unnecessary blank in latex expression of multivariate polynomial

[dkrenn]

In the LaTeX representation of a multivariate polynomial is an unnecessary blank when variable is concatenated with a -, see

sage: R.<X,Y> = ZZ[]
sage: latex(X-Y)
X -  Y
sage: latex(X^2-X)
X^{2} -  X

It does not happen with +

sage: latex(X+Y)
X + Y

or with a constant

sage: latex(X-1) 
X - 1

In the univarate case the problem does not appear.

Note that this "wrong" output is no problem (i.e. a correct input for LaTeX)

[ppurka]

Does this really need a fix? The latex output can be used in two ways:

1. Input to a latex document.
2. Display on the notebook with typeset.

In both the cases, the output will be fine.

[aapitzsch]

The attached patch removes the unnecessary blank.

But now we also have

sage: latex(-X^2-X)
-X^{2} - X

instead of

sage: latex(-X^2-X)
- X^{2} - X

Is this reasonable or should we find a solution to keep the blank between the leading variable and its sign ("-")?

[novoselt]

No, it is not reasonable. As it was pointed out, the current output is correct and the only reason to "fix" is aesthetics. From this point of view, getting rid of a double space and swallowing a single one is not good - better to have more spaces then less.

[aapitzsch]

Currently we have

sage: latex(-x^2-x)
-x^{2} - x
sage: R.<y> = ZZ[]  
sage: latex(-y^2-y)
-y^{2} - y

Notice, there is no blank in the leading term, if it's no multivariate polynomial. So why do we need one if it's a multivariate polynomial?

IMO it should be consistent.

[novoselt]

Replying to aapitzsch:

IMO it should be consistent.

Yes, and this means that there is always exactly one space, if it is missing, it is the same "bug" as with double spaces or even worse since it makes reading more difficult and interferes with word wrapping.

A more useful LaTeX alteration than decreasing spaces would be getting rid of {} around singletons, i.e. instead of x_{1}^{2} it would be better to have x_1^2, again for reasons of readability and subsequent "manual manipulations". Macaulay2 is making such a change, as can be seen in one of the failing tests here: ​http://trac.sagemath.org/sage_trac/ticket/11710#comment:23

[aapitzsch]

Apply only trac_12908_latex_blank.patch.

This patch doesn't fix all "space bugs" in LaTeX expressions, only those which are related to the bug mentioned at the top.

[dsmatth]

I attempted to apply trac_12908_latex_blank.patch to Sage 5.10, but it appears rotted.

~/sage/devel/sage$ hg qimport trac_12908_latex_blank.patch                                                                                   
adding trac_12908_latex_blank.patch to series file                                                                                           
~/sage/devel/sage$ hg qpush                                                                                                                  
applying trac_12908_latex_blank.patch                                                                                                        
patching file doc/de/tutorial/latex.rst                                                                                                      
Hunk #1 succeeded at 73 with fuzz 1 (offset -5 lines).                                                                                       
patching file doc/en/tutorial/latex.rst                                                                                                      
Hunk #1 succeeded at 70 with fuzz 1 (offset -10 lines).                                                                                      
patching file sage/algebras/free_algebra_quotient_element.py                                                                                 
Hunk #1 FAILED at 136                                                                                                                        
1 out of 1 hunks FAILED -- saving rejects to file sage/algebras/free_algebra_quotient_element.py.rej                                         
patching file sage/combinat/free_module.py                                                                                                   
Hunk #1 FAILED at 1032                                                                                                                       
1 out of 1 hunks FAILED -- saving rejects to file sage/combinat/free_module.py.rej                                                           
patching file sage/rings/polynomial/polynomial_quotient_ring_element.py                                                                      
Hunk #1 FAILED at 185                                                                                                                        
1 out of 1 hunks FAILED -- saving rejects to file sage/rings/polynomial/polynomial_quotient_ring_element.py.rej                              
patch failed, unable to continue (try -v)                                                                                                    
patch failed, rejects left in working dir                                                                                                    
errors during apply, please fix and refresh trac_12908_latex_blank.patch 

On a side note, I would like to vote that the space between the sign and leading term should not be there. I am new to sage and trac, but I prefer

-X - Y

as opposed to

- X - Y

The leading negative indicates an inverse where as the subtraction between objects is an operation (adding an inverse). That is:

-X + -Y = -X - Y

Since the original patch has the same problems applying, I will change the status to needs work.

[novoselt]

Are there any particular style guidelines that recommend smaller space for unitary minus? I personally prefer the same space as for binary for consistency.

[rws]

FWIW, Don Knuth doesn't use spaces in his TeXbook. A better style guide you won't get.