hi, everyone: I have a problem about the paralle computing when using parallel computing box. I want to combine the matlab with the HFSS to solve a problem, but I find it can't work when i call the HFSS in one cluster. The program belows: >jm=findResource('scheduler','type','jobmanager','name','isaac','LookupURL','192.168.58.109'); >job=createJob(jm); >set(job,'PathDependencies',{'Y:\'}); >createTask(job,@vbs,1,{2}); >submit(job); >waitForState(job); >results=getAllOutputArguments(job) >destroy(job); >function a=vbs(i) >hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','Y:\pipi.vbs',false,true); >i=i-1; >a=i; When i run this program, the HFSS would open, but never open the pipi.vbs ,why ? where is the problem??? just help me ,thank you !!

0 |

3/26/2010 1:03:02 PM

"tan " <tgn0825@hotmail.com> writes: > I have a problem about the paralle computing when using parallel > computing box. I want to combine the matlab with the HFSS to solve a > problem, but I find it can't work when i call the HFSS in one > cluster. The program belows: >>jm=findResource('scheduler','type','jobmanager','name','isaac','LookupURL','192.168.58.109'); > [...] >>function a=vbs(i) >>hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','Y:\pipi.vbs',false,true); >>i=i-1; >>a=i; > > When i run this program, the HFSS would open, but never open the > pipi.vbs ,why ? where is the problem??? just help me ,thank you !! I think the problem is that you're using the jobmanager and trying to access "pipi.vbs" using a mapped drive-letter. MDCS workers can't see your drive letter mappings. It should work if you provide a UNC path such as \\server-name\path\to\pipi.vbs (We have an undocumented function that can do this mapping for you, dctReplaceDriveWithUNCPath('y:\pipi.vbs') that you must run on your MATLAB client (not on the cluster) - but please be aware of the usual caveats about undocumented stuff - we might choose to remove/change/rename it without notice) Cheers, Edric.

0 |

3/26/2010 1:16:00 PM

Edric M Ellis <eellis@mathworks.com> wrote in message <ytw634j5ilr.fsf@uk-eellis-deb5-64.mathworks.co.uk>... > "tan " <tgn0825@hotmail.com> writes: > > > I have a problem about the paralle computing when using parallel > > computing box. I want to combine the matlab with the HFSS to solve a > > problem, but I find it can't work when i call the HFSS in one > > cluster. The program belows: > > >>jm=findResource('scheduler','type','jobmanager','name','isaac','LookupURL','192.168.58.109'); > > > [...] > > >>function a=vbs(i) > >>hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','Y:\pipi.vbs',false,true); > >>i=i-1; > >>a=i; > > > > When i run this program, the HFSS would open, but never open the > > pipi.vbs ,why ? where is the problem??? just help me ,thank you !! > > I think the problem is that you're using the jobmanager and trying to > access "pipi.vbs" using a mapped drive-letter. MDCS workers can't see > your drive letter mappings. It should work if you provide a UNC path > such as > > \\server-name\path\to\pipi.vbs > > (We have an undocumented function that can do this mapping for you, > dctReplaceDriveWithUNCPath('y:\pipi.vbs') that you must run on your > MATLAB client (not on the cluster) - but please be aware of the usual > caveats about undocumented stuff - we might choose to > remove/change/rename it without notice) > > > Cheers, > > Edric. thank you, Edric. but i have set the path of mapped drive before i run the program. and i am sure that the path of "pipi.vbs" have been set and mapped. when i don't use the mdce, and input the following in matlab command window: >hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','Y:\pipi.vbs',false,true); it can run pipi.vbs,and that indicats there is no problem about the path of pipi.vbs. I hope you can help me, i am tired about this. thank you again.

0 |

3/30/2010 10:47:02 AM

"tan " <tgn0825@hotmail.com> wrote in message <hoskr6$lk$1@fred.mathworks.com>... > Edric M Ellis <eellis@mathworks.com> wrote in message <ytw634j5ilr.fsf@uk-eellis-deb5-64.mathworks.co.uk>... > > "tan " <tgn0825@hotmail.com> writes: > > > > > I have a problem about the paralle computing when using parallel > > > computing box. I want to combine the matlab with the HFSS to solve a > > > problem, but I find it can't work when i call the HFSS in one > > > cluster. The program belows: > > > > >>jm=findResource('scheduler','type','jobmanager','name','isaac','LookupURL','192.168.58.109'); > > > > > [...] > > > > >>function a=vbs(i) > > >>hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','Y:\pipi.vbs',false,true); > > >>i=i-1; > > >>a=i; > > > > > > When i run this program, the HFSS would open, but never open the > > > pipi.vbs ,why ? where is the problem??? just help me ,thank you !! > > > > I think the problem is that you're using the jobmanager and trying to > > access "pipi.vbs" using a mapped drive-letter. MDCS workers can't see > > your drive letter mappings. It should work if you provide a UNC path > > such as > > > > \\server-name\path\to\pipi.vbs > > > > (We have an undocumented function that can do this mapping for you, > > dctReplaceDriveWithUNCPath('y:\pipi.vbs') that you must run on your > > MATLAB client (not on the cluster) - but please be aware of the usual > > caveats about undocumented stuff - we might choose to > > remove/change/rename it without notice) > > > > > > Cheers, > > > > Edric. > > thank you, Edric. but i have set the path of mapped drive before i run the program. and i am sure that the path of "pipi.vbs" have been set and mapped. when i don't use the mdce, and input the following in matlab command window: > >hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','Y:\pipi.vbs',false,true); > it can run pipi.vbs,and that indicats there is no problem about the path of pipi.vbs. > I hope you can help me, i am tired about this. thank you again. I am sorry about the reply to your answer. you are right:"MDCS workers can't see your drive letter mappings. It should work if you provide a UNC path". and now i have solve this problem ,thank you Edric,my friend.

0 |

3/30/2010 11:34:03 AM

first,thank you for your answer. Now I want to start a worker on a remote claster,and on that claster i want to start the hfss, how to do that ?? I use the UNC path,but how to define the path of the hfss.exe?? >hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','\\Isaac\ceshi\pipi.vbs',false,true);

0 |

4/3/2010 12:58:04 AM

"tan " <tgn0825@hotmail.com> writes: > first,thank you for your answer. Now I want to start a worker on a remote > claster,and on that claster i want to start the hfss, how to do that ?? I use > the UNC path,but how to define the path of the hfss.exe?? >>hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','\\Isaac\ceshi\pipi.vbs',false,true); You either need to install hfss.exe on a shared path and refer to it by the UNC path, or else you need to install it locally on the cluster node. Cheers, Edric.

0 |

4/6/2010 8:22:35 AM

Edric M Ellis <eellis@mathworks.com> wrote in message <ytw1vetf0s4.fsf@uk-eellis-deb5-64.mathworks.co.uk>... > "tan " <tgn0825@hotmail.com> writes: > > > first,thank you for your answer. Now I want to start a worker on a remote > > claster,and on that claster i want to start the hfss, how to do that ?? I use > > the UNC path,but how to define the path of the hfss.exe?? > >>hfssExecuteScript('C:\"Program Files"\Ansoft\HFSS10\hfss.exe','\\Isaac\ceshi\pipi.vbs',false,true); > > You either need to install hfss.exe on a shared path and refer to it by > the UNC path, or else you need to install it locally on the cluster > node. > > > Cheers, > > Edric. thank you very much, the problem has been solved!

0 |

4/6/2010 12:46:04 PM

in pmode, I creaed a segment and output is: lab1: 1 2 3 lab2: 11 12 13 lab3: 21 22 23 lab4: 31 32 33 When I am using " whole = codistributed(segment, codistributor()) " The system told " ??? Undefined function or variable 'codistributor'. " What's the problem.... flycod <jby1985@gmail.com> writes: > in pmode, I creaed a segment and output is: > lab1: 1 2 3 > lab2: 11 12 13 > lab3: 21 22 23 > lab4: 31 32 33 > > When I am using " whole = codistributed(segment, codistributor()) " > > The system told " ??? Undef...

Let U[0; 2theta] be a uniform random variable from the interval [0; 2theta] and let A Exp(1) be exponentially distributed with mean 1. Assume and A independent. Compute the mean mX(t) =E[X(t)] and autocorrelation RX(s; t) = E[X(s)X(t)] of the phase-shifted sinusoid.X(t) = A* cos(t +theta ): State also if X(t) is Wide Sense Stationary (WSS). plot 10 realisations of X(t) plotR(s-t,0)as a function of s-t "pramod kumar" <pramod.kilu@gmail.com> wrote in message <jpapso$44b$1@newscl01ah.mathworks.com>... > Let U[0; 2theta] be a uniform random variable ...

Hello, in the last month I encountered big problems to read mail from yahoo.com.au server. There is no chance to connect to e-mail /www based server. Having logged to www.yahoo.com.au mail window doesn't open waiting for an image from au.adserver.yahoo.com server is that server dead, making reading mail completely impossible ? Ping says, that server is unreachable. What can I do to make things better as in previous months ? Ple ...

Hi all, I need your help. I have a Matlab program, which uses the Matlab symbolic toolbox and does exclusively symbolic computations. It calls Maple's library internally. My program is currently non-parallel. I need to run it on a super computer. The super computer does not have Matlab and it is Linux based and supports MPI. I didn't figure out how to convert non-parallel program into parallel exectuables on Linux with MPI. Since I failed to get a nice solution of the above problem, now I am willing to make a comprise, I am going to make it coarse-grained distrubuted computing: (1...

Hi I like to use parallel threads, to use the multiple processor cores. (I have a quadcore system and frankly when using matlab despite all mathworks claims I NEVER EVER spotted it using more than 1 core (i.e. 25% CPU usage!) - yes I enabled multithreading support in the preferences). Anyway. So I want to take things in my own hands. So I started with and example of an "embarresingly parallel task: filtering 12 images. The images are in the variable imgs{...} tic I={}; h = fspecial('gaussian',30,30); for jj= 1:size(imgs,2) I{jj}=imfilter(imgs{jj},h); end toc -> this takes...

The “divide and average” method, an old-time method for approximating the square root of any positive number "a" can be formulated as: x= (x+a/x)/2 (a) Write a well-structured pseudocode to implement this algorithm. Use proper indentation so that the structure is clear. (b) Develop, debug, and document a Matlab function to implement this equation. Its basically a problem that by using this formula or function it would give me the square root of a number inserted. I see you cut the (!) down from 22 to 1. That is a good start. Now, where are you stuck? _U...

for i=1:size(APLocation,1) point=APLocation(i,:); parfor j=1:length (RXpoint) rssi(i,j) = LOSS(point,RXpoint(:,j)'); %%rssi having the signal strength from all AP's end end when i'm running the following code, i get the same value for each element of rssi (serially i get different results) where is my problem? Creating some fake data and and implementing a simple LOSS function, I do not see a different between the serial and parallel execution. Does LOSS use any persistent or global variables? function [ rssi rssi2 ] = foo APLoc...

Any one know how to work this problem? In this exercise you are to compute a power of the diagonalizable matrix A. Let A = pascal(5). Use the eigenvalue decomposition of A to calculate the ninth power of A. You cannot apply the power operator ^ to the matrix A. "Lynn Koch" <ankrunner@aol.com> wrote in message news:jfca37$r46$1@newscl01ah.mathworks.com... > Any one know how to work this problem? > In this exercise you are to compute a power of the diagonalizable matrix > A. Let A = pascal(5). Use the eigenvalue decomposition of A to calculate > the ninth powe...

I get the following error when running my code over distributed computing server; however, I don't get the error when running it with matlabpool set to local: ??? Error using ==> parallel_function>make_general_channel/channel_general at 843 Undefined function or method 'findCentroidsIndep' for input arguments of type 'struct'. Error in ==> parallel_function>distributed_execution at 752 [tags, out] = P.getCompleteIntervals(chunkSize); Error in ==> parallel_function at 564 R = distributed_execution(... Error in ==> processVideoDistributed at 85 n...

Hello friends, I want to know that ,is it possible to do parallel computing with MATLAB on a single computer.I want that different modules communicate and interlinked with each other so that called by other functions also. Hi, have a look at the distributed computing toolbox: http://www.mathworks.com/products/distribtb/ Although mainly intended to seperate the workload to different machines you may also run it on one machine... Titus www.mathworks.com\products\dis "shivi chaturvedi" <shivi79@rediffmail.com> schrieb im Newsbeitrag news:ef3d35a.-1@webcrossing...

hi to all, i have a problem running this code on Matlab 6.5, it running fine on matlab Version 7.1.0.183 (R14) Service Pack 3 (but i have not this version) Matlab 6.5 give me a compilation error: ?? Error: File: C:\Documents and Settings\Roby\Documenti\Tesi\portoni_sd_sa\portoni_sd_sa\portoni_2d.m Line: 23 Column: 19 "identifier" expected, "(" found. row 23 col 19 start from @(points) ... (at line begining with error_function = ) Can u help me to resolve this error? can u give me sugestion how can write this line i'm a beginner and i can't understand why t...

Is any body doing parallel computing with matlab here? ...

Hello, I would like to know if it is possible to run multiple m-files parallelly in MATLAB. For example, suppose I need to run 'n' script files all independent of each other. I needed it for a robotics application. Is it possible to assign preferences to the processes(like in real time processes) so that a process of lesser preference is put on the hold if a more preferred task needs to be run. Please give details, as I am not very used to MATLAB. Thanks in advance. Regards... "ajay " <ajaypai88@gmail.com> writes: > I would like to know if it is possible to run mu...

The program works well on a local jobmanager. When doing it using a jobmanager on a remote cluster computer, the error looks like: Error using ==>parallel_function>make_general_channel/channel_general at 843 Undefined function handle. Error in ==> parallel_function>distributed_execution at 752 [tags, out] = P.getCompleteIntervals(chunkSize); Error in ==> parallel_function at 564 R = distributed_execution(... Could anybody tell me what's wrong? Thanks. YAN <lingyan.sheng@gmail.com> writes: > The program works well on a local jobmanager. When doing it using a ...