Ali Hassani

45 Reputation

4 Badges

8 years, 244 days

MaplePrimes Activity


These are replies submitted by Ali Hassani

@acer Thank you very much

Best wishes

@acer Thanks very much for explaining in detail. 

@tomleslie I always have a problem extracting the coefficients of "diff" functions in the expression. I saw you used "op~" to extract the coefficients in chebyshev(exp(x),x). 
Is it possible to apply a simillar way to take the coeffiecients of the expression including "diff"? For example, let me assume the following expresstion:
eq:=2*ln(x)*diff(y(x),x,x)+2*x*diff(y(x),x)+sin(x)*x^2*y(x);
Now, the coefficeints of diff(y(x),x,x), diff(y(x),x), and y(x) are 2*ln(x), 2*x, and sin(x)*x^2, respectively. 
How can I extract the above coefficients by coding in such a way you used for chebyshev(exp(x),x) or any other way you know?
Thanks

@acer Thanks very much for consuming your time.
Would you please more tell about "specfunc" and its applications?
Best wishes

@tomleslie 

Thanks very much for presenting a simple way to solve my problem. I wish you more explain what "tuning" is and give me an example about.
Best wishes

@acer 

Thanks very much for your attention to my problem
The mentioned procedures, especially the last one, are excellent. I think that the last procedure could be employed by Maple as a built-in procedure to overcome the drawback of the construction of sub- and superscribed variables in 1-D math.

 

@mmcdara 
Thanks veru much

@dharr 
Thanks very much.
Can you define a procedure or module to combine the construction of the sub- and super-scripted variable and the "alias" command in order to simplify calling the mentioned variable in the code?

In this way, one can use the aforementioned procedure (module) to establish a sub-super-scripted variable and address it easily.

@Carl Love Thank you for your reply. I kindly ask you to send me an example of a numerical, ie. non-symbolic, solution of a set of equations in parallel through either ‘solve’ command with ‘numeric’ option or the ‘fsolve’ command. I want to take an idea to numerically solve equations in parallel with that example.

best wishes

@Carl Love 

Dear Carl,

As you said, the coding in Maple is not effective to write a program in shared-memory parallelism. To the best of my knowledge, the shared-memory parallelism is far more effective than an independent-memory one. On the other hand, Maple weakly treats shared-memory parallelism because of the essence of built-in serial programming. It is a major drawback for Maple, unfortunately.

@Carl Love
Dear Carl,
Thank you for your attention to me. I wish there is either an idea or an algorithm to solve the equations in parallel.
Generally, I think Maple is almost weak to convert built-in serial procedures into parallel ones. Unfortunately, many of the commands are not be able to automatically execute in parallel.
Best wishes
 

@Carl Love 

Dear Carl,

If possible, please solve for me a set of algebraic equations by the ‘solve’ in parallel, as an example. I hope that I take an idea through your specimen to solve other equations.

Do you send me the solution of an arbitrary set of equations in parallel?

Bset wishes

@Carl Love

Dear Carl,
Thank you for your attention to my problem. I am not sure the mentioned equation, which I am now involving, can be solved even in serial!. However, the solution of either linear or nonlinear some equation(s) is a permanent challenge for me. It is very important for me to find a way to solve equation(s) in parallel because I almost involve a linear or nonlinear set of equations taking long running time in serial.
Can you help me to overcome the problem forever?
 

Best wishes

@Carl Love 

Dear Carl,
Thanks very much for your refined comments on my summary which is itself useful and informative.

@Carl Love 

 

 
Dear Carl,
before anything, I appreciate it you consumed so much time to reply to my five questions.
I want to re-write what you told me above. If anything is incorrect, let me know.
1) A process is an "independently running" with no sharing the memory and, consequently, needs more memory than the "Threads" package which fundamentally works based on the shared memory.
2) The word "task" is not essentially limited to "Threads". "Task" is the smallest work unit.
3) Too large "tasksize" may be resulted in being idle some processors, and too small "tasksize" could lead to more large CPU-time.
4) If code is threadsafe, "Threads:-Map" command is preceded Grid:-Map. Unfortunately, because of uncontrolled sharing in writing in memory, many commands in Maple are not threadsafe.
 
However, I have a question. What is your goal of talking about “overhead for task initializations”? In other words, what is the meaning of overhead to initial the tasks?
 
best wishes
1 2 Page 1 of 2