| Type: | Package |
| Title: | Space Filling Based Tools for Data Mining |
| Version: | 0.1.0 |
| Author: | Mohamed Laib and Mikhail Kanevski |
| Maintainer: | Mohamed Laib <laib.med@gmail.com> |
| Description: | Contains space filling based tools for machine learning and data mining. Some functions offer several computational techniques and deal with the out of memory for large big data by using the ff package. |
| Imports: | wordspace, doParallel, ff, parallel, stats |
| License: | GPL-3 |
| URL: | https://sites.google.com/site/mohamedlaibwebpage/ |
| BugReports: | https://github.com/mlaib/SFtools/issues |
| Encoding: | UTF-8 |
| LazyData: | true |
| Note: | The GPU computing version can be found soon, on following url https://sites.google.com/site/mohamedlaibwebpage/Software. |
| RoxygenNote: | 6.0.1 |
| NeedsCompilation: | no |
| Packaged: | 2017-06-28 15:23:59 UTC; mlaib |
| Repository: | CRAN |
| Date/Publication: | 2017-06-28 15:53:43 UTC |
SFtools: Space Filling Based Tools for Data Mining
Description
Contains space filling based tools for machine learning and data mining. Some functions offer several computational techniques and deal with the out of memory for large big data by using the ff package.
Author(s)
Mohamed Laib Mohamed.Laib@unil.ch and
Mikhail Kanevski Mikhail.Kanevski@unil.ch,
Maintainer: Mohamed Laib laib.med@gmail.com
References
M. Laib and M. Kanevski (2017). Unsupervised Feature Selection Based on Space Filling Concept, arXiv:1706.08894.
J. A. Royle, D. Nychka, An algorithm for the construction of spatial coverage designs with implementation in Splus, Computers and Geosciences 24 (1997) p. 479–488.
J. Franco, Planification d’expériences numériques en phase exploratoire pour la simulation des phénomènes complexes, Thesis (2008) 282.
D. Dupuy, C. Helbert, J. Franco (2015). DiceDesign and DiceEval: Two R Packages for Design and Analysis of Computer Experiments. Journal of Statistical Software, 65(11), 1-38. Jstatsoft.
See Also
Useful links:
Report bugs at https://github.com/mlaib/SFtools/issues
Simulated dataset
Description
Generates a simulated dataset (the Infinity dataset)
Usage
Infinity(n=1000)
Arguments
n |
Number of generated data points (by default: |
Value
A data.frame of simulated dataset, with 7 features
(4 of them are redundants)
Author(s)
Mohamed Laib Mohamed.Laib@unil.ch
Examples
infinity<-Infinity(n=1000)
plot(infinity$x1,infinity$x2)
## Not run:
#### Visualisation of the infinity dataset (3D) ####
require(rgl)
require(colorRamps)
c <- cut(infinity$z,breaks=100)
cols <- matlab.like(100)[as.numeric(c)]
plot3d(infinity$x1,infinity$x2,infinity$z,radius=0.01, col=cols,
type="s",xlab="x1",ylab="x2",zlab="z",box=F)
grid3d(c("x","y","z"),col="black",lwd=1)
## End(Not run)
UfsCov algorithm for unsupervised feature selection
Description
Applies the UfsCov algorithm based on the space filling concept, by using a sequatial forward search (SFS).
Usage
UfsCov(data)
Arguments
data |
Data of class: |
Details
Since the algorithm is based on pairwise distances, and
according to the computing power of your machine, large number of
data points can take much time and needs more memory.
See UfsCov_par for parellel computing, or
UfsCov_ff for memory efficient storage of large data
on disk and fast access (by using the ff and the ffbase packages).
Value
A list of two elements:
-
CovDa vector containing the coverage measure of each step of the SFS. -
IdRa vector containing the added variables during the selection procedure.
Note
The algorithm does not deal with missing values and constant features. Please make sure to remove them.
Author(s)
Mohamed Laib Mohamed.Laib@unil.ch
References
M. Laib and M. Kanevski (2017). Unsupervised Feature Selection Based on Space Filling Concept, arXiv:1706.08894.
Examples
infinity<-Infinity(n=800)
Results<- UfsCov(infinity)
cou<-colnames(infinity)
nom<-cou[Results[[2]]]
par(mfrow=c(1,1), mar=c(5,5,2,2))
names(Results[[1]])<-cou[Results[[2]]]
plot(Results[[1]] ,pch=16,cex=1,col="blue", axes = FALSE,
xlab = "Added Features", ylab = "Coverage measure")
lines(Results[[1]] ,cex=2,col="blue")
grid(lwd=1.5,col="gray" )
box()
axis(2)
axis(1,1:length(nom),nom)
which.min(Results[[1]])
## Not run:
#### UfsCov on the Butterfly dataset ####
require(IDmining)
N <- 1000
raw_dat <- Butterfly(N)
dat<-raw_dat[,-9]
Results<- UfsCov(dat)
cou<-colnames(dat)
nom<-cou[Results[[2]]]
par(mfrow=c(1,1), mar=c(5,5,2,2))
names(Results[[1]])<-cou[Results[[2]]]
plot(Results[[1]] ,pch=16,cex=1,col="blue", axes = FALSE,
xlab = "Added Features", ylab = "Coverage measure")
lines(Results[[1]] ,cex=2,col="blue")
grid(lwd=1.5,col="gray" )
box()
axis(2)
axis(1,1:length(nom),nom)
which.min(Results[[1]])
## End(Not run)
UfsCov for unsupervised features selection
Description
Applies the UfsCov algorithm based on the space filling concept, by using a sequatial forward search (for memory efficient storage of large data on disk and fast access).
Usage
UfsCov_ff(data, blocks=2)
Arguments
data |
Data of class: |
blocks |
Number of splits to facilitate the computation of the distance matrix (by default: blocks=2). |
Value
A list of two elements:
-
CovDa vector containing the coverage measure of each step of the SFS. -
IdRa vector containing the added variables during the selection procedure.
Note
This function is still under developement.
Author(s)
Mohamed Laib Mohamed.Laib@unil.ch
References
M. Laib and M. Kanevski (2017). Unsupervised Feature Selection Based on Space Filling Concept, arXiv:1706.08894.
Examples
## Not run:
#### Infinity dataset ####
N <- 1000
dat<-Infinity(N)
Results<- UfsCov_ff(dat)
cou<-colnames(dat)
nom<-cou[Results[[2]]]
par(mfrow=c(1,1), mar=c(5,5,2,2))
names(Results[[1]])<-cou[Results[[2]]]
plot(Results[[1]] ,pch=16,cex=1,col="blue", axes = FALSE,
xlab = "Added Features", ylab = "Coverage measure")
lines(Results[[1]] ,cex=2,col="blue")
grid(lwd=1.5,col="gray" )
box()
axis(2)
axis(1,1:length(nom),nom)
which.min(Results[[1]])
#### Butterfly dataset ####
require(IDmining)
N <- 1000
raw_dat <- Butterfly(N)
dat<-raw_dat[,-9]
Results<- UfsCov_ff(dat)
## End(Not run)
UfsCov algorithm for unsupervised feature selection
Description
Applies the UfsCov algorithm based on the space filling concept, by using a sequatial forward search (SFS).This function offers a parellel computing.
Usage
UfsCov_par(data, ncores=2)
Arguments
data |
Data of class: |
ncores |
Number of cores to use (by default: |
Details
Since the algorithm is based on pairwise distances, and
according to the computing power of your machine, large number of
data points needs more memory. See UfsCov_ff for memory
efficient storage of large data on disk and fast access (by using the
ff and the ffbase packages).
Value
A list of two elements:
-
CovDa vector containing the coverage measure of each step of the SFS. -
IdRa vector containing the added variables during the selection procedure.
Note
The algorithm does not deal with missing values and constant
features. Please make sure to remove them. Note that it is not recommanded to
use this function with small data, it takes more time than using the
standard UfsCov function.
Author(s)
Mohamed Laib Mohamed.Laib@unil.ch
References
M. Laib and M. Kanevski (2017). Unsupervised Feature Selection Based on Space Filling Concept, arXiv:1706.08894.
See Also
Examples
N <- 800
dat<-Infinity(N)
Results<- UfsCov_par(dat,ncores=2)
cou<-colnames(dat)
nom<-cou[Results[[2]]]
par(mfrow=c(1,1), mar=c(5,5,2,2))
names(Results[[1]])<-cou[Results[[2]]]
plot(Results[[1]] ,pch=16,cex=1,col="blue", axes = FALSE,
xlab = "Added Features", ylab = "Coverage measure")
lines(Results[[1]] ,cex=2,col="blue")
grid(lwd=1.5,col="gray" )
box()
axis(2)
axis(1,1:length(nom),nom)
which.min(Results[[1]])
## Not run:
N<-5000
dat<-Infinity(N)
## Little comparison:
system.time(Uf<-UfsCov(dat))
system.time(Uf.p<-UfsCov_par(dat, ncores = 4))
## End(Not run)