Package relaimpo

February 20, 2015

Title Relative importance of regressors in linear models
Version 2.2-2
Date 2013-09-11
Description relaimpo provides several metrics for assessing relative importance in linear models. These can be printed, plotted and bootstrapped. The recommended metric is lmg, which provides a decomposition of the model explained variance into non-negative contributions. There is a version of this package available that additionally provides a new and also recommended metric called pmvd. If you are a non-US user, you can download this extended version from Ulrike Groempings web site.
Depends R(>= 2.2.1), MASS, boot, survey, mitools, graphics
Imports methods, corpcor
Maintainer Ulrike Groemping <[email protected]>
License GPL-2
LazyLoad yes
Encoding latin1
URL http://prof.beuth-hochschule.de/groemping/relaimpo/,
http://prof.beuth-hochschule.de/groemping/
Author Ulrike Groemping [aut, cre],
Lehrkamp Matthias [ctb]
NeedsCompilation no
Repository CRAN
Date/Publication 2013-09-11 14:29:23

R topics documented:
relaimpo-package . . . .
booteval.relimp . . . . .
calc.relimp . . . . . . .
classesmethods.relaimpo
mianalyze.relimp . . . .

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
1

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

. 2
. 4
. 10
. 17
. 20

relaimpo-package
relimplm-class . . . . .
relimplmboot-class . .
relimplmbooteval-class
relimplmbootMI-class .

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

Index
relaimpo-package

24
26
27
31
35

Package to calculate relative importance metrics for linear models

Description
relaimpo calculates several relative importance metrics for the linear model. The recommended
metrics are lmg (R2 partitioned by averaging over orders, like in Lindemann, Merenda and Gold
(1980, p.119ff)) and pmvd (a newly proposed metric by Feldman (2005), non-US version only).
For completeness, several other metrics are also on offer. Other packages with related topics:
hier.part, relimp.
Details
relaimpo calculates the metrics and also offers the possibility of bootstrapping them and of displaying results in print and graphically.
It is possible to designate a subset of variables as adjustment variables that always stay in the model
so that relative importance is only assessed among the remaining variables.
Models can have up to 2-way interactions that are treated hierarchically - i.e. an interaction is only
allowed in a model that also contains all its main effects. In case of interactions, only metric lmg
can be used.
Observations with missing values are by default excluded from the analysis for most functions. The
function mianalyze.relimp allows to draw conclusions from a set of multiply imputed data sets.
This function is currently more restrictive than the rest of the package in terms of data types and
models that can be used (when summarizing the multiply imputed samples without calculating confidence intervals, all possibilities available elsewhere are also applicable in mianalyze.relimp).
relaimpo does accomodate complex survey designs by making use of the facilities in package
survey. Currently, interactions and calculated variables cannot be combined with using a complex
survey design in bootstrapping functions.
Acknowlegment
This package uses as an internal function the function nchoosek from vsn, authored by Wolfgang
Huber, available under LGPL. Furthermore, it uses a modified version of the function carscore from
care by Verena Zuber and Korbinian Strimmer.
Warning
lmg and pmvd are computer-intensive. Although they are calculated based on the covariance matrix,
which saves substantial computing time in comparison to carrying out actual regressions, these
methods still take quite long for problems with many regressors. Obviously, this is particularly
relevant in combination with bootstrapping.

relaimpo-package

Note
There are two versions of this package. The version on CRAN is globally licensed under GPL
version 2 (or later). There is an extended version with the interesting additional metric pmvd that is
licensed according to GPL version 2 under the geographical restriction "outside of the US" because
of potential issues with US patent 6,640,204. This version can be obtained from Ulrike Groempings
website (cf. references section). Whenever you load the package, a display tells you, which version
you are loading.

Author(s)
Ulrike Groemping, BHT Berlin

References
Chevan, A. and Sutherland, M. (1991) Hierarchical Partitioning. The American Statistician 45,
9096.
Darlington, R.B. (1968) Multiple regression in psychological research and practice. Psychological
Bulletin 69, 161182.
Feldman, B. (2005) Relative Importance and Value. Manuscript (Version 1.1, March 19 2005),
downloadable at http://www.prismanalytics.com/docs/RelativeImportance050319.pdf
Genizi, A. (1993) Decomposition of R2 in multiple regression with correlated regressors. Statistica Sinica 3, 407420. Downloadable at http://www3.stat.sinica.edu.tw/statistica/
password.asp?vol=3&num=2&art=10
Groemping, U. (2006) Relative Importance for Linear Regression in R: The Package relaimpo Journal of Statistical Software 17, Issue 1. Downloadable at http://www.jstatsoft.org/v17/i01
Lindeman, R.H., Merenda, P.F. and Gold, R.Z. (1980) Introduction to Bivariate and Multivariate
Analysis, Glenview IL: Scott, Foresman.
Zuber, V. and Strimmer, K. (2010) Variable importance and model selection by decorrelation.
Preprint, downloadable at http://www.uni-leipzig.de/strimmer/lab/publications/preprints/
carscore2010.pdf
Go to http://prof.beuth-hochschule.de/groemping/relaimpo/ for further information and
references.

See Also
calc.relimp, booteval.relimp, mianalyze.relimp, classesmethods.relaimpo, package hier.part,
package survey

booteval.relimp

booteval.relimp

Functions to Bootstrap Relative Importance Metrics

Description
These functions provide bootstrap confidence intervals for relative importances. boot.relimp uses
the R package boot to do the actual bootstrapping of requested metrics (which may take quite a
while), while booteval.relimp evaluates the results and provides confidence intervals. Output
from booteval.relimp is printed with a tailored print method, and a plot method produces bar
plots with confidence indication of the relative importance metrics.
Usage
## generic function
boot.relimp(object, ...)
## default S3 method
## Default S3 method:
boot.relimp(object, x = NULL, ..., b = 1000, type = "lmg",
rank = TRUE, diff = TRUE, rela = FALSE, always = NULL,
groups = NULL, groupnames = NULL, fixed=FALSE,
weights = NULL, design = NULL)
## S3 method for formula objects
## S3 method for class 'formula'
boot.relimp(formula, data, weights, na.action, ..., subset = NULL)
## S3 method for objects of class lm
## S3 method for class 'lm'
boot.relimp(object, type = "lmg", groups = NULL, groupnames=NULL, always = NULL,
..., b=1000)
## function for evaluating bootstrap results
booteval.relimp(bootrun, bty = "perc", level = 0.95,
sort = FALSE, norank = FALSE, nodiff = FALSE,
typesel = c("lmg", "pmvd", "last", "first", "betasq", "pratt", "genizi", "car"))
Arguments
object

cf. calc.relimp.

formula

cf. calc.relimp. But note the additional restriction that - in connection with
the design=-option - it is currently not possible to use factors, interactions or
calculated quantities in a formula.

cf. calc.relimp.

is the number of bootstrap runs requested on boot.relimp (default: b=1000).

Make sure to set this to a lower number, if you are simply testing code.

booteval.relimp

type

cf. calc.relimp.

rank

is a logical requesting bootstrapping of ranks (rank=TRUE, default) for each metric from type

diff

is a logical requesting bootstrapping of pairwise differences in relative importance (diff=TRUE, default) for each metric in type

rela

cf. calc.relimp.

always

cf. calc.relimp.

groups

cf. calc.relimp.

groupnames

cf. calc.relimp.

weights

cf. calc.relimp for specification of weights. See also the Details section of
this help page for usage of different types of weights.

design

cf. calc.relimp. But note that there are currently some restrictions regarding
usability of other possibilities, when using a design in boot.relimp: formulae
can only be simpler than usual, and factors, interactions or calculated variables
in formulae are not permitted.
For a description of the bootstrap methods treatment of designs, see the details
section. In the current version, using a design in bootstrapping must be considered EXPERIMENTAL.

fixed

is a logical requesting bootstrapping for a fixed design matrix (if TRUE). The
default is bootstrapping for randomly drawn samples (fixed = FALSE).

data

cf. calc.relimp.

subset

cf. calc.relimp.

na.action

cf. calc.relimp.

...

usable for further arguments, particularly all arguments of the default method
can be given to all other methods

bootrun

is an object of class relimplmboot created by function boot.relimp. It hands

over all relevant information on the bootstrap runs to function booteval.relimp.

bty

is the type of bootstrap interval requested (a character string), as handed over

to the function boot.ci from package boot. Possible choices are bca, perc,
basic and norm. student is not supported.

level

is a single confidence level or a numeric vector of confidence levels.

sort

is a logical requesting output sorted by size of relative contribution (sort=TRUE)

or by variable position in list (sort=FALSE, default).

norank

is a logical that suppresses of rank letters (norank=TRUE) even if ranks have

been bootstrapped.

nodiff

is a logical that suppresses output of confidence intervals for differences (nodiff=TRUE),

even if differences have been bootstrapped.

typesel

provides the metrics that are to be reported. Default: all available ones (intersection of those available in object bootrun and those requested in typesel).
typesel accepts the same values as type.

booteval.relimp

Details
Calculations of metrics are based on the function calc.relimp. Bootstrapping is done with the
R package boot, resampling the full observation vectors by default (combinations of response,
weights and regressors, cf. Fox (2002)). If fixed=TRUE is specified, bootstrapping is based on
residuals rather than full observations, keeping the X-variables fixed.
If the weights option is used, weights are resampled together with the full observations, and
weighted contributions are calculated for each resample (no re-normalization is done within the
resamples.)
Please note that usage of weights in linear models can have very different reasons. One motivation
is a different variability of different observations, where weights are the inverse variances. This
is the way weights are treated in function lm and also in calc.relimp and in boot.relimp, if a
vector of weights is given with the weights option. Specifically, weights do not affect the resampling probability in bootstrapping, i.e. each observation has the same probability to be included in
resamples.
If the weights in a data frame represent the multiplicity of each observation (i.e. there are several
units with identical combination of values in the data, and the weights represent the number of
units with exactly this value pattern for each row of the data frame), they can also be directly used
as weights in calc.relimp for calculating the metrics. However, such frequency weights cannot
be appropriately accomodated in boot.relimp; instead, the data frame with frequencies has to be
expanded to include one row for each unit before using the resampling routine (e.g. using function
untable from package reshape or function expand.table from package epitools.
In survey situations, weights are used to generate a more representative picture of the population: an
observations weight is typically the number of units of the population that this single observed unit
represents. In this situation, there is no reason to consider observations with higher weights as less
variable than observations with lower weights; thus, while estimates can again be obtained treating
the weights in the same way as mentioned before, their usage in estimation of standard errors and in
bootstrapping is different. In order to appropriately accomodate survey weights for these purposes,
it is not sufficient to only provide the weights vector; instead, it is necessary to provide a design
generated with package survey or an object of class svyglm (produced by function svyglm) that
includes the appropriate design information.
Clusters are a way to take care of dependency structures like in longitudinal data. Thus, while
relaimpo does not (currently) cover linear mixed models (e.g. produced by function lme), it is
possible to accomodate clustered data by applying function svyglm with linear link function and
gaussian distribution to a design that contains clusters. The bootstrapping approach subsequently
takes care of the dependence by considering clusters as sampling units. Users who want to use this
approach can mimic the second example below.
If the design option is used (experimental), resampling is done within package survey, and the resampled contributions are also calculated within package survey. The results from these calculations
are plugged into an object from package boot, and confidence interval calculation is subsequently
handled in boot. The approach is considered experimental: so far no simulation studies have been
conducted for complex survey designs, and because of limited experience (in spite of thorough
testing) it is not unlikely that bugs will be found by users who are routinely using complex survey
designs.
The output provides results for all selected relative importance metrics. The output object can be
printed and plotted (description of syntax: classesmethods.relaimpo).

booteval.relimp

Printed output: In addition to the standard output of calc.relimp (one row for each regressor,
one column for each bootstrapped metric), there is a table of confidence intervals for each selected
metric (one row per combination of regressor and metric). This table is enhanced by information
on rank confidence intervals, if ranks have been bootstrapped (rank=TRUE) and norank=FALSE. In
addition, if differences have been bootstrapped (diff=TRUE) and nodiff=FALSE, there is a table of
estimated pairwise differences with confidence intervals.
Graphical output: Application of the plot method to the object created by booteval.relimp yields
barplot representations for all bootstrapped metrics (all in one graphics window). Confidence level
(lev=) and number of characters in variable names to be used (names.abbrev=) can be modified.
Confidence bounds are indicated on the graphs by added vertical lines. par() options can be used
for modifying output (exceptions: mfrow, oma and mar are overridden by the plot method).
Value
The value of boot.relimp is of class relimplmboot. It is designed to be useful as input for
booteval.relimp and is not further described here. booteval.relimp returns an object of class
relimplmbooteval, the items of which can be accessed by the $ or the @ extractors.
In addition to the items described for function calc.relimp, which are also available here, the
following items may be of interest for further calculations:
metric.lower

matrix of lower confidence bounds for metric: one row for each confidence
level, one column for each element of metric. metric can be any of lmg,
lmg.rank, lmg.diff, ... (replace lmg with other available relative importance
metrics, cf. calc.relimp)

metric.upper

matrix of upper confidence bounds for metric: one row for each confidence
level, one column for each element of metric

metric.boot

matrix of bootstrap results for metric: one row for each bootstrap run, one
column for each element of metric. Here, metric can be chosen as any of
the above-mentioned and also as R2

nboot

number of bootstrap runs underlying the evaluations

level

confidence levels

Warning
The bootstrap confidence intervals should be used for exploratory purposes only. They can be
somewhat liberal: Limited simulations for percentile intervals have shown that non-coverage probabilities can be up to twice the nominal probabilities. More investigations are needed.
Be aware that the method itself needs some computing time in case of many regressors. Hence,
bootstrapping should be used with awareness of computing time issues.
relaimpo is a package for univariate linear models. Using relaimpo on objects that inherit from
class lm but are not univariate linear model objects may produce nonsensical results without warning. Objects of class mlm or glm with link functions other than identity or family other than gaussian
lead to an error message.

booteval.relimp

Note
There are two versions of this package. The version on CRAN is globally licensed under GPL
version 2 (or later). There is an extended version with the interesting additional metric pmvd that is
licensed according to GPL version 2 under the geographical restriction "outside of the US" because
of potential issues with US patent 6,640,204. This version can be obtained from Ulrike Groempings
website (cf. references section). Whenever you load the package, a display tells you, which version
you are loading.
Author(s)
Ulrike Groemping, BHT Berlin
References
Chevan, A. and Sutherland, M. (1991) Hierarchical Partitioning. The American Statistician 45,
9096.
Darlington, R.B. (1968) Multiple regression in psychological research and practice. Psychological
Bulletin 69, 161182.
Feldman, B. (2005) Relative Importance and Value. Manuscript (Version 1.1, March 19 2005),
downloadable at http://www.prismanalytics.com/docs/RelativeImportance050319.pdf
Fox, J. (2002) Bootstrapping regression models. An R and S-PLUS Companion to Applied Regression: A web appendix to the book. http://cran.r-project.org/doc/contrib/Fox-Companion/
appendix-bootstrapping.pdf.
Genizi, A. (1993) Decomposition of R2 in multiple regression with correlated regressors. Statistica Sinica 3, 407420. Downloadable at http://www3.stat.sinica.edu.tw/statistica/
password.asp?vol=3&num=2&art=10
Groemping, U. (2006) Relative Importance for Linear Regression in R: The Package relaimpo Journal of Statistical Software 17, Issue 1. Downloadable at http://www.jstatsoft.org/v17/i01
Lindeman, R.H., Merenda, P.F. and Gold, R.Z. (1980) Introduction to Bivariate and Multivariate
Analysis, Glenview IL: Scott, Foresman.
Zuber, V. and Strimmer, K. (2010) Variable importance and model selection by decorrelation.
Preprint, downloadable at http://www.uni-leipzig.de/strimmer/lab/publications/preprints/
carscore2010.pdf
Go to http://prof.beuth-hochschule.de/groemping/relaimpo/ for further information and
references.
See Also
relaimpo, calc.relimp, mianalyze.relimp, classesmethods.relaimpo
Examples
#####################################################################
### Example: relative importance of various socioeconomic indicators
###
for Fertility in Switzerland
### Fertility is first column of data set swiss
#####################################################################

booteval.relimp
data(swiss)
# bootstrapping
bootswiss <- boot.relimp(swiss, b = 100,
type = c("lmg", "last", "first", "pratt"),
rank = TRUE, diff = TRUE, rela = TRUE)
# for demonstration purposes only 100 bootstrap replications
#alternatively, use formula interface
bootsub <- boot.relimp(Fertility~Education+Catholic+Infant.Mortality, swiss,
subset=Catholic>40, b = 100, type = c("lmg", "last", "first", "pratt"),
rank = TRUE, diff = TRUE)
# for demonstration purposes only 100 bootstrap replications
#default output (percentily intervals, as of Version 2 of the package)
booteval.relimp(bootswiss)
plot(booteval.relimp(bootswiss))
#sorted printout, chosen confidence levels, chosen interval method
#store as object
result <- booteval.relimp(bootsub, bty="bca",
sort = TRUE, level=c(0.8,0.9))
#because of only 100 bootstrap replications,
#default bca intervals produce warnings
#output driven by print method
result
#result plotting with default settings
#(largest confidence level, names abbreviated to length 4)
plot(result)
#result plotting with modified settings (chosen confidence level,
#names abbreviated to chosen length)
plot(result, level=0.8,names.abbrev=5)
#result plotting with longer names shown vertically
par(las=2)
plot(result, level=0.9,names.abbrev=6)
#plot does react to options set with par()
#exceptions: mfrow, mar and oma are set within the plot routine itself
#####################################################################
### Example: bootstrapping clustered data
###
data taken from example.lmm of package lmm
### y is change in pulse (heart beats per minute)
###
15 min (occ 1 to 3) and 90 min (occ 4 to 6) after
###
treatment with Placebo (occ 1 or 4) low (occ 2 or 5)
###
or high (occ 3 or 6) dose of marihuana
### each of 9 subjects is observed under most or all
### of the 6 possible conditions
#####################################################################
## create example data from package lmm
y <- c(16,20,16,20,-6,-4,
12,24,12,-6,4,-8,
8,8,26,-4,4,8,
20,8,20,-4,
8,4,-8,22,-8,

10

calc.relimp
10,20,28,-20,-4,-4,
4,28,24,12,8,18,
-8,20,24,-3,8,-24,
20,24,8,12)
occ <- c(1,2,3,4,5,6,
1,2,3,4,5,6,
1,2,3,4,5,6,
1,2,5,6,
1,2,3,5,6,
1,2,3,4,5,6,
1,2,3,4,5,6,
1,2,3,4,5,6,
2,3,4,5)
subj <- c(1,1,1,1,1,1,
2,2,2,2,2,2,
3,3,3,3,3,3,
4,4,4,4,
5,5,5,5,5,
6,6,6,6,6,6,
7,7,7,7,7,7,
8,8,8,8,8,8,
9,9,9,9)
### manual creation of dummies
### reference category placebo after 90min (occ=4)
dumm1 <- as.numeric(occ==1)
dumm2 <- as.numeric(occ==2)
dumm3 <- as.numeric(occ==3)
dumm5 <- as.numeric(occ==5)
dumm6 <- as.numeric(occ==6)
## create data frame
dat <- data.frame(y,dumm1,dumm2,dumm3,dumm5,dumm6,subj)
### create design with clusters
des <- svydesign(id=~subj,data=dat)
### apply bootstrapping
### using the design with subjects as clusters implies that
###
clusters are generally kept in or excluded as a whole
### of course, b=100 is too small, only chosen for speed of package creation
bt <- boot.relimp(y~dumm1+dumm2+dumm3+dumm5+dumm6,data=dat,
design=des,b=100,type=c("lmg","first","last","betasq","pratt"))
### calculate and display results
booteval.relimp(bt,lev=0.9,nodiff=TRUE)

calc.relimp

Function to calculate relative importance metrics for linear models

calc.relimp

11

Description
calc.relimp calculates several relative importance metrics for the linear model. The recommended
metrics are lmg (R2 partitioned by averaging over orders, like in Lindemann, Merenda and Gold
(1980, p.119ff)) and pmvd (a newly proposed metric by Feldman (2005) that is provided in the nonUS version of the package only). For completeness and comparison purposes, several other metrics
are also on offer (cf. e.g. Darlington (1968)).
Usage
## generic function
calc.relimp(object, ...)
## default S3 method
## Default S3 method:
calc.relimp(object, x = NULL, ...,
type = "lmg", diff = FALSE, rank = TRUE, rela = FALSE, always = NULL,
groups = NULL, groupnames = NULL, weights=NULL, design=NULL)
## S3 method for formula object
## S3 method for class 'formula'
calc.relimp(formula, data, weights, na.action, ..., subset=NULL)
## S3 method for objects of class lm
## S3 method for class 'lm'
calc.relimp(object, type = "lmg", groups = NULL, groupnames=NULL, always = NULL, ...)

Arguments
object

The class of this object determines which of the methods is used: There are
special methods for output objects from function lm (or linear model objects
inheriting from class lm generated by other functions like glm and svyglm) and
for formula objects. For all other types of object, the default method is used.
Thus, object can be
a formula (e.g. y\~x1+x2+x3+x2:x3) (cf. below for details)
OR
the output of a linear model call (inheriting from class lm, but not mlm); output
objects from lm, glm, svyglm or aov work (if linear with identity link in case
of glms); there may be further functions that output objects inheriting from lm
which may or may not work reasonably with calc.relimp; for calc.relimp
to be appropriate, the underlying model must at least be linear!
The restrictions on usage of interactions listed under item formula below also
apply to linear model objects.
OR
the covariance matrix of a response y and regressors x, (e.g. obtained by cov(cbind(y,x)),
if y is a column vector of response values and x a corresponding matrix of regressors)

12

calc.relimp
OR
a (raw) data matrix or data frame with the response variable in the first column
OR
a response vector or one-column matrix, if x contains the corresponding matrix
or data frame of regressors.
formula

The first object, if a formula is to be given; one response only.

Interaction terms are currently limited to second-order.
Note: If several interaction terms are given, calculations may be very resource
intensive, if these are all connected (e.g. with A:B, B:C, C:D, all A,B,C,D are
connected, while with A:B, C:D, D:E there are separate groups A,B and C,D,E).
Interaction terms occurring in always do not increase resource usage (but are
only permitted if the respective main effects also occur in always).
Interactions and groups currently cannot be used simultaneously.

a (raw) data matrix or data frame containing the regressors, if object is a response vector or one-column matrix
OR
NULL, if object is anything else

type

can be a character string, character vector or list of character strings. It is the

collection of metrics that are to be calculated. Available metrics: lmg, pmvd
(non-US version only), last, first, betasq, pratt, genizi and car. For brief
sketches of their meaning cf. details section.

diff

logical; if TRUE, pairwise differences between the relative contributions are

calculated; default FALSE

rank

logical; if TRUE, ranks of regressors in terms of relative contributions are calculated; default TRUE

rela

is a logical requesting relative importances summing to 100% (rela=TRUE). If

rela is FALSE (default), some of the metrics sum to R2 (lmg, pmvd, pratt),
others do not have a meaningful sum (last, first, betasq).

always

is a vector of column numbers or names of variables to be always in the model

(adjusted for). Valid numbers are 2 to (number of regressors + 1) (1 is reserved
for the response), valid character strings are all column names of object or x
respectively that refer to regressor variables. Numbers and names cannot be
mixed.
Relative importance is only assessed for the variables not selected in always.
This option currently does not work for metrics genizi and car.

groups

is a list of vectors of column numbers or names of variables to be combined into

groups. If only one group is needed, a vector can be given. The numbers and
character strings needed are of the same form as for always.
Relative importance is only allocated between groups of regressors, no subdivision within groups is calculated. Regressors that do not occur in any group are
included as singletons. A regressor must not occur in always and in groups.
Also, groups cannot be used with a linear model or a formula in case of higher
order effects (interactions). Finally, groups only works with the four metrics
lmg, pmvd, last and first.

calc.relimp

13

groupnames

is a vector of names for the variable groups to be used for annotation of output.

weights

is a vector of case weights for the observations in the data frame (or matrix).
You can EITHER specify weights OR a design. Note that weights must not
be specified for linear model objects (since these should contain their weights as
part of the model).

design

is a design object of class survey.design (cf. package survey). You can EITHER specify a design OR weights. For calc.relimp, the design is used for
calculating weights only. Note that it is discouraged (though possible) to specify a design for a conventional linear model object (since a survey-specific linear
model should be used for survey data, cf. function svyglm).
Also note that care is needed when using subset together with design: the
subset-Option only treats the data handed directly to calc.relimp, the design
has to be equivalently treated beforehand.

data

if first object is of class formula: an optional matrix or data frame that the variables in formula and subset come from; if it is omitted, all names must be meaningful in the environment from which calc.relimp is called

subset

if first object is of class formula: an optional expression indicating the subset

of the observations of data that should be used in the fit. This can be a logical
vector, or a numeric vector indicating which observation numbers are to be included, or a character vector of the row names to be included. All (non-missing)
observations are included by default.

na.action

if first object is of class formula: an optional function that indicates what should
happen when the data contain NAs. The default is first, any na.action attribute
of data, second the setting given in the call to calc.relimp, third the na.action
setting of options. Possible choices are "na.fail", (print an error message and
terminate if there are any incomplete observations), "na.omit" or "na.exclude"
(equivalent for package relaimpo, both analyse complete cases only and print a
warning, this is also what is done the default method ).

...

usable for further arguments, particularly most arguments of default method can
be given to all other methods (exception: weights and design cannot be given to
lm-method)

Details
lmg is the R2 contribution averaged over orderings among regressors, cf. e.g. Lindeman, Merenda
and Gold 1980, p.119ff or Chevan and Sutherland (1991).
pmvd is the proportional marginal variance decomposition as proposed by Feldman (2005) (nonUS version only). It can be interpreted as a weighted average over orderings among regressors,
with data-dependent weights.
last is each variables contribution when included last, also sometimes called usefulness.
first is each variables contribution when included first, which is just the squared covariance between y and the variable.
betasq is the squared standardized coefficient.
pratt is the product of the standardized coefficient and the correlation.
genizi is the R2 decomposition according to Genizi 1993

14

calc.relimp
car is the R2 decomposition according to Zuber and Strimmer 2010, also available from package
care (squares of scores produced by function carscore
Each metric is calculated using the internal function metriccalc, e.g. lmgcalc.
Five of the metrics in calc.relimp (lmg, pmvd, pratt, genizi and car), decompose the
model R2 . calc.relimp (lmg, pmvd, pratt, genizi and car) sum to the R2 that is to be
decomposed, if rela = FALSE and to 100pct if rela = TRUE.
The other metrics also (artificially) sum to 100pct if rela = TRUE. If rela = FALSE, they
are given relative to var(y) (or the conditional variance of y after adjusting out the variables
requested in always) but do not sum to R2 .
If always requests some variables to be always in the model, these are conditioned upon (i.e.
included into the model first). Only the remaining R2 that is not explained by these variables
is decomposed among the other regressors. This currently does not work for metrics genizi
and car.
Four of the metrics, lmg, pmvd, first and last, are related to the order in which the variables
are included into the model. For these it is possible to consider the variables in groups that are
always entered into the model together.
Note that relaimpo can only provide metric lmg for models with interactions (2-way interactions only). It averages only over those orders, for which the interactions enter the model after
both their main effects.
Note that there are different types of weights, weights indicating the variability of the response (observations with a more variable responses receive a lower weight than those with a
less variable response, like in the Aitken estimator), frequency weights indicating the number
of observations with exactly the observed data pattern of the current observation, or weights
indicating the number of population units represented by the current observation (inverse sampling probability, weights typically used in survey situations). All three types of weight alike
can be handed to function calc.relimp using the weights= option. Note, however, that they
have to be treated differently for bootstrapping (cf. boot.relimp).
Data from complex surveys can be treated by providing a survey design with design=-option.
For calc.relimp, it is also sufficient to provide the weights derived from the design using the
weights=-option.
calc.relimp cannot handle data with missing values directly. It applies complete-case analysis, i.e. drops all units with any missing values by default. While this can be appropriate,
if there are only few missing values, data with more severe missingness issues need special
treatment. Package relaimpo offers the function mianalyze.relimp that handles multiplyimputed datasets (that can be created by several other R-packages). Currently, possibilities
in this function are limited due to the fact that it uses complex survey designs and bootstrapping which do not (yet) go together well with factors, interactions and calculated quantities in
formulae.

Value
var.y

the variance of the response

R2

the coefficient of determination, R2

R2.decomp

the part of the coefficient of determination that is decomposed among the variables under investigation

lmg

vector of relative contributions obtained from the lmg method, if lmg has been
requested in type

calc.relimp

15

lmg.diff

vector of pairwise differences between relative contributions obtained from the

lmg method, if lmg has been requested in type and diff=TRUE

lmg.rank

rank of the regressors relative contributions obtained from the lmg method, if
lmg has been requested in type and rank=TRUE

metric, metric.diff, metric.rank

analogous to lmg for other metrics
ave.coeffs

average coefficients for variables not not requested by always only for models
of different sizes;
note that coefficients refer to modeling residuals after adjusting out variables
listed in always (both from response and other explanatory variables)

namen

names of variables, starting with response

type

character vector of metrics available

rela

Have metrics been normalized to sum 100% ?

always

column numbers of variables always in the model; in case of factors, the column
numbers given here are not identical to those in the call to calc.relimp, but
refer to the columns of the model matrix

alwaysnam

names of variables always in the model

call

contains the call that generated the object

Warning
lmg and pmvd are computer-intensive. Although they are calculated based on the covariance matrix,
which saves substantial computing time in comparison to carrying out actual regressions, these
methods still take quite long for problems with many regressors.
relaimpo is a package for univariate linear models. Using relaimpo on objects that inherit from
class lm but are not univariate linear model objects may produce nonsensical results without warning. Objects of class mlm or glm with link functions other than identity or family other than gaussian
lead to an error message.

Note
There are two versions of this package. The version on CRAN is globally licensed under GPL
version 2 (or later). There is an extended version with the interesting additional metric pmvd that is
licensed according to GPL version 2 under the geographical restriction "outside of the US" because
of potential issues with US patent 6,640,204. This version can be obtained from Ulrike Groempings
website (cf. references section). Whenever you load the package, a display tells you, which version
you are loading.

Author(s)
Ulrike Groemping, BHT Berlin

16

calc.relimp

References
Chevan, A. and Sutherland, M. (1991) Hierarchical Partitioning. The American Statistician 45,
9096.
Darlington, R.B. (1968) Multiple regression in psychological research and practice. Psychological
Bulletin 69, 161182.
Feldman, B. (2005) Relative Importance and Value. Manuscript (Version 1.1, March 19 2005),
downloadable at http://www.prismanalytics.com/docs/RelativeImportance050319.pdf
Genizi, A. (1993) Decomposition of R2 in multiple regression with correlated regressors. Statistica Sinica 3, 407420. Downloadable at http://www3.stat.sinica.edu.tw/statistica/
password.asp?vol=3&num=2&art=10
Groemping, U. (2006) Relative Importance for Linear Regression in R: The Package relaimpo Journal of Statistical Software 17, Issue 1. Downloadable at http://www.jstatsoft.org/v17/i01
Lindeman, R.H., Merenda, P.F. and Gold, R.Z. (1980) Introduction to Bivariate and Multivariate
Analysis, Glenview IL: Scott, Foresman.
Zuber, V. and Strimmer, K. (2010) Variable importance and model selection by decorrelation.
Preprint, downloadable at http://www.uni-leipzig.de/strimmer/lab/publications/preprints/
carscore2010.pdf
Go to http://prof.beuth-hochschule.de/groemping/relaimpo/ for further information and
references.
See Also
relaimpo, booteval.relimp, mianalyze.relimp, classesmethods.relaimpo
Examples
#####################################################################
### Example: relative importance of various socioeconomic indicators
###
for Fertility in Switzerland
### Fertility is first column of data set swiss
#####################################################################
data(swiss)
calc.relimp(swiss,
type = c("lmg", "last", "first", "betasq", "pratt", "genizi", "car") )
# calculation of all available relative importance metrics
# non-US version offers the additional metric "pmvd",
# i.e. call would be
# calc.relimp(cov(swiss),
# type = c("lmg", "pmvd", "last", "first", "betasq, "pratt"),
# rela = TRUE )
## same analysis with formula or lm method and a few modified options
crf <- calc.relimp(Fertility~Agriculture+Examination+Education+Catholic+Infant.Mortality,swiss,
subset = Catholic>40,
type = c("lmg", "last", "first", "betasq", "pratt"), rela = TRUE )
crf
linmod <- lm(Fertility~Agriculture+Examination+Education+Catholic+Infant.Mortality,swiss)
crlm <- calc.relimp(linmod,
type = c("lmg", "last", "first", "betasq", "pratt", "genizi", "car"), rela = TRUE )

classesmethods.relaimpo

17

plot(crlm)
# bar plot of the relative importance metrics
#of statistical interest in this context: correlation matrix
cor(swiss)
#demonstration of conditioning on one regressor using always
calc.relimp(swiss,
type = c("lmg", "last", "first", "betasq", "pratt"), rela = FALSE,
always = "Education" )
# using calc.relimp with grouping of two regressors
# and weights (not reasonable here, purely for demo purposes)
calc.relimp(swiss,
type = c("lmg", "last", "first"), rela = FALSE,
groups = c("Education","Examination"), weights = abs(-23:23) )
# using calc.relimp with grouping of two regressors
# and a design object (not reasonable here, purely for demo purposes)
des <- svydesign(~1, data=swiss, weights=~abs(-23:23))
calc.relimp(swiss,
type = c("lmg", "last", "first"), rela = FALSE,
groups = c("Education","Examination"), groupnames ="EduExam", design = des)
# calc.relimp with factors (betasq and pratt not possible)
# (calc.relimp would not be necessary here,
# because the experiment is balanced)
calc.relimp(1/time~poison+treat,data=poisons, rela = FALSE,
type = c("lmg", "last", "first"))
# including also the interaction (lmg possible only)
calc.relimp(1/time~poison*treat,data=poisons, rela = FALSE)

classesmethods.relaimpo
Classes and Methods in Package relaimpo

Description
Output objects from package relaimpo have classes relimplm (output from calc.relimp), relimplmboot
(output from boot.relimp), relimplmbooteval (output from booteval.relimp) or relimplmbootMI.
For classes relimplm, relimplmbooteval and relimplmbootMI, there are methods for plotting
and printing, usage of which is described below. For class relimplmbootMI, there is in addition a
summary-method, which produces a less detailed output than the show / print - method. For classes
relimplm, relimplmbooteval and relimplmbootMI, there is in addition a method for extracting
slots of the class with \$.
Usage
## S3 method for class 'relimplm'

18

classesmethods.relaimpo
print(x,...,show.coeffs = ifelse(any(c("lmg", "pmvd") %in% x@type)
& is.null(x@always), TRUE, FALSE))
## S3 method for class 'relimplm'
plot(x,...,names.abbrev=4, ylim=NULL, main=NULL, cex.title=1.5)
## S3 method for class 'relimplmbooteval'
print(x,...)
## S3 method for class 'relimplmbootMI'
print(x,...)
## S3 method for class 'relimplmbooteval'
plot(x, ..., lev=max(x@level), names.abbrev=4, ylim=NULL,
main=NULL, cex.title=1.5)
## S3 method for class 'relimplmbootMI'
plot(x, ..., lev=max(x@level), names.abbrev=4, ylim=NULL,
main=NULL, cex.title=1.5)
## S3 method for class 'relimplmbootMI'
summary(object, ..., lev = max(object@level))

Arguments
x

x is an output object from package relaimpo of the required class

show.coeffs

show.coeffs, if set to FALSE, suppresses printing of averaged coefficients, which

are otherwise printed, if lmg and/or pmvd are among the metrics in x and always
is NULL. See details for a discussion of the averaged coefficients, if always is
not NULL.

object

object is an object of class relimplmbootMI (output from function mianalyze.relimp)

...

further arguments to functions

names.abbrev

names.abbrev is an integer that provides the number of characters to which the

bar labels are shortened (default: 4).

ylim

The plot routines try to use appropriate scaling. If adjustments are needed, ylim
can be used like usually on plot.

main

The plot routine uses a default title based on the reponses name. If adjustments
are desired, main can be used for specifying a different title. Note that only the
first title is affected (in case of plot.relimpbooteval, there is also a sub title that
cannot be changed).

cex.title

cex.title specifies the text size for the overall title. Thus, the par option
cex.main can be used for specifying the size of individual plot titles.

lev

lev is a numeric that provides the confidence level to be plotted or displayed

respectivly (default: maximum available confidence level; 0.5<lev<1 ).

classesmethods.relaimpo

19

Details
This documentation part describes S3 methods. In addition there are S4 methods for show which
coincide with the S3 methods for print and an S4 method for coercing objects of relimplm to lists
(of their numeric elements).
Print (and show) methods produce annotated output for calc.relimp, booteval.relimp, and
mianalyze.relimp (or the objects produced by these functions). Since version 2.1, calc.relimp
provides averaged coefficients for different sub model sizes (slot ave.coeffs of class relimplm),
if metrics based on averaging over orderings (lmg and/or pmvd) are calculated. These are per default
printed if the slot always of x is NULL. If some variables were forced into all models (non-NULL
always), the averaged coefficients refer to the adjusted model after taking residuls from regressions
on the always-columns of the X-matrix for both response and the other columns of the X-matrix.
The reason is that these could be easily and cheaply implemented into the existing code and do
correspond to sub models relevant for lmg and pmvd. Users who are interested in these coefficients,
can set option show.coeffs=TRUE in spite of non-NULL always.
The plot methods produce barplots of relative contributions, either of the metrics alone for output
objects of class relimplm from function calc.relimp, or of the metrics with lines indicating confidence intervals for output objects of class relimplmbooteval from function booteval.relimp
or relimplmbootMI from function booteval.relimp.
Most par() options can be set and should work on plot. Exceptions: mfrow, oma and mar are set
by the plot function, depending on the number of metrics to plot and the amount of annotating text
required.
The summary-method for class relimplmbootMI allows to quickly display brief output and to
change the confidence level versus the level used in the original run (with interval bounds stored in
the metric.lower and metric.upper slots and displayed by print and show methods).
Because of a defined S3-extraction method, slots of classes relimplm, relimplmbooteval and
relimplmbootMI can be extracted not only with the @ extractor but also with \$. Hence, output elements from functions calc.relimp, booteval.relimp, and mianalyze.relimp can be extracted
as though the output objects were lists.
Note that there also is an an internally-used class relimplmtest that permits the internal function
calc.relimp.default.intern to output further detail needed for usage from within other funtions.
Author(s)
Ulrike Groemping, BHT Berlin
References
Go to http://prof.tfh-berlin.de/groemping/ for further information and references.
See Also
relaimpo, calc.relimp, booteval.relimp, mianalyze.relimp, relimplm-class, relimplmboot-class,
relimplmbooteval-class, relimplmbootMI-class

20

mianalyze.relimp

mianalyze.relimp

Function to do relative importance calculations based on multiply imputed datasets

Description
The function mianalyze.relimp takes a list of imputed data frames (or matrices), calculates relative
importance metrics for each of these and combines these metrics into overall estimates with estimated variances according to the method by Rubin (1987). The output object can be summarized,
printed and plotted.
Usage
mianalyze.relimp(implist, level = 0.95, sort = FALSE, ..., b = 50, type = "lmg",
diff = TRUE, no.CI = FALSE, rela = FALSE, always = NULL, groups = NULL,
groupnames = NULL, deslist = NULL, bootlist.out = FALSE, formula = NULL,
weights = NULL, strata=NULL, ids=NULL)
Arguments
implist

list of data frames or matrices containing multiply-imputed datasets, or object

of class imputationList
If no formula is given, the first column of each data frame/matrix is assumed to
be the response variable, the other columns are regressors.
If a list of designs is also given, the variables component of each design must
consist of the necessary columns from the respective entry in implist; if no formula is given, the variables component of each design must coincide (except
for the order of columns) with the respective entry in implist.

level

is a single confidence level (between 0.5 and 1)

sort

is a logical requesting output sorted by size of relative contribution (sort=TRUE)

or by variable position in list (sort=FALSE, default).

...

Further arguments, currently none available

is the number of bootstrap runs requested on boot.relimp (default: b=50). Make

sure to set this to a higher number, if you want to subsequently use the bootlist
slot for calculating further confidence intervals with function booteval.relimp.

type

cf. calc.relimp.

diff

is a logical requesting bootstrapping of pairwise differences in relative importance (diff=TRUE, default) for each metric in type

no.CI

if set to TRUE, suppresses calculation of confidence intervals and only averages

estimated metrics from all imputed data sets in implist. Currently, no.CI = TRUE
is the only setting for which mianalyze.relimp works when using models with
factors, groups or interactions.

rela

cf. calc.relimp.

always

cf. calc.relimp.

mianalyze.relimp

21

groups

cf. calc.relimp.

groupnames

cf. calc.relimp.

deslist

is a list of design object of class survey.design (cf. package survey). You can
EITHER specify a deslist OR weights and/or strata and/or ids. Note that
the design list must contain the same data objects (in the variables element)
that are listed in implist, so that a lot of storage space is needed in case of large
datasets.
If deslist is not given, the function creates a list of designs using weights,
strata, and ids information. Whenever the required designs are simple enough
to be covered by specifying weights, strata, and ids, this is by far preferrable
in terms of storage.

bootlist.out

If TRUE, the individual bootstrap results for each multiply imputed data set are
stored in the bootlist slot of the output object (may be storage-intensive).

formula

cf. boot.relimp; NOTE: If no.CI = FALSE, i.e. confidence intervals are not
suppressed, formula has to follow the same restrictions as mentioned under item
design for boot.relimp (no calculated variables, no interaction terms, no factors), since confidence interval calculations in mianalyze.relimp are designbased, even if no deslist-option is given.

weights

is a vector of case weights for the observations in the data frame (or matrix).
You can EITHER specify weights OR a deslist. If weights is NULL, equal
weights are assumed, unless otherwise specified in deslist. For the different
types of weights and their appropriate treatment for obtaining confidence intervals, cf. the Details section of boot.relimp.

strata

is a strata request that will be handed to function svydesign for defining the
strata in a survey design (to be given to mianalyze without the \~). You can
EITHER specify strata OR a deslist. If strata is NULL, one stratum is
assumed, unless otherwise specified in deslist.

ids

is an id-request that will be handed to function svydesign for defining the clusters in a survey design (to be given to mianalyze without the \~). You can
EITHER specify ids OR a deslist. If ids is NULL, it is assumed that there
are no clusters, unless otherwise specified in deslist.

Details
Multiple imputation is a contemporary method for handling data with a substantial missing value
problem. It produces a number of completed data sets (e.g. 10) the inference from which is subsequently combined. The most frequently used way of combination is the one by Rubin: estimates
from the different completed data sets are averaged, and the variance is estimated by combining
the average over the estimated variances (within imputation variance) with an appropriately-scaled
variance between estimates, and confidence intervals are obtained by using a t-distribution with
appropriately chosen degrees of freedom.
The variance-covariance matrix of the vector of estimates for each individual completed data set is
obtained from function withReplicates in package survey based on surveys bootstrap replication
weights. On request (bootlist.out=TRUE), the underlying bootstrap resamples are also stored in
the bootlist-slot of the output object. In this case, list elements of the bootlist-slot are objects of class relimplmboot and can be processed by function booteval.relimp. This can help in

22

mianalyze.relimp
getting an impression whether the overall aggregated confidence intervals are heavily distorted towards symmetry. If such sanity-checking is intended, the default value for b should be substantially
increased.
Function mianalyze.relimp needs a list of multiply-imputed data sets or an object of class imputationList
for input. Multiply imputed data sets can - within R - be obtained from various packages. Hints
for creating lists of the form needed for mianalyze.relimp are given below for users of functions
aregImpute, mice, and amelia. Users of packages norm, cat, mix, or pan (who have managed to
operate these extremely uncomfortable packages) can of course also produce lists of imputed data
sets (only less comfortably).
For an object imp of class mids obtained from function mice in package mice, the code
lapply(as.list(1:imp$m),function(obj) complete(imp,action=obj))
produces a list of multiply-imputed data sets as needed for function mianalyze.relimp. For an
object f of class aregImpute produced by function aregImpute in package Hmisc,
lapply(as.list(1:f$m),function(obj) impute(imp,imputation=obj))
produces the required list of multiply-imputed data sets. For an object output produced by function
amelia in package Amelia, the code
output[1:output$amelia.args$m]
produces the list of multiply-imputed data sets as needed for function mianalyze.relimp.
For multiple imputation, practice is in many cases ahead of theory; this is no different with function mianalyze.relimp. Users should note that the validity of confidence intervals has only been
proven for likelihood-based analyses. Since the metrics calculated in relaimpo are not strictly
likelihood-based, the confidence intervals from function mianalyze.relimp must be considered
approximate and experimental.

Value
The value returned by function mianalyze.relimp is an object of class relimplmbootMI (if no.CI = FALSE,
default) or an object of class relimplm (if no.CI=TRUE). It can be printed, plotted and summarized
using special methods. For extracting its items, the @ or $ extractors can be used.
In addition to the items described for function calc.relimp, which are also available here, the
following items from class relimplmbootMI may be of interest for further calculations:
metric.lower

matrix of lower confidence bounds for metric: one row for each confidence
level, one column for each element of metric. metric can be any of lmg,
lmg.rank, lmg.diff, ... (replace lmg with other available relative importance
metrics, cf. calc.relimp)

metric.upper

matrix of upper confidence bounds for metric: one row for each confidence
level, one column for each element of metric

nboot

number of bootstrap runs underlying the evaluations

level

confidence level

MIresult

object of class MIresult that can be processed by the function summary.MIresult

from package survey

bootlist

only available if bootlist.out=TRUE has been chosen; list of objects of class

boot.relimp; each list element can be input to function booteval.relimp

mianalyze.relimp

23

Warning
The confidence intervals produced here should be used for exploratory purposes only. They can be
somewhat liberal and are likely to be too symmetric particularly for small data sets. The confidence
intervals produced by function mianalyze.relimp need further research into their behaviour and
are currently considered experimental.
Be aware that the methods themselves (lmg and even more pmvd) need some computing time in case
of many regressors. Hence, bootstrapping of multiple data sets should be used with awareness of
computing time issues.
Note
There are two versions of this package. The version on CRAN is globally licensed under GPL
version 2 (or later). There is an extended version with the interesting additional metric pmvd that is
licensed according to GPL version 2 under the geographical restriction "outside of the US" because
of potential issues with US patent 6,640,204. This version can be obtained from Ulrike Groempings
website (cf. references section). Whenever you load the package, a display tells you, which version
you are loading.
Author(s)
Ulrike Groemping, BHT Berlin
References
Chevan, A. and Sutherland, M. (1991) Hierarchical Partitioning. The American Statistician 45,
9096.
Darlington, R.B. (1968) Multiple regression in psychological research and practice. Psychological
Bulletin 69, 161182.
Feldman, B. (2005) Relative Importance and Value. Manuscript (Version 1.1, March 19 2005),
downloadable at http://www.prismanalytics.com/docs/RelativeImportance050319.pdf
Genizi, A. (1993) Decomposition of R2 in multiple regression with correlated regressors. Statistica Sinica 3, 407420. Downloadable at http://www3.stat.sinica.edu.tw/statistica/
password.asp?vol=3&num=2&art=10
Groemping, U. (2006) Relative Importance for Linear Regression in R: The Package relaimpo Journal of Statistical Software 17, Issue 1. Downloadable at http://www.jstatsoft.org/v17/i01
Lindeman, R.H., Merenda, P.F. and Gold, R.Z. (1980) Introduction to Bivariate and Multivariate
Analysis, Glenview IL: Scott, Foresman.
Little, R.J.A. and Rubin, D.B. (2002) Statistical Analysis with Missing Data, Wiley, New York.
Zuber, V. and Strimmer, K. (2010) Variable importance and model selection by decorrelation.
Preprint, downloadable at http://www.uni-leipzig.de/strimmer/lab/publications/preprints/
carscore2010.pdf
Go to http://prof.beuth-hochschule.de/groemping/ for further information and references.
See Also
relaimpo, calc.relimp, booteval.relimp, classesmethods.relaimpo

24

relimplm-class

Examples
## smi contains a list of 5 imputed datasets (class imputationList) from package mitools
## (first element of smi is list of data frames)
## it is not a well-suited example for relative importance but easily available for demonstrating
##
multiple imputation-related functionality
data(smi)
## obtain averaged estimates only, without confidence intervals
## works with factors and interactions
mianalyze.relimp(smi[[1]], formula = cistot ~ drkfre+sex+wave, no.CI = TRUE)
## for obtaining all individual estimates, use lapply:
smi.cr.list <- lapply(smi[[1]], function(obj) calc.relimp(cistot ~ drkfre+sex+wave, data=obj))
## display result for first individual imputed data set
smi.cr.list[[1]]
## obtain confidence intervals,
## currently only usable for models without calculated variables, factors, groups, interactions
## call without using weights, strata, clusters or a design list
mianalyze.relimp(smi[[1]], formula = cistot ~ mdrkfre+sex+wave)
## call using the id column (identical in all smi data sets) for cluster structure
ident <- smi[[1]][[1]]$id
mitest <- mianalyze.relimp(smi[[1]], formula = cistot ~ mdrkfre+sex+wave, ids=ident)
mitest
## postprocess: look at intervals with different confidence level
summary(mitest,lev=0.8)
## call with design list
deslist = lapply(smi[[1]], function(obj) svydesign(~id,strata=~sex,weights=~cistot,data=obj))
mitest <- mianalyze.relimp(smi[[1]], formula = cistot ~ mdrkfre+sex+wave, deslist=deslist,
level=c(0.8))
mitest

relimplm-class

Class relimplm

Description
This is the class of output objects from the function calc.relimp in package relaimpo. Its elements
are described in the documentation of calc.relimp.
Objects from the Class
Objects should only be created by calls to the function calc.relimp.
Slots
var.y: Object of class "numeric"
R2: Object of class "numeric"

relimplm-class

25

R2.decomp: Object of class "numeric"

lmg: Object of class "numeric"
pmvd: Object of class "numeric"
first: Object of class "numeric"
last: Object of class "numeric"
betasq: Object of class "numeric"
pratt: Object of class "numeric"
genizi: Object of class "numeric"
car: Object of class "numeric"
lmg.rank: Object of class "numeric"
pmvd.rank: Object of class "numeric"
first.rank: Object of class "numeric"
last.rank: Object of class "numeric"
betasq.rank: Object of class "numeric"
pratt.rank: Object of class "numeric"
genizi.rank: Object of class "numeric"
car.rank: Object of class "numeric"
lmg.diff: Object of class "numeric"
pmvd.diff: Object of class "numeric"
first.diff: Object of class "numeric"
last.diff: Object of class "numeric"
betasq.diff: Object of class "numeric"
pratt.diff: Object of class "numeric"
genizi.diff: Object of class "numeric"
car.diff: Object of class "numeric"
namen: Object of class "character": variable names, starting with name for response
nobs: Object of class "numeric": number of valid observations used in analysis (available only if
calculations are based on raw data)
ave.coeffs: Object of class "matrix": average coefficients for models of different sizes
type: Object of class "character": metrics that have been calculated
rela: Object of class "logical": TRUE means that metrics have been normalized to sum 100pct
always: Object of class "numintnull": variables always in the model (adjusted for), in terms of
columns in the model matrix (i.e. in case of factors in the model, there are more entries than
in the call to function calc.relimp; first possible column is 2 (1 reserved for response))
alwaysnam: Object of class "charnull": variable names of variables always in model
groupdocu: Object of class "list": information on group definitions
call: Object of class "langnull": the call that created the object

26

relimplmboot-class

Methods
This documentation section documents S4 methods only. There are also S3-methods for printing,
plotting and extracting slots with \$. For their usage, see classesmethods.relaimpo.
coerce S4-method: If an object of this class is coerced to list by as(object, "list"), the slots
are output to the list in the following order: var.y, R2, lmg, lmg.rank, lmg.diff, next three
in non-US version only: pmvd, pmvd.rank, pmvd.diff, last, last.rank, last.diff, ...,
pratt, pratt.rank, pratt.diff, namen, type, rela, always, alwaysnam
show S4-method: identical to print
Author(s)
Ulrike Groemping, BHT Berlin
See Also
calc.relimp, print.relimplm, plot.relimplm, relaimpo

relimplmboot-class

Class relimplmboot

Description
This is the class for output objects from function boot.relimp. It is needed as input to function
booteval.relimp.
Objects from the Class
Objects should only be created by calls to the function boot.relimp.
Slots
boot: Object of class "boot", output from bootstrapping
type: Object of class "character", metrics that have been bootstrapped
nboot: Object of class "numeric", number of bootstrap runs
rank: Object of class "logical", have ranks been bootstrapped ?
diff: Object of class "logical", have differences been bootstrapped ?
rela: Object of class "logical", have the metrics been forced to add up to 100pct (TRUE) or not?
fixed: Object of class "logical", has bootstrapping been done for fixed (TRUE) or random
(FALSE) regressors?
namen: Object of class "character": variable names of all variables (y and the regressors)
nobs: Object of class "numeric": number of valid observations used in calculations
always: Object of class "numintnull": variables always in the model
alwaysnam: Object of class "charnull": variable names of variables always in model

relimplmbooteval-class

27

groupdocu: Object of class "list": information on group definitions

wt: Object of class "numintnull": observation weights
vcov: Object of class "numintmatnull": variance-covariance matrix of the estimate
call: Object of class "langnull": the call that created the object
Methods
This documentation section documents S4 methods only. There is an analogous S3-method for
printing.
show S4 method: identical to print
Author(s)
Ulrike Groemping, BHT Berlin
See Also
boot.relimp, booteval.relimp, relimplmbooteval-class,relimplmbootMI-class

relimplmbooteval-class
Class relimplmbooteval

Description
Output object from function booteval.relimp, described there.
Objects from the Class
Objects should only be created by calls to function booteval.relimp.
Slots
lmg.lower: Object of class "matrix" ~~
lmg.upper: Object of class "matrix" ~~
lmg.rank.lower: Object of class "matrix" ~~
lmg.rank.upper: Object of class "matrix" ~~
lmg.diff.lower: Object of class "matrix" ~~
lmg.diff.upper: Object of class "matrix" ~~
pmvd.lower: Object of class "matrix" ~~
pmvd.upper: Object of class "matrix" ~~
pmvd.rank.lower: Object of class "matrix" ~~
pmvd.rank.upper: Object of class "matrix" ~~

28

relimplmbooteval-class
pmvd.diff.lower: Object of class "matrix" ~~
pmvd.diff.upper: Object of class "matrix" ~~
last.lower: Object of class "matrix" ~~
last.upper: Object of class "matrix" ~~
last.rank.lower: Object of class "matrix" ~~
last.rank.upper: Object of class "matrix" ~~
last.diff.lower: Object of class "matrix" ~~
last.diff.upper: Object of class "matrix" ~~
first.lower: Object of class "matrix" ~~
first.upper: Object of class "matrix" ~~
first.rank.lower: Object of class "matrix" ~~
first.rank.upper: Object of class "matrix" ~~
first.diff.lower: Object of class "matrix" ~~
first.diff.upper: Object of class "matrix" ~~
betasq.lower: Object of class "matrix" ~~
betasq.upper: Object of class "matrix" ~~
betasq.rank.lower: Object of class "matrix" ~~
betasq.rank.upper: Object of class "matrix" ~~
betasq.diff.lower: Object of class "matrix" ~~
betasq.diff.upper: Object of class "matrix" ~~
pratt.lower: Object of class "matrix" ~~
pratt.upper: Object of class "matrix" ~~
pratt.rank.lower: Object of class "matrix" ~~
pratt.rank.upper: Object of class "matrix" ~~
pratt.diff.lower: Object of class "matrix" ~~
pratt.diff.upper: Object of class "matrix" ~~
genizi.lower: Object of class "matrix" ~~
genizi.upper: Object of class "matrix" ~~
genizi.rank.lower: Object of class "matrix" ~~
genizi.rank.upper: Object of class "matrix" ~~
genizi.diff.lower: Object of class "matrix" ~~
genizi.diff.upper: Object of class "matrix" ~~
car.lower: Object of class "matrix" ~~
car.upper: Object of class "matrix" ~~
car.rank.lower: Object of class "matrix" ~~
car.rank.upper: Object of class "matrix" ~~
car.diff.lower: Object of class "matrix" ~~

relimplmbooteval-class
car.diff.upper: Object of class "matrix" ~~
var.y.boot: Object of class "numeric" ~~
R2.boot: Object of class "numeric" ~~
R2.decomp.boot: Object of class "numeric" ~~
lmg.boot: Object of class "matrix" ~~
pmvd.boot: Object of class "matrix" ~~
last.boot: Object of class "matrix" ~~
first.boot: Object of class "matrix" ~~
betasq.boot: Object of class "matrix" ~~
pratt.boot: Object of class "matrix" ~~
genizi.boot: Object of class "matrix" ~~
car.boot: Object of class "matrix" ~~
lmg.rank.boot: Object of class "matrix" ~~
pmvd.rank.boot: Object of class "matrix" ~~
last.rank.boot: Object of class "matrix" ~~
first.rank.boot: Object of class "matrix" ~~
betasq.rank.boot: Object of class "matrix" ~~
pratt.rank.boot: Object of class "matrix" ~~
genizi.rank.boot: Object of class "matrix" ~~
car.rank.boot: Object of class "matrix" ~~
lmg.diff.boot: Object of class "matrix" ~~
pmvd.diff.boot: Object of class "matrix" ~~
last.diff.boot: Object of class "matrix" ~~
first.diff.boot: Object of class "matrix" ~~
betasq.diff.boot: Object of class "matrix" ~~
pratt.diff.boot: Object of class "matrix" ~~
genizi.diff.boot: Object of class "matrix" ~~
car.diff.boot: Object of class "matrix" ~~
est: Object of class "numintnull" ~~
vcov: Object of class "numintmatnull" ~~
level: Object of class "numeric" ~~
nboot: Object of class "numeric" ~~
diffnam: Object of class "character" ~~
rank: Object of class "logical" ~~
diff: Object of class "logical" ~~
rela: Object of class "logical" ~~
fixed: Object of class "logical" ~~

29

30

relimplmbooteval-class
type: Object of class "character" ~~
sort: Object of class "logical" ~~
bty: Object of class "character" ~~
mark: Object of class "matrix" ~~
markdiff: Object of class "matrix" ~~
var.y: Object of class "numeric" ~~
R2: Object of class "numeric" ~~
R2.decomp: Object of class "numeric" ~~
lmg: Object of class "numeric" ~~
pmvd: Object of class "numeric" ~~
first: Object of class "numeric" ~~
last: Object of class "numeric" ~~
betasq: Object of class "numeric" ~~
pratt: Object of class "numeric" ~~
genizi: Object of class "numeric" ~~
car: Object of class "numeric" ~~
lmg.rank: Object of class "numeric" ~~
pmvd.rank: Object of class "numeric" ~~
first.rank: Object of class "numeric" ~~
last.rank: Object of class "numeric" ~~
betasq.rank: Object of class "numeric" ~~
pratt.rank: Object of class "numeric" ~~
genizi.rank: Object of class "numeric" ~~
car.rank: Object of class "numeric" ~~
lmg.diff: Object of class "numeric" ~~
pmvd.diff: Object of class "numeric" ~~
first.diff: Object of class "numeric" ~~
last.diff: Object of class "numeric" ~~
betasq.diff: Object of class "numeric" ~~
pratt.diff: Object of class "numeric" ~~
genizi.diff: Object of class "numeric" ~~
car.diff: Object of class "numeric" ~~
namen: Object of class "character" ~~
nobs: Object of class "numeric" ~~
ave.coeffs: Object of class "numintmatnull" ~~
always: Object of class "numintnull" ~~
alwaysnam: Object of class "charnull" ~~
groupdocu: Object of class "list" ~~
call: Object of class "langnull" ~~

relimplmbootMI-class

31

Extends
Class "relimplm", directly.
Methods
This documentation section documents S4 methods only. There are also S3-methods for printing,
plotting and extracting slots with \$. For their usage, see classesmethods.relaimpo.
show S4 method: identical to print
Author(s)
Ulrike Groemping, BHT Berlin
See Also
relimplm-class, relimplmboot-class, relimplmbootMI-class, booteval.relimp, print.relimplmbooteval,
plot.relimplmbooteval

relimplmbootMI-class

Class relimplmbootMI

Description
Output object from function mianalyze.relimp, described there.
Objects from the Class
Objects should only be created by calls to function mianalyze.relimp.
Slots
lmg.lower: Object of class "matrix"
lmg.upper: Object of class "matrix"
lmg.rank.lower: Object of class "matrix"
lmg.rank.upper: Object of class "matrix"
lmg.diff.lower: Object of class "matrix"
lmg.diff.upper: Object of class "matrix"
pmvd.lower: Object of class "matrix"
pmvd.upper: Object of class "matrix"
pmvd.rank.lower: Object of class "matrix"
pmvd.rank.upper: Object of class "matrix"
pmvd.diff.lower: Object of class "matrix"
pmvd.diff.upper: Object of class "matrix"

32

relimplmbootMI-class
last.lower: Object of class "matrix"
last.upper: Object of class "matrix"
last.rank.lower: Object of class "matrix"
last.rank.upper: Object of class "matrix"
last.diff.lower: Object of class "matrix"
last.diff.upper: Object of class "matrix"
first.lower: Object of class "matrix"
first.upper: Object of class "matrix"
first.rank.lower: Object of class "matrix"
first.rank.upper: Object of class "matrix"
first.diff.lower: Object of class "matrix"
first.diff.upper: Object of class "matrix"
betasq.lower: Object of class "matrix"
betasq.upper: Object of class "matrix"
betasq.rank.lower: Object of class "matrix"
betasq.rank.upper: Object of class "matrix"
betasq.diff.lower: Object of class "matrix"
betasq.diff.upper: Object of class "matrix"
pratt.lower: Object of class "matrix"
pratt.upper: Object of class "matrix"
pratt.rank.lower: Object of class "matrix"
pratt.rank.upper: Object of class "matrix"
pratt.diff.lower: Object of class "matrix"
pratt.diff.upper: Object of class "matrix"
genizi.lower: Object of class "matrix"
genizi.upper: Object of class "matrix"
genizi.rank.lower: Object of class "matrix"
genizi.rank.upper: Object of class "matrix"
genizi.diff.lower: Object of class "matrix"
genizi.diff.upper: Object of class "matrix"
car.lower: Object of class "matrix"
car.upper: Object of class "matrix"
car.rank.lower: Object of class "matrix"
car.rank.upper: Object of class "matrix"
car.diff.lower: Object of class "matrix"
car.diff.upper: Object of class "matrix"
MIresult: Object of class "MIresult"

relimplmbootMI-class
est: Object of class "numintnull"
vcov: Object of class "nummatnull"
bootlist: Object of class "listnull", list of objects of class relimplmboot or null
level: Object of class "numeric"
nboot: Object of class "numeric"
diffnam: Object of class "character"
rank: Object of class "logical"
diff: Object of class "logical"
rela: Object of class "logical"
fixed: Object of class "logical"
type: Object of class "character"
sort: Object of class "logical"
bty: Object of class "character"
mark: Object of class "matrix"
markdiff: Object of class "matrix"
var.y: Object of class "numeric"
R2: Object of class "numeric"
R2.decomp: Object of class "numeric"
lmg: Object of class "numeric"
pmvd: Object of class "numeric"
first: Object of class "numeric"
last: Object of class "numeric"
betasq: Object of class "numeric"
pratt: Object of class "numeric"
genizi: Object of class "numeric"
car: Object of class "numeric"
lmg.rank: Object of class "numeric"
pmvd.rank: Object of class "numeric"
first.rank: Object of class "numeric"
last.rank: Object of class "numeric"
betasq.rank: Object of class "numeric"
pratt.rank: Object of class "numeric"
genizi.rank: Object of class "numeric"
car.rank: Object of class "numeric"
lmg.diff: Object of class "numeric"
pmvd.diff: Object of class "numeric"
first.diff: Object of class "numeric"

33

34

relimplmbootMI-class
last.diff: Object of class "numeric"
betasq.diff: Object of class "numeric"
pratt.diff: Object of class "numeric"
genizi.diff: Object of class "numeric"
car.diff: Object of class "numeric"
namen: Object of class "character"
nobs: Object of class "numeric": number of valid observations used in analysis
always: Object of class "numintnull": variables always in the model
alwaysnam: Object of class "charnull": variable names of variables always in model

Extends
Class "relimplm", directly.
Methods
This documentation section documents S4 methods only. There are also S3-methods for printing,
plotting, extracting slots with \$ and producing a brief summary. For their usage, see classesmethods.relaimpo.
show S4-method: identical to print
to the slot MIresult for a briefer output than the standard printout
Author(s)
Ulrike Groemping, BHT Berlin
See Also
relimplm-class, relimplmbooteval-class, booteval.relimp, print.relimplmbootMI, plot.relimplmbootMI,
summary.relimplmbootMI

Index
coerce,relimplm,list-method
(relimplm-class), 24

Topic classes
relimplm-class, 24
relimplmboot-class, 26
relimplmbooteval-class, 27
relimplmbootMI-class, 31
Topic htest
booteval.relimp, 4
classesmethods.relaimpo, 17
mianalyze.relimp, 20
relaimpo-package, 2
Topic methods
relimplm-class, 24
Topic models
booteval.relimp, 4
calc.relimp, 10
classesmethods.relaimpo, 17
mianalyze.relimp, 20
relaimpo-package, 2
Topic multivariate
booteval.relimp, 4
calc.relimp, 10
classesmethods.relaimpo, 17
mianalyze.relimp, 20
relaimpo-package, 2
$.relimplm (classesmethods.relaimpo), 17
$.relimplmbootMI
(classesmethods.relaimpo), 17
$.relimplmbooteval
(classesmethods.relaimpo), 17

glm, 11
lm, 11
mianalyze.relimp, 2, 3, 8, 16, 19, 20, 31
plot.relimplm, 26
plot.relimplm
(classesmethods.relaimpo), 17
plot.relimplmbooteval, 31
plot.relimplmbooteval
(classesmethods.relaimpo), 17
plot.relimplmbootMI, 34
plot.relimplmbootMI
(classesmethods.relaimpo), 17
print.relimplm, 26
print.relimplm
(classesmethods.relaimpo), 17
print.relimplmbooteval, 31
print.relimplmbooteval
(classesmethods.relaimpo), 17
print.relimplmbootMI, 34
print.relimplmbootMI
(classesmethods.relaimpo), 17
relaimpo, 8, 16, 19, 23, 26
relaimpo (relaimpo-package), 2
relaimpo-package, 2
relimplm, 31
relimplm (classesmethods.relaimpo), 17
relimplm-class, 24
relimplmboot (classesmethods.relaimpo),
17
relimplmboot-class, 26
relimplmbooteval
(classesmethods.relaimpo), 17
relimplmbooteval-class, 27

boot.relimp, 14, 21, 27

boot.relimp (booteval.relimp), 4
booteval.relimp, 3, 4, 16, 19, 22, 23, 27, 31,
34
calc.relimp, 35, 7, 8, 10, 1924, 26
carscore, 14
classesmethods.relaimpo, 3, 6, 8, 16, 17,
23, 26, 31, 34
35

36
relimplmbootMI
(classesmethods.relaimpo), 17
relimplmbootMI-class, 31
relimplmtest (classesmethods.relaimpo),
17
show,relimplm-method (relimplm-class),
24
show,relimplmboot-method
(relimplmboot-class), 26
show,relimplmbooteval-method
(relimplmbooteval-class), 27
show,relimplmbootMI-method
(relimplmbootMI-class), 31
summary.relimplmbootMI, 34
summary.relimplmbootMI
(classesmethods.relaimpo), 17

INDEX