 1 0.120 0.035 0.001
x∗7 0 0 0 1 1 1 1 1 0 −1 1 0.252 0.034 0.000
x∗8 0 −1 0 0 0 0 0 0 1 0 0 − 0.343 0.043 0.000
x∗9 0 1 0 −1 −1 −1 −1 −1 1 −1 −1 0.962 0.044 0.000
x∗10 −1 1 −1 1 1 1 1 1 1 1 1 1.239 0.068 0.000

123
851
 852

123
Table 12 Least squares regression model coefficients and null p values (last three columns) for a tailored ilr representation of the major oxide composition (represented in
the first ten columns) to explain the balance of clay against sand and silt (note: intercept not included)

SiO2 TiO2 Al2 O3 Fe2 O3 MnO MgO CaO Na2 O K2 O P2 O5 LOI Estimate Std.error p value

x1∗ 0 0 −1 0 0 0 1 0 0 0 0 0.566 0.055 0.000

x2∗ 0 0 0 0 −1 1 0 0 0 0 0 − 0.146 0.028 0.000
x3∗ 0 0 0 0 0 0 0 −1 0 1 0 − 0.192 0.030 0.000
x4∗ −1 0 0 0 0 0 0 0 0 0 1 0.635 0.033 0.000
x5∗ 0 −1 0 0 1 1 0 0 0 0 0 − 0.801 0.042 0.000
x6∗ 0 1 −1 0 1 1 −1 0 0 0 0 0.793 0.052 0.000
x7∗ 1 0 0 0 0 0 0 −1 0 −1 1 − 0.284 0.028 0.000
x8∗ 0 1 1 −1 1 1 1 0 0 0 0 0.040 0.068 0.560
x9∗ 0 1 1 1 1 1 1 0 −1 0 0 − 0.431 0.048 0.000
∗
x10 −1 1 1 1 1 1 1 −1 1 −1 −1 0.589 0.023 0.000
Math Geosci (2021) 53:823–858
Math Geosci (2021) 53:823–858 853

−1.5 −1.0 −0.5 0.0 0.5 1.0 1.5

0.6

1.5
Na2O

0.4
P2O5

1.0
LOI
0.2

silt

0.5
TiO2
Fe2O3
sand
0.0

0.0
SiO2
CaO

−0.5
−0.2

MnO
clay
MgO
K2O

−1.0
−0.4

Al2O3

−1.5
−0.4 −0.2 0.0 0.2 0.4 0.6

Fig. 11 Simultaneous plot of the left and (scaled) right singular vectors of the regression coefficients
matrix, expressed in clr coefficients on both image and origin compositional spaces

picture for y2∗ is completely different: here we can only hope to simplify one coeffi-
cient, namely x8∗ , giving the balance of Fe with respect to the other mafic components.
This is nevertheless irrelevant for the sake of subcompositional independence testing,
because the rest of the balances between mafic components (x2∗ , x5∗ and x7∗ ) do show
significant coefficients.
Another way of looking at the model coefficients is to express them via the singu-
lar value decomposition of Eq. (11). A naive simultaneous plot of the left and right
singular vectors (these last ones scaled by the singular values) is given in Fig. 11. In
this diagram, links joining two variables represent the direction (on the origin or on
the image simplexes, resp. Sx or S y ) associated with fastest change of the logratio
of the two variables involved. A pair of parallel links, one involving components of
Sx and the other linking components of S y , suggests that the logratio between the
involved response variables is strongly controlled by the logratio of the explanatory
variables. For instance, the silt–clay link is reasonably parallel to the link Na2 O–
Al2 O3 ; the same can be said of the links silt–sand versus TiO2 –SiO2 , or of clay–sand
versus K2 O–SiO2 . An analogous reasoning applies for orthogonal links: they indicate
lack of dependence between the two sets of variables involved. In other words, by
finding orthogonal links we identify subcompositions to test for potential subcom-
positional independence. For instance, the link sand-silt is roughly orthogonal to the
sets SiO2 –Al2 O3 and CaO–Fe2 O3 –MgO–Na2 O–LOI(–MnO), that is to the subcom-
positions that were previously tested. Similarly, the diagram suggests as well tests
for subcompositional independence of sand-clay with respect to the subcompositions
SiO2 –P2 O5 –Na2 O or Al2 O3 –Fe2 O3 (–LOI), or even K2 O–TiO2 .

123
854 Math Geosci (2021) 53:823–858

sand silt clay

0.8

0.5
0.8

0.7

0.4
0.6
0.6

0.3
0.5
Fitted

Fitted

Fitted
0.4

0.2
0.4

0.3

0.1
0.2
0.2

0.1

0.0
0.2 0.4 0.6 0.8 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.0 0.1 0.2 0.3 0.4 0.5

Observed Observed Observed

Fig. 12 Observed versus fitted values of the grain size composition, using the major oxides as predictors.
Points on the dashed lines would indicate a perfect match between the observed and fitted values. The
contour lines indicate the density of the point distribution

Finally, we illustrate the model in terms of the fitted values. For this purpose one
can use any ilr coordinates representing the grain size composition, and any ilr coor-
dinates representing the major oxides, and perform LS regression to obtain the fitted
values in ilr coordinates. The appropriate inverse ilr transformation leads to the back-
transformed fitted values of the grain size distribution, which can be compared to
the observed values in Fig. 12 (as kernel density estimates). The same results will
be obtained with any other logratio transformation. The linear model has at least
some predictive power, and one can see a clearer relationship with sand and clay, and
a weaker with silt. This suggests that the major oxides are affecting the grain size
composition mainly by its sand and clay proportions. Obviously, several factors do
contribute to this discrepancy, among other the information effect (the regression line
of true values as a function of predicted values cannot lie above the 1:1 bisector), the
presence of outliers, the bad quality of the input grain size data, the non-linearity of
the back-transformation of predictions from logratios to original components, or the
highly complex relations between chemistry, mineralogy and texture that form the
basis to attempt such a prediction.With respect to outliers, the predictive power can
be improved by using a robust estimator.The non-linearity of the back-transformation
is something that can easily be corrected by means of Hermitian integration of the
conditional distribution of the soil grain size composition provided by Eq. (10), as
proposed by Aitchison (1986). But much more important than those effects are the
uncertainty on the textural data and the complexity of the relation we are trying to
capture here. Indeed, if the goal of the study would be that prediction, linear regression
might not be the most appropriate technique. Tackling this complexity is a matter of
predictive models, beyond the scope of this contribution.

6 Conclusions

The purpose of this contribution was to outline the concept of regression analysis
for compositional data, and to show how the analysis can be carried out in practice
with real data. We distinguished three types of regression models: Type 1, where

123
Math Geosci (2021) 53:823–858 855

the response is a composition and the explanatory variable(s) is a (are) real non-
compositional variable(s); Type 2 with a composition as explanatory variables and a
real response, and Type 3 where both the responses and the explanatory variables are
compositions. Note that one could also consider the case where regression is done
within one composition, by splitting the compositional parts into a group forming the
responses, and a group representing the explanatory variables. This case has not been
treated here because it requires a so-called errors-in-variables model, see Hrůzová
et al. (2016) for details.
For all three types of models it is essential how the composition is treated for regres-
sion modeling. A geometrically sound approach is in terms of orthonormal coordinates,
so-called balances, which can be constructed in order to obtain an interpretation of
the regression coefficients and for testing different hypotheses. If the interest is not in
the statistical inference but only in the model fit and in the fitted values, any logra-
tio coordinates would be appropriate to represent the composition. Note that the clr
transformation would not be appropriate for Type 2 or Type 3 regression models, since
the resulting data matrix is singular, leading to problems for the parameter estimation
when the composition plays the role of the explanatory variables.
Classical least-squares (LS) regression as well as robust MM regression have been
considered to estimate the regression parameters and the corresponding p values for
the hypothesis tests. If the model requirements are fulfilled, the LS regression esti-
mator is the so-called best linear unbiased estimator (BLUE) with the corresponding
optimality properties (see, e.g., Johnson and Wichern 2007), but in that case also MM
regression leads to an estimator with high statistical efficiency. However, in case of
model violations, e.g., due to data outliers, these optimality properties are no longer
valid. Still, the MM estimator is reliable because it is highly robust against outliers,
both in the explanatory and in the response variables. In practical applications it might
not always be clear if outliers are present in the data at hand. In this case it could be
recommended to carry out both types of analysis and compare the results. In particular,
one could inspect diagnostics plots from robust regression (as it was done in Sect. 5.5)
in order to identify potential outliers that could have affected the LS estimator, see
Maronna et al. (2006).
The different regression types and estimators have been applied to an example data
set from the GEMAS project (Reimann et al. 2014a, b). All presented examples are
only for illustrative purposes, but they show how balances can be constructed and
how hypotheses can be tested. For the robust estimators, functions are available in the
R packages robustbase (Maechler et al. 2018) and FRB (Van Aelst and Willems
2013). It is important to note that not only the regression parameters are estimated
robustly with these packages, but robust estimation is also carried out for estimating
the standard errors and for hypothesis testing, for the residual variance, the multiple R 2
measure, etc. We demonstrate the possibilities of regression diagnostics in Sect. 5.5.
In most examples, a comparison of LS and MM regression has been provided.
An important issue in the regression context is the problem of variable selection, or
subcompositional independence. In particular for Type 2 and 3 where the explanatory
variables are originating from a composition, it is not straightforward how to end up
with the “best subset” of compositional parts that does not contain non-informative
parts and still yields a model with similar predictive power as the full model. There are

123
856 Math Geosci (2021) 53:823–858

approaches available in the literature to reduce the number of components, see, e.g.,
Pawlowsky-Glahn et al. (2011), Hron et al. (2013), Mert et al. (2015) and Greenacre
(2019). However, there are no methods of subcompositional independence which work
equivalently to non-compositional methods, such as forward or backward variable
selection; only a brief outlook for those in the compositional context was sketched in
Filzmoser et al. (2018). Those methods will be treated in our future research.

Acknowledgements Karel Hron and Peter Filzmoser gratefully acknowledge the support by
Czech Science Foundation GA19-01768S.

Funding Open access funding provided by ZHAW Zurich University of Applied Sciences

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use
is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission
directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/
by/4.0/.

References
Aitchison J (1986) The statistical analysis of compositional data. Monographs on Statistics and Applied
Probability, London (UK): Chapman & Hall, London. (Reprinted in 2003 with additional material by
The Blackburn Press), ISBN 0-412-28060-4
Aitchison J (1997) The one-hour course in compositional data analysis or compositional data analysis
is simple. In: Pawlowsky-Glahn V (ed) Proceedings of IAMG’97—The III annual conference of
the international association for mathematical geology, volume I, II and addendum, Barcelona (E):
International Center for Numerical Methods in Engineering (CIMNE), Barcelona (E), ISBN 84-87867-
97-9, pp 3–35
Aitchison J, Greenacre M (2002) Biplots for compositional data. J R Stat Soc Ser C (Appl Stat) 51(4):375–
392
Aitchison J, Barceló-Vidal C, Egozcue JJ, Pawlowsky-Glahn V (2002) A concise guide for the algebraic-
geometric structure of the simplex, the sample space for compositional data analysis. In: Bayer U,
Burger H, Skala W (eds) Proceedings of IAMG’02—The eigth annual conference of the International
Association for Mathematical Geology, volume I and II, Selbstverlag der Alfred-Wegener-Stiftung,
Berlin, pp 387–392, ISSN 0946-8978
Anderson TW, Darling DA (1952) Asymptotic theory of certain “goodness-of-fit” criteria based on stochastic
processes. Ann Math Stat 23:193–212
Barceló-Vidal C, Martín-Fernández JA, Pawlowsky-Glahn V (2001) Mathematical foundations of compo-
sitional data analysis. In Ross G (ed) Proceedings of IAMG’01—The VII annual conference of the
international association for mathematical geology, Cancun (Mex)
Baritz R, Fuchs M, Hartwich R, Krug D, Richter S (2005) Soil regions of the European Union and adjacent
countries 1:5,000,000 (Version 2.0)—Europaweite thematische Karten und Datensätze. European Soil
Bureau Network
Billheimer D, Guttorp P, Fagan W (2001) Statistical interpretation of species composition. J Am Stat Assoc
96(456):1205–1214
Coenders G, Martín-Fernández J, Ferrer-Rosell B (2017) When relative and absolute information matter:
compositional predictor with a total in generalized linear models. Stat Model 17(6):494–512
Daunis-i-Estadella J, Egozcue JJ, Pawlowsky-Glahn V (2002) Least squares regression in the Simplex.
In: Bayer U, Burger H, Skala W (eds) Proceedings of IAMG’02—the eigth annual conference of

123
Math Geosci (2021) 53:823–858 857

the International Association for Mathematical Geology, volume I and II, Selbstverlag der Alfred-
Wegener-Stiftung, Berlin, ISSN 0946-8978, pp 411–416
Egozcue JJ, Pawlowsky-Glahn V (2005) Groups of parts and their balances in compositional data analysis.
Math Geol 37:795–828
Egozcue J J, Pawlowsky-Glahn V (2011) Basic concepts and procedures. In: Pawlowsky-Glahn V, Buccianti
A (eds) Compositional data analysis: theory and applications, Wiley, ISBN 978-0-470-71135-4, pp
12–28
Egozcue JJ, Pawlowsky-Glahn V, Mateu-Figueras G, Barceló-Vidal C (2003) Isometric logratio transfor-
mations for compositional data analysis. Math Geol 35(3):279–300, ISSN 0882-8121
Egozcue JJ, Daunis-i-Estadella J, Pawlowsky-Glahn V, Hron K, Filzmoser P (2012) Simplicial regression.
The normal model. J Appl Probab Stat 6(1&2):87–108
Egozcue J, Lovell D, Pawlowsky-Glahn V (2013) Regression between compositional data sets. In: Hron
K, Filzmoser P, Templ M (eds) Proceedings of the 5th international workshop on compositional data
analysis, Vorau
Filzmoser P, Hron K, Templ M (2018) Applied compositional data analysis: with worked examples in R.
Springer, Cham
Fišerová E, Donevska S, Hron K, Bábek O, Vaňkátová K (2016) Practical aspects of log-ratio coordinate
representations in regression with compositional response. Meas Sci Rev 16(5):235–243
Graffelman J, van Eeuwijk F (2005) Calibration of multivariate scatter plots for exploratory analysis of
relations within and between sets of variables in genomic research. Biom J 47(6):863–879
Greenacre M (2019) Variable selection in compositional data using pairwise logratios. Math Geosc 51:649–
682
Hampel F, Ronchetti E, Rousseeuw P, Stahel W (1986) Robust statistics. The approach based on influence
functions. Wiley, New York
Hron K, Donevska S, Fišerová E, Filzmoser P (2013) Covariance-based variable selection for compositional
data. Math Geosci 45(4):487–498
Hrůzová K, Todorov V, Hron K, Filzmoser P (2016) Classical and robust orthogonal regression between
parts of compositional data. Stat A J Theor Appl Stat 50(6):1261–1275
Johnson R, Wichern D (2007) Applied multivariate statistical analysis, 6th edn. Prentice Hall, New York,
p 800
Maechler M, Rousseeuw P, Croux C, Todorov V, Ruckstuhl A, Salibian-Barrera M, Verbeke T, Koller M,
Conceicao ELT, Anna di Palma M (2018) robustbase: basic robust statistics. R package version 0.93-3
Maronna R, Martin R, Yohai V (2006) Robust statistics: theory and methods. Wiley, New York
Mateu-Figueras G, Pawlowsky-Glahn V (2008) A critical approach to probability laws in geochemistry.
Math Geosci 40(5):489–502
Mert C, Filzmoser P, Hron K (2015) Sparse principal balances. Stat Model 15(2):159–174
Mood AM, Graybill FA, Boes DC (1974) Introduction to the theory of statistics, 3rd edn. McGraw-Hill,
New York
Pawlowsky-Glahn V (2003) Statistical modelling on coordinates. In: Thió-Henestrosa S, Martín-Fernández
JA (eds) Proceedings of CoDaWork’03, The 1st Compositional Data Analysis Workshop, Girona (E).
Universitat de Girona, ISBN 84-8458-111-X, http://ima.udg.es/Activitats/CoDaWork2003/
Pawlowsky-Glahn V, Egozcue JJ (2001a) Geometric approach to statistical analysis on the simplex. Stoch
Environ Res Risk Assess (SERRA) 15(5):384–398
Pawlowsky-Glahn V, Egozcue JJ (2001b) Geometric approach to statistical analysis on the simplex. Stoch
Environ Res Risk Assess (SERRA) 15(5):384–398
Pawlowsky-Glahn V, Egozcue JJ, Tolosana-Delgado R (2011) Principal balances. In: Egozcue JJ, Tolosana-
Delgado R, Ortego MI (eds) Proceedings of the 4th international workshop on compositional data
analysis (2011), CIMNE, Barcelona, Spain, ISBN 978-84-87867-76-7
Pawlowsky-Glahn V, Egozcue J, Tolosana-Delgado R (2015) Modeling and analysis of compositional data.
Wiley, Chichester
R Development Core Team (2019) R: a language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna
Reimann C, Birke M, Demetriades A, Filzmoser P, O’Connor P (eds) (2014a) Chemistry of Europe’s
agricultural soils—part A: methodology and interpretation of the GEMAS data set. Geologisches
Jahrbuch (Reihe B 102). Schweizerbarth, Hannover

123
858 Math Geosci (2021) 53:823–858

Reimann C, Birke M, Demetriades A, Filzmoser P, O’Connor P (eds) (2014b) Chemistry of Europe’s

agricultural soils—part B: general background information and further analysis of the GEMAS data
set. Geologisches Jahrbuch (Reihe B 103). Schweizerbarth, Hannover
Rousseeuw PJ, van Zomeren BC (1990) Unmasking multivariate outliers and leverage points. J Am Stat
Assoc 8:633–639
Salibian-Barrera M, Zamar R (2002) Bootstrapping robust estimates of regression. Ann Stat 30:556–582
Salibian-Barrera M, Van Aelst S, Willems G (2008) Fast and robust bootstrap. Stat Methods Appl 17:41–71
Simonoff Jeffrey S (2003) Analyzing categorical data. Springer, Berlin
Tolosana-Delgado R, van den Boogaart K (2011) Linear models with compositions in R. In: Pawlowsky-
Glahn V, Buccianti A (eds) Compositional data analysis: theory and applications. Wiley, Chichester,
pp 356–371
Van Aelst S, Willems G (2013) Fast and robust bootstrap for multivariate inference: the R package FRB. J
Stat Softw 53(3):1–32
van den Boogaart K, Tolosana-Delgado R (2013) Analyzing compositional data with R. Springer, Heidelberg
Yohai V (1987) High breakdown-point and high efficiency estimates for regression. Ann Stat 15:642–665

123