PolyhedralGeometry: Enhance docs of solve_mixed

[lkastner]

No description provided.

[HechtiDerLachs]

Seems reasonable, thanks for the work. The original issue which finally lead to this PR was the following: The docstring says we solve a system A*x = b under some condition. Here A and B are both matrices. Hence, given what I was used to from the context of lift-like commands from commutative algebra, I was expecting also x to be a matrix; probably the one who's j-th column multiplied with A becomes the j-th column of b. What was confusing me was that what I got was a matrix x which I could not even multiply with A! Example:

julia> A = ZZ[1 0; 1 -1]
[1    0]
[1   -1]

julia> B = ZZ[1; 1]
[1]
[1]

julia> C = ZZ[1 0; 0 1]
[1   0]
[0   1]

julia> x = solve_mixed(ZZMatrix, A, B, C)
[1   0]

julia> A*x
ERROR: Incompatible matrix dimensions
Stacktrace:
 [1] error(s::String)
   @ Base ./error.jl:35
 [2] *(x::ZZMatrix, y::ZZMatrix)
   @ Nemo ~/.julia/packages/Nemo/Tj33G/src/flint/fmpz_mat.jl:312
 [3] top-level scope
   @ REPL[95]:1

This is something that I personally still find confusing. But I understand that a lot of demands have already gone into the design of this function and I am not the only one who is confused about something. The docstring now might have helped me better to overcome my confusion, so I think this is an improvement.

[codecov]

Codecov Report

Merging #2798 (35b3545) into master (ebee98c) will decrease coverage by 0.05%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #2798      +/-   ##
==========================================
- Coverage   73.61%   73.56%   -0.05%     
==========================================
  Files         455      455              
  Lines       64469    64469              
==========================================
- Hits        47459    47429      -30     
- Misses      17010    17040      +30     

Files Changed	Coverage
src/PolyhedralGeometry/solving_integrally.jl	ø