Another property I find interesting, but I haven't quite understood the value of, are the "extended" patterns like these.
I think of them as transformations or modulations or shifts(?). Maybe VQC will chip in with a more useful / descriptive name (if they're useful). But when you look at (1, 1), (4, 4), (9, 9) you can see how they are all (1, 1), but the a-b connections get "moved" or "shifted". Like in (1, 1, 1) you have a=1, b=5, but in (4, 4, 1) you have a=1, b=13. Moving squares "shifts" the patterns.
This can be applied to any record. Take (1, 5). You have:
(1, 5, 4.0, 3, 1.0, 17.0)
(1, 5, 12.0, 7, 5.0, 29.0)
(1, 5, 30.0, 13, 17.0, 53.0)
(1, 5, 46.0, 17, 29.0, 73.0)
(1, 5, 76.0, 23, 53.0, 109.0)
(Mind the structure of the cells)
Multiply the square by the e and n values, example 1 * 4, 5 * 4 =(4, 20):
(4, 20, 7.0, 6, 1.0, 53.0)
(4, 20, 19.0, 14, 5.0, 73.0)
(4, 20, 43.0, 26, 17.0, 109.0)
(4, 20, 63.0, 34, 29.0, 137.0)
(4, 20, 99.0, 46, 53.0, 185.0)
Here you can see how the values are shifted. This can be done continuously by multiplying squares to the e and n.
Again, except we multiply 9, giving (1 * 9, 5 * 9) =(9, 45)
(9, 45, 10.0, 9, 1.0, 109.0)
(9, 45, 26.0, 21, 5.0, 137.0)
(9, 45, 56.0, 39, 17.0, 185.0)
(9, 45, 80.0, 51, 29.0, 221.0)
(9, 45, 122.0, 69, 53.0, 281.0)
Just thinking out loud here. But when you look at (e, c) (when you treat c like n in column e) the number of chains (or sequences) tend to match the number of factors. In this case we "extend" the number of factors (because we increase the number of chains / sequences).
It looks like, by multiplying (e, n) by 4 or 9 or 16, it's the same as multiplying those factors by those squares. Maybe I'm just rambling here and this is just a result of the way the grid is structured. VQC, you're free to fill inn (even if these patterns aren't strictly related to our problem).