SWP 6.0.26 Computation Issues

jm's picture

The attached .sci file illustrates some problems I have with Compute in SWP 6.0.26.

Am I doing something illegitimate? Or are these bugs in SWP 6.0.26?

SWP6026ComputationIssue.sci143.05 KB
pviton's picture

At least on Win7, what seems

At least on Win7, what seems to happen is that SWP doesn't really cease to respond: it's still working on the SolveExact problem, as you can see if you open the Task Manager and look at the memory utilized by the MuPad kernel. Eventually it just runs out of memory. In the meantime, I can actually type in text and successfully save the revised document.

It's not clear how the program *should* respond in such a situation, ie when it's given a really difficult computational task. The Stop button on the computation toolbar is active and will stop the computation if you get tired of waiting.

Of course, on the Mac behavior may be different, I don't know. But on Windows I don't think the behavior is unreasonable.

jm's picture

On my Mac, SWP really does

On my Mac, SWP really does cease to respond. The screen shows the twirling ball that indicates non-response, the Stop button on the computation toolbar is inactive, and I cannot enter any text or save the document. Right-clicking the SWP icon gives the message "Application Not Responding". The only way I can get out of the situation is using "Force Quit" on SWP.

Did you try the Compute > Simplify issue that I also mentioned to see if you get the right answer to that using Win7?

pviton's picture

I tried jm's first (solve)

I tried jm's first (solve) problem in stand-alone
Maple18; the response was extremely fast,
though I'm not sure how useful the solution is.

I also tried to simplify the lhs of jm's equation:
this was also extremely fast, but again I'm not sure
how useful it is.

See the attached zip archive for the session as
an html file; I've also included the maple code as
mws files with and without the computational results,
in case they're helpful.

jm - I was surprised at just how fast all this was.
Given this, you might want to check that I've entered
your equation correctly; if not, let me know what needs
to be changed, and I'll be glad to try again.

Two more general points: first, except for plotting,
does SWP6 have any significant computational advantages
over SWP5? The computation menus look exactly the same.
(Also, the date of the folders in the MuPad tar file is
Apr 2011, ie six years old: is MuPad still being developed?)

Second, I seem to recall that one of the advantages
of SWP6 was that it was intended to make interfacing
with other programs easier. Has anything been done
in this area? For example, has anyone taken a look
at the free Sage system, which seems to be able to
produce LaTeX output (which could presumably in principle
be read by swp6), and (a) contains interfaces to many
other applications, including R (which would I think be
greatly welcomed by empirical researchers using SWP6)
and some specialized computation engines (eg for group
theory) for the mathematicians; and (b) is being
aggressively developed.