doctest failure in doc/en/constructions/calculus.rst due to upgrade to Maxima 5.19.1

[sagetrac-drkirkby]

On Solaris (SPARC), the following test failed. Both ECL and Maxima were updated - ECL version 9.8.4, Maxima version 5.19.1

----------------------------------------------------------------------
| Sage Version 4.1.1, Release Date: 2009-08-14                       |
| Type notebook() for the GUI, and license() for information.        |
----------------------------------------------------------------------
Thu Aug 20 20:02:37 BST 2009
dsage-trial tmp directory doesn't exist - creating ...
This script will run the unit tests for DSage

File "/export/home/drkirkby/sage/sage-4.1.1/devel/sage/doc/en/constructions/calculus.rst", line 117:
    sage: maxima(f).powerseries(x,0)
Expected:
    ('sum((-1)^i2*2^(2*i2)*bern(2*i2)*x^(2*i2)/(i2*(2*i2)!),i2,1,inf))/2
Got:
    'sum((-1)^i2*2^(2*i2-1)*bern(2*i2)*x^(2*i2)/(i2*(2*i2)!),i2,1,inf)

Component: calculus

Keywords: maxima

Issue created by migration from https://trac.sagemath.org/ticket/6782

[aghitza]

comment:2

This one is trivial; see attached patch.

[aghitza]

Changed keywords from none to maxima

[aghitza]

Attachment: trac_6782.patch.gz

apply after spkg's at #6564 and #6699

[sagetrac-drkirkby]

comment:3

The log was generated with Maxima 5.19.1, not 5.19.0, so I'm changing the title slightly to reflect this.

[sagetrac-drkirkby]

comment:4

Are these mathematically equivalent? It seems to me the factor in the numerator has changed from (2i2) to (2i2-1) and denominator from 2 to 1. That seems a different result to me.

I tried what I think the test is doing in Mathematica and got:

In[7]:= f=Log[Sin[x]/x]

            Sin[x]
Out[7]= Log[------]
              x

In[8]:=  Series[f,{x,0,12}]

          2    4      6      8       10           12
        -x    x      x      x       x        691 x           13
Out[8]= --- - --- - ---- - ----- - ------ - ---------- + O[x]
         6    180   2835   37800   467775   3831077250

In[9]:=

I've no idea what's right or not.

Dave

[aghitza]

comment:5

The difference is 2^(2*i2)/2 versus 2^(2*i2-1). Yes, they are the same.

[sagetrac-mvngu]

comment:7

This is fixed by #6699.