1,360 reputation
522
bio website
location
age
visits member for 1 year, 11 months
seen Oct 21 at 0:44

Dec
10
asked efficient way to give nearest neighbour and next nearest neighbour of every point in a point set
Dec
8
revised How to constrain memory usage on HPC?
added 1635 characters in body
Dec
8
comment How to constrain memory usage on HPC?
@OleksandrR. Though the pmem option should suffice for ordinary usage, I must say to constrain memory inside mathematica is much more adaptive. First, there are some small clusters which are not equipped with a job management system. Second, the way to use the HPC is variable. Very often we can use mathemaica interactively on HPC by submitting an interactive PBS job, and if the pmem is set at the beginning, then all the calculations doing in this interactive job is restricted per process, while we should have been able to use more memory if we didn't set the pmem when submit the job.
Dec
8
comment How to constrain memory usage on HPC?
@OleksandrR. Thank you so much! You're right.Generally, an HPC has job management system. For example, the PBSpro provide pmem option to constrain memory per process. I actually forgot this before, thank you for reminding me this.
Dec
7
revised How to parallel the computation tasks according to machines instead of kernels
added 97 characters in body
Dec
7
answered How to parallel the computation tasks according to machines instead of kernels
Dec
7
awarded  Nice Question
Dec
7
revised How to constrain memory usage on HPC?
added 57 characters in body
Dec
7
comment How to constrain memory usage on HPC?
@halirutan No error message. It simply seems that the watchdog didn't launch, and free physical memory is not constrained, even not on HPC.
Dec
7
asked How to constrain memory usage on HPC?
Dec
3
comment How to solve this FindFit::nrlnum:
Oh,I understand why. I made a silly mistake, what a shame! I give Mathematica an improper fitting function, I should have add a constant term. Thank you Nasser!
Dec
3
comment How to solve this FindFit::nrlnum:
Interesting, your method works in MM9 while not work in MM8. And I really don't understand that the data points are so trivial, why I have to add so many options to make the fit work?
Dec
3
asked How to solve this FindFit::nrlnum:
Dec
1
comment How to parallel the computation tasks according to machines instead of kernels
@ssch Thank you ssch! You enlightened me, though your comment is not 100% right. Try ParallelTable[Eigenvalues[Randomreal[{1,2},{6000,6000}],{1}], this won't automatic multi-thread. But I figured out just now that ParallelTable[SetSystemOptions["MKLThreads" -> 8];Eigenvalues[Randomreal[{1,2},{6000,6000}],{1}]will use 8 cores. So problem solved. When we want to use multi-thread feature, we have to "parallel" SetSystemOptions["MKLThreads" -> 8] to each kernels.
Dec
1
awarded  Promoter
Nov
29
comment How to parallel the computation tasks according to machines instead of kernels
@tchronis well, the most time consuming part is eigenvalues computation, it can't be compiled. Eigenvalues is already a built-in numeric function, it should not be much slower than C or fortran
Nov
29
revised How to parallel the computation tasks according to machines instead of kernels
added 56 characters in body
Nov
29
asked How to parallel the computation tasks according to machines instead of kernels
Nov
25
accepted How to Clear variables represented as a list of strings?
Nov
25
comment How to Clear variables represented as a list of strings?
@YvesKlett I don't know how to write a proper title. Could you help to edit my title? Thank you