1,154 reputation
217
bio website
location
age
visits member for 1 years, 5 months
seen 13 hours ago

Aug
30
asked How to force an evaluation to use only one core?
Aug
30
comment How to construct this matrix fast?
Sorry for late reply. I became busy yesterday. you method works. But I don't understand the syntax mem : matC[n_Integer?Positive] := mem = . never seen it before
Aug
28
comment How to construct this matrix fast?
Actually, I am gonna NIntegrating one of the matrix element with variable kx under fixed n
Aug
28
revised Is the CompiledFunctionCall WVM opcode efficient?
added 22 characters in body
Aug
28
asked Is the CompiledFunctionCall WVM opcode efficient?
Aug
28
answered Compiling more functions that don't call MainEvaluate
Aug
28
comment Compiling more functions that don't call MainEvaluate
@Szabolcs Why you said "If f and g are compiled functions themselves, use the "InlineCompiledFunctions" -> True option as well."? I don't see this option is necessary. InlineExternalDefinitions is enough.
Aug
28
comment How to construct this matrix fast?
Hi, Mr.Wizard! just a remind, maybe you have got an idea
Aug
28
comment how to Compile user-defined function like this
Hi, Halirutan! Thank you for your reply!. I was reading mathematica doc today. And found that there exists simpler solution. Just set CompilationOptions -> {"ExpressionOptimization" -> True} will do all the work.
Aug
27
comment how to Compile user-defined function like this
@IstvánZachar Hello, thank you for your comment. Well, I read several other post, but still don't understand "mainEvaluate" well. CompilePrint will give T(C2)0 = MainEvaluate[ Hold[Inverse][ T(C2)14]] and there is a saying MainEvaluate tells us that something is going through the evaluator and wasn't compilable. So this means Inverse is not being compiled? What do you mean "invmat still uses Det which cannot be compiled"?? I see both Inverse and Det were in the second list in mathematica.stackexchange.com/questions/1096/…;. I really don't understand
Aug
27
accepted is there a method to send different cell evaluation to different core
Aug
27
comment is there a method to send different cell evaluation to different core
OK! I see. Thank you very much,Alexey Popkov! Forgive my ignorance about CellPrint. And I found if the function is user defined, I have to include all the related definition into the cell print. if the definition is long that will be a little inconvience. But anyway, it works in principle. You solved my problem.
Aug
27
asked Whether NIntegrate evaluation is multithreaded or not?
Aug
27
comment how to Compile user-defined function like this
@halirutan I have edited my question
Aug
27
revised how to Compile user-defined function like this
added 2967 characters in body
Aug
27
comment how to Compile user-defined function like this
finally someone replied. Thank you!. Well actually I don't expect to compile Inverse. What I want to compile is the matrix, you see there is a lot similar term in it. And I have test it, mat is compiled, you can see it if you copy the code and paste it into Mathematica and run. But mat2 didn't compile
Aug
27
comment is there a method to send different cell evaluation to different core
Are you sure? I tried CellPrint@ Cell[BoxData[MakeBoxes[Do[Print[n]; Pause[1], {n, 20}]]], "Input", Evaluator -> "My"], it don't print the number, it print the Do expression
Aug
27
comment is there a method to send different cell evaluation to different core
I tried. But add a local kernel with a different name, doesn't work at all. Use`ExpressionCell[expr, "Input", Evaluator -> "Local"]`, two cell with different local evaluator still run in sequence.
Aug
27
comment is there a method to send different cell evaluation to different core
What does several local kernel mean? I only have one local kernel
Aug
27
comment is there a method to send different cell evaluation to different core
Hi, thank you for your reply. But I'm afraid that Evaluator doesn't related to my problem. I know the Kernel Configuration. But this option configure remote kernels. What I am talking about is multicore in one kernel.