Re: [R] Matrix package,solve() errors and crashes Please help
Hello Martin,
Thank you very much for the reply. Thanks for solving the problem earlier
I was not aware of it.I did not receive any message and that is what made me
to post it again. Thank you once again.
I will let the person know about the solve(a,Matrix(b))
I still have this doubt.I still get the error message
*Error in solve(a, Matrix(b)) :
cs_lu(A) failed: near-singular A (or out of memory)*
**
Am I doing any mistake.
I made changes to your program and tried still I got the above error. So, I
planned to stick to the program that you sent and still I get the same
error.
simplify()ing graph.. [Ok]
d - get.adjacency(., sparse) .. [Ok]
d - as(d, symmetricMatrix) .. [Ok]
M - (id - exponent * d) .. [Ok]
b - degree(graph,.) .. [Ok]
solve(M, b).. Error in solve(a, Matrix(b)) :
cs_lu(A) failed: near-singular A (or out of memory)
Is there anything I am missing from your program or something I should do
before running your program
Please help
Thanks in Advance
Nathan
On Sat, May 16, 2009 at 7:31 AM, Martin Maechler maech...@stat.math.ethz.ch
wrote:
SS == Surendar Swaminathan surendar.swaminat...@gmail.com
on Fri, 15 May 2009 15:55:23 -0700 writes:
Hello All,
SS Please help me with this problem.I have been having this problem for
over a
SS month now and I could not find any information.I later realised that
error
SS is with MATRIX package.
[...]
SS This is my graph object.
SS The file 'Bonacich Power.RData' (1.5 MB) is available for download
at
SS
SS
http://dropbox.unl.edu/uploads/20090522/7a0d0313f21fd6a8/Bonacich%20Power.RData
[...]
and you have asked almost the idetnical question 4 weeks ago,
and got a first answer, but then not a *real* answer.
I'm sorry for that, since I had started answering you,
and even more as I *did* solve the problem back then.. (April 22).
more below ...
SS computer Configuration
SS *WIndows XP service Pack 3 .0 GB RAM*
SS I am using SPARSE matrix to solve the problem
SS This is the code I use to obtain bonpower using Sparse Matrix
SS alternatively the code is in the following website
SS http://igraph.wikidot.com/r-recipes#toc6
I'm really appaled that R-recipes on such a website give
such an unprofessional advice
{the following is so grabled because of your (implicit)
insistence on using HTML-ified e-mail .. }
SS *bonpow.sparse - function(graph, nodes=V(graph), loops=FALSE,
SS exponent=1, rescale=TRUE, tol=1e-07) {*
SS * ## remove loops if requested
SS *
SS * ## sparse adjacency matrix
SS d - get.adjacency(graph, sparse=TRUE)*
SS * ## sparse identity matrix
SS id - spMatrix(vcount(graph), vcount(graph),
SS i=1:vcount(graph), j=1:vcount(graph),
SS x=rep(1, vcount(graph)))
SS id - as(id, dgCMatrix)*
SS * ## solve it
SS ev - solve(id - exponent * d, tol=tol) %*% degree(graph,
mode=all)*
SS * if (rescale) {
SS ev - ev/sum(ev)
SS } else {
SS ev - ev * sqrt(vcount(graph)/sum((ev)^2))
SS }*
SS * ev[as.numeric(nodes) + 1]
SS }*
SS **
SS I realised that the error is in Matrix Package
SS *Error: cannot allocate vector of size 3.3 Gb
SS In addition: Warning messages:
SS 1: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)
SS 2: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)
SS 3: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)
SS 4: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)*
SS **
SS sessionInfo()
SS R version 2.9.0 (2009-04-17)
[ . ] {irrelevant here}
SS Please help Matrix Experts
Well, I've marked the one line above
ev - solve(id - exponent * d, tol=tol) %*% degree(graph, mode=all)*
begin abusive language; excuse in advance ..
which really contains stupid advice.
Yes, this is strongly put, but I think people who tell you to
solve
A x = b (A [n x n] matrix)
in R by x - solve(A) %*% b
should not give advice on programming at all.
end abusive language
One of the proper advices would use
x - solve(A, b)
and indeed that's even more crucial in the case where A is
sparse matrix: solve(A) is never sparse for sparse A (*),
but solve(A, b) can happen with a fast (and memory-efficient)
algorithm, and that's what happens also when you use the Matrix package.
I append a version of your function that is also slightly
improved in other places.
(*) and that's why your memory blows up when A is large and sparse !
SS Thanks in advance
SS Nathan
SS [[alternative HTML version deleted]]
(still It would *really* help if you used a more sensible e-mail
configuration }
Re: [R] Matrix package,solve() errors and crashes Please help
SS == Surendar Swaminathan surendar.swaminat...@gmail.com
on Fri, 15 May 2009 15:55:23 -0700 writes:
Hello All,
SS Please help me with this problem.I have been having this problem for
over a
SS month now and I could not find any information.I later realised that
error
SS is with MATRIX package.
[...]
SS This is my graph object.
SS The file 'Bonacich Power.RData' (1.5 MB) is available for download at
SS
SS
http://dropbox.unl.edu/uploads/20090522/7a0d0313f21fd6a8/Bonacich%20Power.RData
[...]
and you have asked almost the idetnical question 4 weeks ago,
and got a first answer, but then not a *real* answer.
I'm sorry for that, since I had started answering you,
and even more as I *did* solve the problem back then.. (April 22).
more below ...
SS computer Configuration
SS *WIndows XP service Pack 3 .0 GB RAM*
SS I am using SPARSE matrix to solve the problem
SS This is the code I use to obtain bonpower using Sparse Matrix
SS alternatively the code is in the following website
SS http://igraph.wikidot.com/r-recipes#toc6
I'm really appaled that R-recipes on such a website give
such an unprofessional advice
{the following is so grabled because of your (implicit)
insistence on using HTML-ified e-mail .. }
SS *bonpow.sparse - function(graph, nodes=V(graph), loops=FALSE,
SS exponent=1, rescale=TRUE, tol=1e-07) {*
SS * ## remove loops if requested
SS *
SS * ## sparse adjacency matrix
SS d - get.adjacency(graph, sparse=TRUE)*
SS * ## sparse identity matrix
SS id - spMatrix(vcount(graph), vcount(graph),
SS i=1:vcount(graph), j=1:vcount(graph),
SS x=rep(1, vcount(graph)))
SS id - as(id, dgCMatrix)*
SS * ## solve it
SS ev - solve(id - exponent * d, tol=tol) %*% degree(graph, mode=all)*
SS * if (rescale) {
SS ev - ev/sum(ev)
SS } else {
SS ev - ev * sqrt(vcount(graph)/sum((ev)^2))
SS }*
SS * ev[as.numeric(nodes) + 1]
SS }*
SS **
SS I realised that the error is in Matrix Package
SS *Error: cannot allocate vector of size 3.3 Gb
SS In addition: Warning messages:
SS 1: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)
SS 2: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)
SS 3: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)
SS 4: In solve(id - exponent * d, tol = tol) :
SS Reached total allocation of 1535Mb: see help(memory.size)*
SS **
SS sessionInfo()
SS R version 2.9.0 (2009-04-17)
[ . ] {irrelevant here}
SS Please help Matrix Experts
Well, I've marked the one line above
ev - solve(id - exponent * d, tol=tol) %*% degree(graph, mode=all)*
begin abusive language; excuse in advance ..
which really contains stupid advice.
Yes, this is strongly put, but I think people who tell you to
solve
A x = b (A [n x n] matrix)
in R by x - solve(A) %*% b
should not give advice on programming at all.
end abusive language
One of the proper advices would use
x - solve(A, b)
and indeed that's even more crucial in the case where A is
sparse matrix: solve(A) is never sparse for sparse A (*),
but solve(A, b) can happen with a fast (and memory-efficient)
algorithm, and that's what happens also when you use the Matrix package.
I append a version of your function that is also slightly
improved in other places.
(*) and that's why your memory blows up when A is large and sparse !
SS Thanks in advance
SS Nathan
SS [[alternative HTML version deleted]]
(still It would *really* help if you used a more sensible e-mail
configuration }
Regards,
Martin Maechler
---
Here's the corrected bonpow.sparse() function , actually with
extra code for diagnostic output, etc.
You should probably simplify it again :
## I got help from IGRAPH community to use sparse Matrix
##
## http://igraph.wikidot.com/r-recipes#toc6
##
### Enhanced by Martin Maechler:
bonpow.sparse - function(graph, nodes = V(graph), loops = FALSE,
exponent = 1, adj.type = both,
trace = TRUE,
rescale=FALSE, tol=1e-07)
{
stopifnot(require(igraph),
require(Matrix))
if(trace) {
c.width - 30
C1 - function(s) cat(sprintf(%-*s .. , c.width, s))
C2 - function() cat([Ok]\n)
} else { C1 - C2 - function(...) {} }
## remove loops if requested
if (!loops) {
C1(simplify()ing graph)
graph - simplify(graph, remove.multiple=FALSE, remove.loops=TRUE)
C2()
}
## sparse adjacency matrix
C1(d - get.adjacency(., sparse))
d - get.adjacency(graph, type = adj.type, sparse=TRUE); C2()
if(trace = 2)
cat(class(d): , class(d),\n)
Re: [R] Matrix package,solve() errors and crashes Please help
Hello All,
Please help me with this problem.I have been having this problem for over a
month now and I could not find any information.I later realised that error
is with MATRIX package.
I am working on graph object using IGRAPH version 0.5.2-2 package wanted
to do Bonacich Power.
What I have is undirected GRAPH.
This is my graph object.
The file 'Bonacich Power.RData' (1.5 MB) is available for download at
http://dropbox.unl.edu/uploads/20090522/7a0d0313f21fd6a8/Bonacich%20Power.RData
Graph size
*Vertices: 20924
Edges: 146938
Directed: FALSE
No graph attributes.
Vertex attributes: name.
No edge attributes.
*
computer Configuration
*WIndows XP service Pack 3 .0 GB RAM*
I am using SPARSE matrix to solve the problem
This is the code I use to obtain bonpower using Sparse Matrix
alternatively the code is in the following website
http://igraph.wikidot.com/r-recipes#toc6
*bonpow.sparse - function(graph, nodes=V(graph), loops=FALSE,
exponent=1, rescale=TRUE, tol=1e-07) {*
* ## remove loops if requested
*
* ## sparse adjacency matrix
d - get.adjacency(graph, sparse=TRUE)*
* ## sparse identity matrix
id - spMatrix(vcount(graph), vcount(graph),
i=1:vcount(graph), j=1:vcount(graph),
x=rep(1, vcount(graph)))
id - as(id, dgCMatrix)*
* ## solve it
ev - solve(id - exponent * d, tol=tol) %*% degree(graph, mode=all)*
* if (rescale) {
ev - ev/sum(ev)
} else {
ev - ev * sqrt(vcount(graph)/sum((ev)^2))
}*
* ev[as.numeric(nodes) + 1]
}*
**
I realised that the error is in Matrix Package
*Error: cannot allocate vector of size 3.3 Gb
In addition: Warning messages:
1: In solve(id - exponent * d, tol = tol) :
Reached total allocation of 1535Mb: see help(memory.size)
2: In solve(id - exponent * d, tol = tol) :
Reached total allocation of 1535Mb: see help(memory.size)
3: In solve(id - exponent * d, tol = tol) :
Reached total allocation of 1535Mb: see help(memory.size)
4: In solve(id - exponent * d, tol = tol) :
Reached total allocation of 1535Mb: see help(memory.size)*
**
sessionInfo()
R version 2.9.0 (2009-04-17)
i386-pc-mingw32
locale:
LC_COLLATE=English_United States.1252;LC_CTYPE=English_United
States.1252;LC_MONETARY=English_United
States.1252;LC_NUMERIC=C;LC_TIME=English_United States.1252
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] Matrix_0.999375-24 lattice_0.17-22igraph_0.5.2-2
loaded via a namespace (and not attached):
[1] grid_2.9.0
Please help Matrix Experts
Thanks in advance
Nathan
[[alternative HTML version deleted]]
__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
