Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Support
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
P
plsgenomics
Project overview
Project overview
Details
Activity
Releases
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
0
Issues
0
List
Boards
Labels
Milestones
Merge Requests
0
Merge Requests
0
Packages
Packages
Container Registry
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Commits
Issue Boards
Open sidebar
DURIF Ghislain
plsgenomics
Commits
26f1de07
Commit
26f1de07
authored
Feb 19, 2019
by
GD
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
automatic generation of the doc
parent
23f5e65a
Changes
13
Hide whitespace changes
Inline
Sidebyside
Showing
13 changed files
with
104 additions
and
82 deletions
+104
82
pkg/man/logit.spls.Rd
pkg/man/logit.spls.Rd
+7
6
pkg/man/logit.spls.cv.Rd
pkg/man/logit.spls.cv.Rd
+7
6
pkg/man/logit.spls.stab.Rd
pkg/man/logit.spls.stab.Rd
+8
7
pkg/man/multinom.spls.Rd
pkg/man/multinom.spls.Rd
+11
10
pkg/man/multinom.spls.cv.Rd
pkg/man/multinom.spls.cv.Rd
+7
6
pkg/man/multinom.spls.stab.Rd
pkg/man/multinom.spls.stab.Rd
+8
7
pkg/man/sample.bin.Rd
pkg/man/sample.bin.Rd
+8
7
pkg/man/sample.multinom.Rd
pkg/man/sample.multinom.Rd
+7
6
pkg/man/spls.Rd
pkg/man/spls.Rd
+7
6
pkg/man/spls.cv.Rd
pkg/man/spls.cv.Rd
+7
6
pkg/man/spls.stab.Rd
pkg/man/spls.stab.Rd
+8
7
pkg/man/stability.selection.Rd
pkg/man/stability.selection.Rd
+16
5
pkg/man/stability.selection.heatmap.Rd
pkg/man/stability.selection.heatmap.Rd
+3
3
No files found.
pkg/man/logit.spls.Rd
View file @
26f1de07
...
...
@@ 99,14 +99,14 @@ new observations in \code{Xtest}, that are used to predict the
\description{
The function \code{logit.spls} performs compression and variable selection
in the context of binary classification (with possible prediction)
using Durif et al. (201
7
) algorithm based on Ridge IRLS and sparse PLS.
using Durif et al. (201
8
) algorithm based on Ridge IRLS and sparse PLS.
}
\details{
The columns of the data matrices \code{Xtrain} and \code{Xtest} may
not be standardized, since standardizing can be performed by the function
\code{logit.spls} as a preliminary step.
The procedure described in Durif et al. (201
7
) is used to compute
The procedure described in Durif et al. (201
8
) is used to compute
latent sparse components that are used in a logistic regression model.
In addition, when a matrix \code{Xtest} is supplied, the procedure
predicts the response associated to these new values of the predictors.
...
...
@@ 147,10 +147,11 @@ sum(model1$hatYtest!=Ytest) / length(index.test)
}
\references{
Durif G., Modolo L., Michaelsson J., Mold J. E., LambertLacroix S.,
Picard F. (2017). High Dimensional Classification with combined Adaptive
Sparse PLS and Logistic Regression, (in prep),
available on (\url{http://arxiv.org/abs/1502.05933}).
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
}
\seealso{
\code{\link{spls}}, \code{\link{logit.spls.cv}}
...
...
pkg/man/logit.spls.cv.Rd
View file @
26f1de07
...
...
@@ 88,14 +88,14 @@ error rate averaged over the folds. \code{cv.grid} is NULL if
The
function
\
code
{
logit
.
spls
.
cv
}
chooses
the
optimal
values
for
the
hyper

parameter
of
the
\
code
{
logit
.
spls
}
procedure
,
by
minimizing
the
averaged
error
of
prediction
over
the
hyper

parameter
grid
,
using
Durif
et
al
.
(
201
7
)
LOGIT

SPLS
algorithm
.
using
Durif
et
al
.
(
201
8
)
LOGIT

SPLS
algorithm
.
}
\
details
{
The
columns
of
the
data
matrices
\
code
{
X
}
may
not
be
standardized
,
since
standardizing
is
performed
by
the
function
\
code
{
logit
.
spls
.
cv
}
as
a
preliminary
step
.
The
procedure
is
described
in
Durif
et
al
.
(
201
7
).
The
K

fold
The
procedure
is
described
in
Durif
et
al
.
(
201
8
).
The
K

fold
cross

validation
can
be
summarize
as
follow
:
the
train
set
is
partitioned
into
K
folds
,
for
each
value
of
hyper

parameters
the
model
is
fit
K
times
,
using
each
fold
to
compute
the
prediction
error
rate
,
and
fitting
the
...
...
@@ 141,10 +141,11 @@ str(cv1)
}
\
references
{
Durif
G
.,
Modolo
L
.,
Michaelsson
J
.,
Mold
J
.
E
.,
Lambert

Lacroix
S
.,
Picard
F
.
(
2017
).
High
Dimensional
Classification
with
combined
Adaptive
Sparse
PLS
and
Logistic
Regression
,
(
in
prep
),
available
on
(\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}).
Durif
,
G
.,
Modolo
,
L
.,
Michaelsson
,
J
.,
Mold
,
J
.
E
.,
Lambert

Lacroix
,
S
.,
Picard
,
F
.,
2018.
High
dimensional
classification
with
combined
adaptive
sparse
PLS
and
logistic
regression
.
Bioinformatics
34
,
485

493.
\
url
{
https
://
doi
.
org
/
10.1093
/
bioinformatics
/
btx571
}.
Available
at
\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}.
}
\
seealso
{
\
code
{\
link
{
logit
.
spls
}},
\
code
{\
link
{
logit
.
spls
.
stab
}}
...
...
pkg/man/logit.spls.stab.Rd
View file @
26f1de07
...
...
@@ 67,7 +67,7 @@ the seed for pseudorandom number generation is set accordingly.}
\value{
An object with the following attributes
\item{q.Lambda}{A table with values of q.Lambda (c.f. Durif
et al. (201
7
) for the notation), being the averaged number of covariates
et al. (201
8
) for the notation), being the averaged number of covariates
selected among the entire grid of hyperparameters candidates values,
for increasing size of hyperparameter grid.}
\item{probs.lambda}{A table with estimated probability of selection for each
...
...
@@ 81,7 +81,7 @@ candidate values \code{(ncomp, lambda.l1, lambda.ridge)} of hyperparameters
on multiple subsamplings in the data. The stability selection procedure
selects the covariates that are selected by most of the models among the
grid of hyperparameters, following the procedure described in
Durif et al. (201
7
). Candidates values for \code{ncomp}, \code{lambda.l1}
Durif et al. (201
8
). Candidates values for \code{ncomp}, \code{lambda.l1}
and \code{lambda.l2} are respectively given by
the input arguments \code{ncomp.range}, \code{lambda.l1.range}
and \code{lambda.l2.range}.
...
...
@@ 91,7 +91,7 @@ The columns of the data matrices \code{X} may not be standardized,
since standardizing is performed by the function \code{logit.spls.stab}
as a preliminary step.
The procedure is described in Durif et al. (201
7
). The stability selection
The procedure is described in Durif et al. (201
8
). The stability selection
procedure can be summarize as follow (c.f. Meinshausen and Buhlmann, 2010).
(i) For each candidate values \code{(ncomp, lambda.l1, lambda.ridge)} of
...
...
@@ 155,10 +155,11 @@ stability.selection(stab1, piThreshold=0.6, rhoError=10)
}
\references{
Durif G., Modolo L., Michaelsson J., Mold J. E., LambertLacroix S.,
Picard F. (2017). High Dimensional Classification with combined Adaptive
Sparse PLS and Logistic Regression, (in prep),
available on (\url{http://arxiv.org/abs/1502.05933}).
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
Meinshausen, N., Buhlmann P. (2010). Stability Selection. Journal of the
Royal Statistical Society: Series B (Statistical Methodology)
...
...
pkg/man/multinom.spls.Rd
View file @
26f1de07
...
...
@@ 82,18 +82,18 @@ converge in less than \code{maxIter} iterations or not.}
\
item
{
X
.
score
}{
list
of
nclass

1
different
(
n
x
ncomp
)
matrices
being
the
observations
coordinates
or
scores
in
the
new
component
basis
produced
for
each
class
in
the
multinomial
model
by
the
SPLS
step
(
sparse
PLS
),
see
Durif
et
al
.
(
201
7
)
for
details
.}
see
Durif
et
al
.
(
201
8
)
for
details
.}
\
item
{
X
.
weight
}{
list
of
nclass

1
different
(
p
x
ncomp
)
matrices
being
the
coefficients
of
predictors
in
each
components
produced
for
each
class
in
the
multinomial
model
by
the
sparse
PLS
,
see
Durif
et
al
.
(
201
7
)
for
details
.}
see
Durif
et
al
.
(
201
8
)
for
details
.}
\
item
{
X
.
score
.
full
}{
a
((
n
x
(
nclass

1
))
x
ncomp
)
matrix
being
the
observations
coordinates
or
scores
in
the
new
component
basis
produced
by
the
SPLS
step
(
sparse
PLS
)
in
the
linearized
multinomial
model
,
see
Durif
et
al
.
(
201
7
).
Each
column
t
.
k
of
\
code
{
X
.
score
}
is
a
SPLS
component
.}
Durif
et
al
.
(
201
8
).
Each
column
t
.
k
of
\
code
{
X
.
score
}
is
a
SPLS
component
.}
\
item
{
X
.
weight
.
full
}{
a
(
p
x
ncomp
)
matrix
being
the
coefficients
of
predictors
in
each
components
produced
by
sparse
PLS
in
the
linearized
multinomial
model
,
see
Durif
et
al
.
(
201
7
).
Each
column
w
.
k
of
model
,
see
Durif
et
al
.
(
201
8
).
Each
column
w
.
k
of
\
code
{
X
.
weight
}
verifies
t
.
k
=
Xtrain
x
w
.
k
(
as
a
matrix
product
).}
\
item
{
lambda
.
ridge
}{
the
Ridge
hyper

parameter
used
to
fit
the
model
.}
\
item
{
lambda
.
l1
}{
the
sparse
hyper

parameter
used
to
fit
the
model
.}
...
...
@@ 111,7 +111,7 @@ new observations in \code{Xtest}, that are used to predict the
\
description
{
The
function
\
code
{
multinom
.
spls
}
performs
compression
and
variable
selection
in
the
context
of
multi

label
(
'nclass'
>
2
)
classification
(
with
possible
prediction
)
using
Durif
et
al
.
(
201
7
)
algorithm
(
with
possible
prediction
)
using
Durif
et
al
.
(
201
8
)
algorithm
based
on
Ridge
IRLS
and
sparse
PLS
.
}
\
details
{
...
...
@@ 119,7 +119,7 @@ The columns of the data matrices \code{Xtrain} and \code{Xtest} may
not
be
standardized
,
since
standardizing
can
be
performed
by
the
function
\
code
{
multinom
.
spls
}
as
a
preliminary
step
.
The
procedure
described
in
Durif
et
al
.
(
201
7
)
is
used
to
compute
The
procedure
described
in
Durif
et
al
.
(
201
8
)
is
used
to
compute
latent
sparse
components
that
are
used
in
a
multinomial
regression
model
.
In
addition
,
when
a
matrix
\
code
{
Xtest
}
is
supplied
,
the
procedure
predicts
the
response
associated
to
these
new
values
of
the
predictors
.
...
...
@@ 161,10 +161,11 @@ sum(model1$hatYtest!=Ytest) / length(index.test)
}
\
references
{
Durif
G
.,
Modolo
L
.,
Michaelsson
J
.,
Mold
J
.
E
.,
Lambert

Lacroix
S
.,
Picard
F
.
(
2017
).
High
Dimensional
Classification
with
combined
Adaptive
Sparse
PLS
and
Logistic
Regression
,
(
in
prep
),
available
on
(\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}).
Durif
,
G
.,
Modolo
,
L
.,
Michaelsson
,
J
.,
Mold
,
J
.
E
.,
Lambert

Lacroix
,
S
.,
Picard
,
F
.,
2018.
High
dimensional
classification
with
combined
adaptive
sparse
PLS
and
logistic
regression
.
Bioinformatics
34
,
485

493.
\
url
{
https
://
doi
.
org
/
10.1093
/
bioinformatics
/
btx571
}.
Available
at
\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}.
}
\
seealso
{
\
code
{\
link
{
spls
}},
\
code
{\
link
{
logit
.
spls
}},
...
...
pkg/man/multinom.spls.cv.Rd
View file @
26f1de07
...
...
@@ 89,14 +89,14 @@ error rate averaged over the folds. \code{cv.grid} is NULL if
The
function
\
code
{
multinom
.
spls
.
cv
}
chooses
the
optimal
values
for
the
hyper

parameter
of
the
\
code
{
multinom
.
spls
}
procedure
,
by
minimizing
the
averaged
error
of
prediction
over
the
hyper

parameter
grid
,
using
Durif
et
al
.
(
201
7
)
multinomial

SPLS
algorithm
.
using
Durif
et
al
.
(
201
8
)
multinomial

SPLS
algorithm
.
}
\
details
{
The
columns
of
the
data
matrices
\
code
{
X
}
may
not
be
standardized
,
since
standardizing
is
performed
by
the
function
\
code
{
multinom
.
spls
.
cv
}
as
a
preliminary
step
.
The
procedure
is
described
in
Durif
et
al
.
(
201
7
).
The
K

fold
The
procedure
is
described
in
Durif
et
al
.
(
201
8
).
The
K

fold
cross

validation
can
be
summarize
as
follow
:
the
train
set
is
partitioned
into
K
folds
,
for
each
value
of
hyper

parameters
the
model
is
fit
K
times
,
using
each
fold
to
compute
the
prediction
error
rate
,
and
fitting
the
...
...
@@ 143,10 +143,11 @@ str(cv1)
}
\
references
{
Durif
G
.,
Modolo
L
.,
Michaelsson
J
.,
Mold
J
.
E
.,
Lambert

Lacroix
S
.,
Picard
F
.
(
2017
).
High
Dimensional
Classification
with
combined
Adaptive
Sparse
PLS
and
Logistic
Regression
,
(
in
prep
),
available
on
(\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}).
Durif
,
G
.,
Modolo
,
L
.,
Michaelsson
,
J
.,
Mold
,
J
.
E
.,
Lambert

Lacroix
,
S
.,
Picard
,
F
.,
2018.
High
dimensional
classification
with
combined
adaptive
sparse
PLS
and
logistic
regression
.
Bioinformatics
34
,
485

493.
\
url
{
https
://
doi
.
org
/
10.1093
/
bioinformatics
/
btx571
}.
Available
at
\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}.
}
\
seealso
{
\
code
{\
link
{
multinom
.
spls
}},
\
code
{\
link
{
multinom
.
spls
.
stab
}}
...
...
pkg/man/multinom.spls.stab.Rd
View file @
26f1de07
...
...
@@ 68,7 +68,7 @@ the seed for pseudorandom number generation is set accordingly.}
\value{
An object with the following attributes
\item{q.Lambda}{A table with values of q.Lambda (c.f. Durif
et al. (201
7
) for the notation), being the averaged number of covariates
et al. (201
8
) for the notation), being the averaged number of covariates
selected among the entire grid of hyperparameters candidates values,
for increasing size of hyperparameter grid.}
\item{probs.lambda}{A table with estimated probability of selection for each
...
...
@@ 82,7 +82,7 @@ each candidate values \code{(ncomp, lambda.l1, lambda.ridge)} of
hyperparameters on multiple subsamplings in the data. The stability
selection procedure selects the covariates that are selected by most of the
models among the grid of hyperparameters, following the procedure
described in Durif et al. (201
7
). Candidates values for \code{ncomp},
described in Durif et al. (201
8
). Candidates values for \code{ncomp},
\code{lambda.l1} and \code{lambda.l2} are respectively given by
the input arguments \code{ncomp.range}, \code{lambda.l1.range}
and \code{lambda.l2.range}.
...
...
@@ 92,7 +92,7 @@ The columns of the data matrices \code{X} may not be standardized,
since standardizing is performed by the function \code{multinom.spls.stab}
as a preliminary step.
The procedure is described in Durif et al. (201
7
). The stability selection
The procedure is described in Durif et al. (201
8
). The stability selection
procedure can be summarize as follow (c.f. Meinshausen and Buhlmann, 2010).
(i) For each candidate values \code{(ncomp, lambda.l1, lambda.ridge)} of
...
...
@@ 160,10 +160,11 @@ stability.selection(stab1, piThreshold=0.6, rhoError=10)
}
\references{
Durif G., Modolo L., Michaelsson J., Mold J. E., LambertLacroix S.,
Picard F. (2017). High Dimensional Classification with combined Adaptive
Sparse PLS and Logistic Regression, (in prep),
available on (\url{http://arxiv.org/abs/1502.05933}).
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
Meinshausen, N., Buhlmann P. (2010). Stability Selection. Journal of the
Royal Statistical Society: Series B (Statistical Methodology)
...
...
pkg/man/sample.bin.Rd
View file @
26f1de07
...
...
@@ 5,8 +5,8 @@
\title{Generates covariate matrix X with correlated block of covariates and
a binary random reponse depening on X through a logistic model}
\usage{
sample.bin(n, p, kstar, lstar, beta.min, beta.max, mean.H = 0, sigma.H,
sigma.F, seed = NULL)
sample.bin(n, p, kstar, lstar, beta.min, beta.max, mean.H = 0, sigma.H
= 1
,
sigma.F
= 1
, seed = NULL)
}
\arguments{
\item{n}{the number of observations in the sample.}
...
...
@@ 94,7 +94,7 @@ the ones with null coefficients are not.
The response is generated as by drawing one observation of n different
Bernoulli random variables of parameters logit^\{1\}(XB).
The details of the procedure are developped by Durif et al. (201
7
).
The details of the procedure are developped by Durif et al. (201
8
).
}
\examples{
### load plsgenomics library
...
...
@@ 111,10 +111,11 @@ str(sample1)
}
\references{
Durif G., Modolo L., Michaelsson J., Mold J. E., LambertLacroix S.,
Picard F. (2017). High Dimensional Classification with combined Adaptive
Sparse PLS and Logistic Regression, (in prep),
available on (\url{http://arxiv.org/abs/1502.05933}).
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
}
\seealso{
\code{\link{sample.cont}}
...
...
pkg/man/sample.multinom.Rd
View file @
26f1de07
...
...
@@ 6,7 +6,7 @@
a multilabel random reponse depening on X through a multinomial model}
\usage{
sample.multinom(n, p, nb.class = 2, kstar, lstar, beta.min, beta.max,
mean.H = 0, sigma.H
, sigma.F
, seed = NULL)
mean.H = 0, sigma.H
= 1, sigma.F = 1
, seed = NULL)
}
\arguments{
\item{n}{the number of observations in the sample.}
...
...
@@ 96,7 +96,7 @@ the ones with null coefficients are not.
The response is generated as by drawing one observation of n different
Bernoulli random variables of parameters logit^\{1\}(XB).
The details of the procedure are developped by Durif et al. (201
7
).
The details of the procedure are developped by Durif et al. (201
8
).
}
\examples{
### load plsgenomics library
...
...
@@ 115,10 +115,11 @@ str(sample1)
}
\references{
Durif G., Modolo L., Michaelsson J., Mold J. E., LambertLacroix S.,
Picard F. (2017). High Dimensional Classification with combined Adaptive
Sparse PLS and Logistic Regression, (in prep),
available on (\url{http://arxiv.org/abs/1502.05933}).
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
}
\seealso{
\code{\link{sample.cont}}
...
...
pkg/man/spls.Rd
View file @
26f1de07
...
...
@@ 121,14 +121,14 @@ step was adaptive or not.}
\description{
The function \code{spls.adapt} performs compression and variable selection
in the context of linear regression (with possible prediction)
using Durif et al. (201
7
) adaptive SPLS algorithm.
using Durif et al. (201
8
) adaptive SPLS algorithm.
}
\details{
The columns of the data matrices \code{Xtrain} and \code{Xtest} may
not be standardized, since standardizing can be performed by the function
\code{spls} as a preliminary step.
The procedure described in Durif et al. (201
7
) is used to compute
The procedure described in Durif et al. (201
8
) is used to compute
latent sparse components that are used in a regression model.
In addition, when a matrix \code{Xtest} is supplied, the procedure
predicts the response associated to these new values of the predictors.
...
...
@@ 183,10 +183,11 @@ points(1000:1000,1000:1000, type="l")
}
\references{
Durif G., Modolo L., Michaelsson J., Mold J. E., LambertLacroix S.,
Picard F. (2017). High Dimensional Classification with combined Adaptive
Sparse PLS and Logistic Regression, (in prep),
available on (\url{http://arxiv.org/abs/1502.05933}).
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
Chun, H., & Keles, S. (2010). Sparse partial least squares regression for
simultaneous dimension reduction and variable selection. Journal of the
...
...
pkg/man/spls.cv.Rd
View file @
26f1de07
...
...
@@ 83,14 +83,14 @@ error rate over the folds.
The
function
\
code
{
spls
.
cv
}
chooses
the
optimal
values
for
the
hyper

parameter
of
the
\
code
{
spls
}
procedure
,
by
minimizing
the
mean
squared
error
of
prediction
over
the
hyper

parameter
grid
,
using
Durif
et
al
.
(
201
7
)
adaptive
SPLS
algorithm
.
using
Durif
et
al
.
(
201
8
)
adaptive
SPLS
algorithm
.
}
\
details
{
The
columns
of
the
data
matrices
\
code
{
Xtrain
}
and
\
code
{
Xtest
}
may
not
be
standardized
,
since
standardizing
can
be
performed
by
the
function
\
code
{
spls
.
cv
}
as
a
preliminary
step
.
The
procedure
is
described
in
Durif
et
al
.
(
201
7
).
The
K

fold
The
procedure
is
described
in
Durif
et
al
.
(
201
8
).
The
K

fold
cross

validation
can
be
summarize
as
follow
:
the
train
set
is
partitioned
into
K
folds
,
for
each
value
of
hyper

parameters
the
model
is
fit
K
times
,
using
each
fold
to
compute
the
prediction
error
rate
,
and
fitting
the
...
...
@@ 136,10 +136,11 @@ cv1$ncomp.opt
}
\
references
{
Durif
G
.,
Modolo
L
.,
Michaelsson
J
.,
Mold
J
.
E
.,
Lambert

Lacroix
S
.,
Picard
F
.
(
2017
).
High
Dimensional
Classification
with
combined
Adaptive
Sparse
PLS
and
Logistic
Regression
,
(
in
prep
),
available
on
(\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}).
Durif
,
G
.,
Modolo
,
L
.,
Michaelsson
,
J
.,
Mold
,
J
.
E
.,
Lambert

Lacroix
,
S
.,
Picard
,
F
.,
2018.
High
dimensional
classification
with
combined
adaptive
sparse
PLS
and
logistic
regression
.
Bioinformatics
34
,
485

493.
\
url
{
https
://
doi
.
org
/
10.1093
/
bioinformatics
/
btx571
}.
Available
at
\
url
{
http
://
arxiv
.
org
/
abs
/
1502.05933
}.
}
\
seealso
{
\
code
{\
link
{
spls
}}
...
...
pkg/man/spls.stab.Rd
View file @
26f1de07
...
...
@@ 69,7 +69,7 @@ the seed for pseudorandom number generation is set accordingly.}
\value{
An object with the following attributes
\item{q.Lambda}{A table with values of q.Lambda (c.f. Durif
et al. (201
7
) for the notation), being the averaged number of covariates
et al. (201
8
) for the notation), being the averaged number of covariates
selected among the entire grid of hyperparameters candidates values,
for increasing size of hyperparameter grid.}
\item{probs.lambda}{A table with estimated probability of selection for each
...
...
@@ 83,7 +83,7 @@ candidate values \code{(ncomp, lambda.l1)} of hyperparameters
on multiple subsamplings in the data. The stability selection procedure
selects the covariates that are selected by most of the models among the
grid of hyperparameters, following the procedure described in
Durif et al. (201
7
). Candidates values for \code{ncomp} and \code{lambda.l1}
Durif et al. (201
8
). Candidates values for \code{ncomp} and \code{lambda.l1}
are respectively given by the input arguments \code{ncomp.range} and
\code{lambda.l1.range}.
}
...
...
@@ 92,7 +92,7 @@ The columns of the data matrices \code{X} may not be standardized,
since standardizing is performed by the function \code{spls.stab}
as a preliminary step.
The procedure is described in Durif et al. (201
7
). The stability selection
The procedure is described in Durif et al. (201
8
). The stability selection
procedure can be summarize as follow (c.f. Meinshausen and Buhlmann, 2010).
(i) For each candidate values \code{(ncomp, lambda.l1)} of
...
...
@@ 149,10 +149,11 @@ stability.selection(stab1, piThreshold=0.6, rhoError=10)
}
\references{
Durif G., Modolo L., Michaelsson J., Mold J. E., LambertLacroix S.,
Picard F. (2017). High Dimensional Classification with combined Adaptive
Sparse PLS and Logistic Regression, (in prep),
available on (\url{http://arxiv.org/abs/1502.05933}).
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
Meinshausen, N., Buhlmann P. (2010). Stability Selection. Journal of the
Royal Statistical Society: Series B (Statistical Methodology)
...
...
pkg/man/stability.selection.Rd
View file @
26f1de07
...
...
@@ 12,10 +12,10 @@ stability.selection(stab.out, piThreshold = 0.9, rhoError = 10)
\code{\link{logit.spls.stab}} or \code{\link{multinom.spls.stab}}.}
\item{piThreshold}{a value in (0,1], corresponding to the threshold
probability used to select covariate (c.f. Durif et al., 201
7
).}
probability used to select covariate (c.f. Durif et al., 201
8
).}
\item{rhoError}{a positive value used to restrict the grid of
hyperparameter candidate values (c.f. Durif et al., 201
7
).}
hyperparameter candidate values (c.f. Durif et al., 201
8
).}
}
\value{
An object with the following attributes:
...
...
@@ 28,12 +28,12 @@ hyperparameters.}
\description{
The function \code{stability.selection} returns the list of selected
covariates, when following the stability selection procedure described in
Durif et al. (201
7
). In particular, it selects covariates that are selected
Durif et al. (201
8
). In particular, it selects covariates that are selected
by most of the sparse PLS, the logitSPLS or the multinomialSPLS models
when exploring the grid of hyperparameter candidate values.
}
\details{
The procedure is described in Durif et al. (201
7
). The stability selection
The procedure is described in Durif et al. (201
8
). The stability selection
procedure can be summarize as follow (c.f. Meinshausen and Buhlmann, 2010).
(i) For each candidate values of hyperparameters, a model is trained
...
...
@@ 48,7 +48,7 @@ function \code{\link{stability.selection.heatmap}}.
set of covariates that were selected by most of the training among the
grid of hyperparameters candidate values, based on a threshold probability
\code{piThreshold} and a restriction of the grid of hyperparameters based
on \code{rhoError} (c.f. Durif et al., 201
7
, for details).
on \code{rhoError} (c.f. Durif et al., 201
8
, for details).
This function achieves the second step (ii) of the stability selection
procedure. The first step (i) is achieved by the functions
...
...
@@ 87,6 +87,17 @@ str(stab1)
stability.selection(stab1, piThreshold=0.6, rhoError=10)
}
}
\references{
Durif, G., Modolo, L., Michaelsson, J., Mold, J.E., LambertLacroix, S.,
Picard, F., 2018. High dimensional classification with combined
adaptive sparse PLS and logistic regression. Bioinformatics 34,
485493. \url{https://doi.org/10.1093/bioinformatics/btx571}.
Available at \url{http://arxiv.org/abs/1502.05933}.
Meinshausen, N., Buhlmann P. (2010). Stability Selection. Journal of the
Royal Statistical Society: Series B (Statistical Methodology)
72, no. 4, 417473.
}
\seealso{
\code{\link{spls.stab}}, \code{\link{logit.spls.stab}},
...
...
pkg/man/stability.selection.heatmap.Rd
View file @
26f1de07
...
...
@@ 22,10 +22,10 @@ The function \code{stability.selection.heatmap} allows to visualize
estimated probabilities to be selected for each covariate depending on the
value of hyperparameters in the spls, logitspls or multinomialspls models.
These estimated probabilities are used in the stability selection procedure
described in Durif et al. (201
7
).
described in Durif et al. (201
8
).
}
\details{
The procedure is described in Durif et al. (201
7
). The stability selection
The procedure is described in Durif et al. (201
8
). The stability selection
procedure can be summarize as follow (c.f. Meinshausen and Buhlmann, 2010).
(i) For each candidate values of hyperparameters, a model is trained
...
...
@@ 40,7 +40,7 @@ function \code{\link{stability.selection.heatmap}}.
set of covariates that were selected by most of the training among the
grid of hyperparameters candidate values, based on a threshold probability
\code{piThreshold} and a restriction of the grid of hyperparameters based
on \code{rhoError} (c.f. Durif et al., 201
7
, for details).
on \code{rhoError} (c.f. Durif et al., 201
8
, for details).
This function allows to visualize probabalities estimated at the first
step (i) of the stability selection by the functions \code{\link{spls.stab}},
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment