Introduction to Principal

Components and FactorAnalysis

Multivariate Analysis often starts out with
data involving a substantial number of
correlated variables.
Principal Component Analysis (PCA) is a
dimension-reduction tool that can be used to
reduce a large set of variables to a small set
that still contains most of the information in
the large set.
 Principal component analysis (PCA) is a mathematical
procedure that transforms a number of (possibly)
correlated variables into a (smaller) number of
uncorrelated variables called principal components.
The first principal component accounts for as much of
the variability in the data as possible, and each
succeeding component accounts for as much of the
remaining variability as possible.
Principal components analysis is similar to another
multivariate procedure called Factor Analysis. They are
often confused and many scientists do not understand
the difference between the two methods or what types
of analyses they are each best suited.
 Traditionally, principal component analysis is performed
on a square symmetric matrix.
It can be a SSCP matrix (pure sums of squares and
cross products), Covariance matrix (scaled sums of
squares and cross products), or Correlation matrix (sums
of squares and cross products from standardized data).
The analysis results for objects of type SSCP and
Covariance do not differ, since these objects only differ
in a global scaling factor.
A correlation matrix is used if the variances of
individual variates differ much, or if the units of
measurement of the individual variates differ.
Geometry of Principal Components
Campbell and Atchley
 Objectives of principal component analysis
PCA reduces attribute space from a larger number of
variables to a smaller number of factors and as such is a
"non-dependent" procedure (that is, it does not assume a
dependent variable is specified).

PCA is a dimensionality reduction or data compression

method. The goal is dimension reduction and there is no
guarantee that the dimensions are interpretable (a fact
often not appreciated by (amateur) statisticians).

To select a subset of variables from a larger set, based

on which original variables have the highest correlations
with the principal component.
 Factor Analysis & Principal Components
Definitions for Beginners
Principal component analysis: Factor model in which the factors are based on
summarizing the total variance. With PCA, unities are used in the diagonal of the
correlation matrix computationally implying that all the variance is common or
shared. Algorithm lacking underlying model.

Common factor analysis: Factor model explores a reduced correlation matrix. That
is, communalities (r2) are inserted on the diagonal of the correlation matrix, and the
extracted factors are based only on the common variance, with specific and error
variances excluded. Explores underlying latent structure of data. Model
assumes variability partitionable into common and unique components

Common variance: Variance shared with other variables in the factor analysis.

Specific or unique variance: Variance of each variable unique to that variable and
not explained or associated with other variables in the factor analysis.

Communality: Total amount of variance an original variable shares with all other
variables included in the analysis.

Eigenvalue: Column sum of squared loadings for a factor, i.e., the latent root. It
conceptually represents that amount of variance accounted for by a factor.
FAfB continued - 2
Sphericity test: Statistical test for the overall significance of all correlations
within a correlation matrix

Factor: Linear combination (variate) of the original variables. Factors also

represent the underlying dimensions (constructs) that summarize or account for
the original set of observed variables.

Factor loadings: Correlation between the original variables and the factors, and
the key to understanding the underlying nature of a particular factor. Squared
factor loadings indicate what percentage of the variance in an original variable is
explained by a factor.

Factor matrix: Table displaying the factor loadings of all variables on each
factor.

Factor score: Composite measure created for each observation on each factor
extracted in the factor analysis. The factor weights are used in conjunction with
the original variable values to calculate each observation's score. The factor
scores are standardized to reflect a z-score. Factor scores place each variable in
a plane of multivariate variability.
Principal components analysis (PCA): PCA seeks a linear combination of
variables such that the maximum variance is extracted from the variables.
It then removes this variance and seeks a second linear combination which
explains the maximum proportion of the remaining variance, and so on.
This is called the principal axis method and results in orthogonal
(uncorrelated) factors. PCA analyzes total (common and unique) variance.
Eigenvectors: Principal components (from PCA - principal components
analysis) reflect both common and unique variance of the variables and
may be seen as a variance-focused approach seeking to reproduce both
the total variable variance with all components and to reproduce the
correlations.
PCA is far more common than PFA, however, and it is common to use
"factors" interchangeably with "components."
The principal components are linear combinations of the original variables
weighted by their contribution to explaining the variance in a particular
orthogonal dimension
Eigenvalues: Also called characteristic roots. The eigenvalue for a
given factor measures the variance in all the variables which is
accounted for by that factor.

The ratio of eigenvalues is the ratio of explanatory importance of

the factors with respect to the variables. If a factor has a low
eigenvalue, then it is contributing little to the explanation of
variances in the variables and may be ignored as redundant with more
important factors.

Eigenvalues measure the amount of variation in the total sample

accounted for by each factor.

A factor's eigenvalue may be computed as the sum of its squared

factor loadings for all the variables.

Note that the eigenvalues associated with the unrotated and rotated
solution will differ, though their total will be the same.
Factor loadings (factor or component coefficients) : The
factor loadings, also called component loadings in PCA, are the
correlation coefficients between the variables (rows) and
factors (columns).
Analogous to Pearson's r, the squared factor loading is the
percent of variance in that variable explained by the factor.
To get the percent of variance in all the variables accounted
for by each factor, add the sum of the squared factor
loadings for that factor (column) and divide by the number of
variables. (Note the number of variables equals the sum of
their variances as the variance of a standardized variable is
1.) This is the same as dividing the factor's eigenvalue by the
number of variables.
PC scores: Also called component scores in PCA,
these scores are the scores of each case (row) on
each factor (column). To compute the factor
score for a given case for a given factor, one
takes the case's standardized score on each
variable, multiplies by the corresponding factor
loading of the variable for the given factor, and
sums these products.
 Eigenvalues of the Covariance Matrix
Eigenvalue Difference Proportion Cumulative

1 19.2196613 7.497092 0.3559 0.3559

2 11.7225694 3.7876295 0.2171 0.573
3 7.9349399 3.1810705 0.1469 0.7199
4 4.7538694 2.5056124 0.088 0.808
5 2.248257 0.2362884 0.0416 0.8496
6 2.0119686 0.498058 0.0373 0.8869
7 1.5139106 0.3555246 0.028 0.9149
Portion of PCA Analysis of 54 Amino Acid Physio-Chemical Attributes
Variable PCA1 PCA2 PCA3 PCA4
MDS3 0.130 -0.007 0.243 -0.093
MDS16 0.171 -0.110 0.174 -0.005
MDS17 -0.104 0.228 0.105 -0.001
MDS19 0.146 0.196 0.059 0.121
MDS24 -0.188 -0.003 0.089 0.106
MDS25 0.203 -0.073 0.114 -0.037
MDS29 0.105 -0.064 0.266 -0.103
MDS30 -0.073 0.239 0.097 0.092
MDS31 0.084 -0.137 -0.178 0.233
MDS32 0.097 -0.020 -0.172 -0.031
MDS36 -0.066 0.239 0.128 -0.076
MDS43 0.117 0.193 0.054 0.145
MDS44 0.061 0.174 0.018 0.161
MDS45 0.089 0.038 -0.156 -0.242
PCA Coefficient
Introduction to Factor
Analysis
Factor analysis is a statistical procedure to identify interrelationships that
exist among a large number of variables, i.e., to identify how suites of
variables are related.

Factor analysis can be used for exploratory or confirmatory purposes.

As an exploratory procedure, factor analysis is used to search for a

possible underlying structure in the variables. In confirmatory research,
the researcher evaluates how similar the actual structure of the data, as
indicated by factor analysis, is to the expected structure.

The major difference between exploratory and confirmatory factor

analysis is that researcher has formulated hypotheses about the
underlying structure of the variables when using factor analysis for
confirmatory purposes.

As an exploratory tool, factor analysis doesn't have many statistical

assumptions. The only real assumption is presence of relatedness
between the variables as represented by the correlation coefficient. If
there are no correlations, then there is no underlying structure.
 Steps in conducting a factor analysis
There are five basic factor analysis steps:
data collection and generation of the
correlation matrix
partition of variance into common and
unique components (unique may include
random error variability)
extraction of initial factor solution
rotation and interpretation
construction of scales or factor scores to
use in further analyses
 Factor Analysis & Principal Components
Definitions for Beginners
Principal component analysis: Factor model in which the factors are based on
the total variance. With PCA, unities are used in the diagonal of the correlation
matrix computationally implying that all the variance is common or shared.

Common factor analysis: Factor model explores a reduced correlation matrix.

That is, communalities (r2) are inserted on the diagonal of the correlation matrix,
and the extracted factors are based only on the common variance, with specific
and error variances excluded.

Common variance: Variance shared with other variables in the factor analysis.

Specific or unique variance: Variance of each variable unique to that variable

and not explained or associated with other variables in the factor analysis.

Communality: Total amount of variance an original variable shares with all other
variables included in the analysis.

Eigenvalue: Column sum of squared loadings for a factor; = the latent root. It
conceptually represents that amount of variance accounted for by a factor.
FFB continued - 2
Sphericity test: Statistical test for the overall significance of all correlations
within a correlation matrix

Factor: Linear combination (variate) of the original variables. Factors also

represent the underlying dimensions (constructs) that summarize or account for
the original set of observed variables.

Factor loadings: Correlation between the original variables and the factors, and
the key to understanding the nature of a particular factor. Squared factor
loadings indicate what percentage of the variance in an original variable is
explained by a factor.

Factor matrix: Table displaying the factor loadings of all variables on each
factor.

Factor score: Composite measure created for each observation on each factor
extracted in the factor analysis. The factor weights are used in conjunction with
the original variable values to calculate each observation's score. The factor
scores are standardized to according to a z-score.
FFB continued - 3
Factor rotation: Process of manipulation or adjusting the factor axes
to achieve a simpler and pragmatically more meaningful factor
solution.
Oblique factor rotation: Factor rotation computed so that the
extracted factors are correlated. Rather than arbitrarily constraining
the factor rotation to an orthogonal (90 degree angle) solution, the
oblique solution identifies the extent to which each of the factors are
correlated.
Orthogonal factor rotation: Factor rotation in which the factors are
extracted so that their axes are maintained at 90 degrees. Each factor
is independent of, or orthogonal to, all other factors. The correlation
between teh factors is determined to be zero.
VARIMAX: One of the most popular orthogonal factor rotation
methods.
 Factor Rotation

Unrotated Axes Rotated Axes

Each variable lies somewhere in the plane formed by these two factors. The
factor loadings, which represent the correlation between the factor and the
variable, can also be thought of as the variable's coordinates on this plane.
In unrotated factor solution the Factor "axes" may not line up very well with the
pattern of variables and the loadings may show no clear pattern. Factor axes can
be rotated to more closely correspond to the variables and therefore become
more meaningful. Relative relationships between variables are preserved.
The rotation can be either orthogonal or oblique
 Rotation of Factors to Simple Structure

Variable Factor 1 Factor 2 Variable Factor 1 Factor 2

1 .31 .19 1 .04 .32
2 .44 .39 2 .02 .54
3 .40 -.21 3 .47 .12
4 .18 -.30 4 .32 -.06

Factors subjected to Varimax orthogonal rotation to simple structure.

Simple structure attempts to clarify the relationships among variables by
producing factors with either very high or very low coefficients and variables
with high coefficients on only one variable.
Rotation to simple structure generally simplifies the relationships among the
variables and clarifies the interpretation of the factors.
 Summary of Factor (not PCA) Analyses
Correlation Matrix Factor matrix
Amino acid attributes I II III
c1 Ala X1 I X1 II X1 III
r12 c2 Arg X2 I X2 II X2 III
r13 r23 c3 Asn X3 I X3 II X3 III
r1n r2n r3n cn Val Xn I Xn II Xn III

Transform sequences Factor scores

C-MYC ALRDQIPELE Ala -0.158 -1.385 0.205
L-MYC ALRDQVPTLA Arg 1.096 0.726 2.467
N-MYC TLRDHVPELV Asn 1.383 -0.340 -1.065
1. Decompose variances and covariances
2. Manova and discriminant analysis
3. Model amino acid and protein behaviors
 Portion of factor pattern matrix of 54 amino acid attributes.

Amino Acid Attributes F1 F2 F3 F4 F5 Comm

Average non-bonded energy per atom 1.028 0.074 0.152 0.047 -0.079 0.982
Percentage of exposed residues 1.024 0.016 0.194 0.095 0.025 0.965
Average accessible surface area 1.005 -0.034 0.159 0.059 0.153 0.994
Residue accessible surface area 0.950 0.098 0.178 0.039 0.237 0.961
Number of hydrogen bond donors 0.809 0.021 0.122 0.021 0.357 0.808
Polarity 0.790 -0.044 -0.388 0.027 -0.092 0.956
Hydrophilicity value 0.779 -0.153 -0.333 0.213 0.023 0.862
Polar requirement 0.775 -0.128 -0.335 -0.020 -0.245 0.939
Long range non-bonded energy 0.725 -0.024 -0.394 0.189 -0.104 0.905
Negative charge 0.451 -0.218 -0.024 -0.052 -0.714 0.737
Positive charge 0.442 -0.246 -0.225 -0.085 0.708 0.730
Size 0.440 -0.112 0.811 -0.144 0.108 0.915
Normalized relative frequency of bend 0.435 0.674 -0.225 0.082 -0.118 0.912
Normalized frequency of beta-turn 0.416 0.648 -0.346 -0.019 -0.079 0.969
Molecular weight 0.363 -0.091 0.657 -0.504 -0.047 0.923
Relative mutability 0.337 -0.172 -0.183 0.297 -0.296 0.416
Normalized frequency of coil 0.271 0.863 0.028 0.123 0.073 0.860
Average volume of buried residue 0.269 -0.153 0.766 -0.340 0.016 0.928
Conformational parameter of beta-turn 0.243 0.693 -0.185 -0.439 0.078 0.837
Residue volume 0.225 -0.172 0.794 -0.292 0.036 0.946
Isoelectric point 0.224 -0.060 -0.049 0.163 0.967 0.955
Propensity to form reverse turn 0.224 -0.005 -0.433 0.319 -0.194 0.563
Chou-Fasman coil conformation 0.201 0.780 -0.338 -0.052 0.048 0.948

Factors describe latent structure of amino acid attributes