marked mpoly_all() as not tested

[demosd235]

Demo at SD23.5, also increases coverage!

---

Apply trac_9494-doctest_mpoly_all.2.patch​ to the Sage library.

[demosd235]

Add a comment to a ticket here.

[demosd235]

Is this ready yet?

[aapitzsch]

According to ​http://www.sagemath.org/doc/developer/conventions.html#docstring-markup-with-rest-and-sphinx it should be EXAMPLES: instead of EXAMPLE::.

It's enough to put only 3 dots between <BLANKLINE> and System but this is okay as well.

[rbeezer]

If an "EXAMPLES" block just has one example, then it is OK to drop the S. And if there is no prose before the test, then a double-colon can be used to indicate verbatim text next.

There are excess ... but they cause no harm. This ticket looks good to me. Am I missing something? It looks ready for a positive review to me.

I am working on this as a high-priority ticket for Bug Days 32.

Rob

[benjaminfjones]

The patch looks good to me and passes doctests:

bjones@sage:~/sage/sage-4.7.1/devel/sage$ ../../sage -t sage/tests/benchmark.py 
sage -t  "devel/sage-9494/sage/tests/benchmark.py"          
	 [8.6 s]
 
----------------------------------------------------------------------
All tests passed!
Total time for all tests: 8.6 seconds

I'm giving it a positive review.

[leif]

Actually, mpoly_all() needs work, since it is apparently unable to interrupt (e.g.) Mathematica a second time.

On redhawk, where /home still isn't mounted, Mathematica is installed but fails to startup (prompting the user), such that I get:

leif@redhawk:/$ time sage -c "from sage.tests.benchmark import mpoly_all; mpoly_all()" 



Compute (x_0 + ... + x_99) * (x_100 + ... + x_200) over Rational Field (use singular for Sage mult.)
  System      min         avg         max         trials          cpu or wall
* sage        0.049767    0.049767    0.049767    1                  0.049767w
Unable to start magma
computation timed out because alarm was set for 60 seconds
Unable to start macaulay2 because the command 'M2 --no-debug --no-readline --silent -e 'ZZ#{Standard,Core#"private dictionary"#"InputPrompt"} = lineno -> "_EGAS_ : ";ZZ#{Standard,Core#"private dictionary"#"InputContinuationPrompt"} = lineno -> "_EGAS_ : ";printWidth = 0;lineNumber = 10^9;'' failed.




Compute (x_0 + ... + x_199) * (x_200 + ... + x_400) over Rational Field (use singular for Sage mult.)
  System      min         avg         max         trials          cpu or wall
* sage        0.089736    0.089736    0.089736    1                  0.089736w
Unable to start magma
Interrupting Mathematica...

which does not terminate.

What is this doctest supposed to "test" at all?

The output is random, so it doesn't really test anything, except perhaps which external programs are installed, but that's nothing subject to a doctest, so IMHO this ticket should either be closed as invalid, or the "doctest" (i.e., the example) should be marked # not tested, at least until mpoly_all() got fixed.

[benjaminfjones]

marked mpoly_all() as not tested

[benjaminfjones]

I marked mpoly_all() as # not tested. The new patch applied cleanly to 5.0.beta12 and passes tests on sage.math.washington.edu where Mma is installed. I'm running tests on redhawk now to double check.

[benjaminfjones]

There seems to still be an issue with the Mma installation on redhawk and the mpoly_all() test is still unable to interrupt the crashed process all the time:

Compute (x_1 + 2*x_2 + 3*x_3 + ... + 128*x_128) * (129 * x_129 + ... + 257*x_257) over Rational Field (use singular for Sage mult.)
  System      min         avg         max         trials          cpu or wall
* sage        0.049826    0.049826    0.049826    1                  0.049826w
Unable to start magma
Mathematica crashed -- automatically restarting.
computation timed out because alarm was set for 60 seconds
Unable to start macaulay2 because the command 'M2 --no-debug --no-readline --silent -e 'ZZ#{Standard,Core#"private dictionary"#"InputPrompt"} = lineno -> "_EGAS_ : ";ZZ#{Standard,Core#"private dictionary"#"InputContinuationPrompt"} = lineno -> "_EGAS_ : ";printWidth = 0;lineNumber = 10^9;'' failed.




Compute (x_1 + 2*x_2 + 3*x_3 + ... + 256*x_256) * (257 * x_257 + ... + 513*x_513) over Rational Field (use singular for Sage mult.)
  System      min         avg         max         trials          cpu or wall
* sage        0.149662    0.149662    0.149662    1                  0.149662w
Unable to start magma
Interrupting Mathematica...
^C
Unable to start macaulay2 because the command 'M2 --no-debug --no-readline --silent -e 'ZZ#{Standard,Core#"private dictionary"#"InputPrompt"} = lineno -> "_EGAS_ : ";ZZ#{Standard,Core#"private dictionary"#"InputContinuationPrompt"} = lineno -> "_EGAS_ : ";printWidth = 0;lineNumber = 10^9;'' failed.

But I'd say that's an issue for another ticket.

[rws]

Moved patch to git.

---

New commits:

​1803137

trac #9494: marked mpoly_all() as not tested

[git]

Branch pushed to git repo; I updated commit sha1 and set ticket back to needs_review. New commits:

​6f109fa

Merge branch 'develop' into ticket/9494