unexpected behavior of issl card

[Bernard Berthomieu]

Hello,

The iscc script below (a simplified version of a much larger one) returns count 188.
But if I'm replacing in it the lines defining U an V by these:

  U = 6 e22 + 6 e52 + 7 e32 + 7 e42 + 8 e12 and
  V = 6 e21 + 6 e51 + 7 e31 + 7 e41 + 8 e12 and

then iscc returns count 192.

I can't explain why swapping the components of a sum would change the cardinality
of that set, so what am I doing wrong ?

I'm using iscc from the barvinok-0.41 distribution.

Thanks for any hint,
  Bernard.
 

---

card {

[U, V, x1, x2, x3, x4, x5] :

exists e11, e12, e21, e22, e31, e32, e41, e42, e51, e52 :

# result is 188
U = 8 e12 + 6 e22 + 7 e32 + 7 e42 + 6 e52 and
V = 8 e11 + 6 e21 + 7 e31 + 7 e41 + 6 e51 and

# result is 192
# U = 6 e22 + 6 e52 + 7 e32 + 7 e42 + 8 e12 and
# V = 6 e21 + 6 e51 + 7 e31 + 7 e41 + 8 e12 and

x1 + e11 + e12 = 1 and
x2 + e21 + e22 = 1 and
x3 + e31 + e32 = 1 and
x4 + e41 + e42 = 1 and
x5 + e51 + e52 = 1 and

x1 >= 0 and
x2 >= 0 and
x3 >= 0 and
x4 >= 0 and
x5 >= 0 and

e11 >= 0 and e12 >= 0 and
e21 >= 0 and e22 >= 0 and
e31 >= 0 and e32 >= 0 and
e41 >= 0 and e42 >= 0 and
e51 >= 0 and e52 >= 0

};

[Sven Verdoolaege]

On Thu, Mar 23, 2023 at 09:18:07AM -0700, Bernard Berthomieu wrote:
> Hello,
>
> The iscc script below (a simplified version of a much larger one) returns
> count 188.
> But if I'm replacing in it the lines defining U an V by these:
>
> U = 6 e22 + 6 e52 + 7 e32 + 7 e42 + 8 e12 and
> V = 6 e21 + 6 e51 + 7 e31 + 7 e41 + 8 e12 and
>
> then iscc returns count 192.
>
> I can't explain why swapping the components of a sum would change the
> cardinality
> of that set, so what am I doing wrong ?

What is a "component of a sum"?

The two pairs of (U, V) expressions have different values and
apparently also different numbers of values,
so it seems perfectly reasonable for the two sets
to have different numbers of elements.

skimo

[Bernard Berthomieu]

Oops, much sorry, just spotted a typo in my sums.

Thanks !