[Newbies] Arrays within arrays

[Charles Gray]

<snip>

> > strength := #( #(0 0 0 0) #(0 0 0 0) #(0 0 0 0) #(0 0 0 0) #(0 0 0 0) #(0
> 0 0 0)
> > #(2 1 0 0) #( 8 4 0 1) #(14 6 0 2) #(30 12 1 4)).
>
> <snip>
>  The literal syntax has a bit of context sensitive-ness.  In a
> literal array, the # is not necessary to mean nested literal array.
> So,
>
>   strength := #((0 0 0 0) (0 0 0 0) (0 0 0 0) (0 0 0 0) (0 0 0 0) (0 0 0 0)
>                (2 1 0 0) (8 4 0 1) (14 6 0 2) (30 12 1 4)).
>
> Also works.
>
> <snip>
>
>  It doesn't refuse the assignment, right?  And you can't use a
> variable in a literal array to mean the contents of the variable.
>
>  Or, the above is actually a simplified code, and some other object
> is holding onto the reference to the "Array new: 10" array?  If you
> assign a new array created from the literal array syntax into the same
> name variable, the guy who is holding onto the reference to "Array
> new: 10" wouldn't see the change.
>
> -- Yoshiki
>
>  Just curious, but what do these numbers represent?
>
> Gee, that was easy. Thanks. I never guessed that you had to leave the #'s
off inside the array. Those particular numbers don't mean anything. They
were just an example of my problem. In my bridge dealer program, there are
four sets of values assigned to each card. The first two numbers give the
relative strength of a card when playing in a suit contract and when playing
in a notrump contract. The numbers are based on the extensive double-dummy
analysis performed by Thomas Andrew. The third number is is what some
players of the Moscito bidding system call AKQ or slam points, and the
fourth number is simply a binary representation that I use for determining
loser count. I wasn't trying to be sneaky. The real numbers from 8 up to ace
are:
#((2 1 0 0) (4 3 0 1) (10 6 0 2) (23 14 0 4) (43 27 1 8) (74 51 2 16) (115
82 3 16)).
The totals are #(271 184 6 31). In the case of the first two numbers, I add
them up for each suit, multiply by ten and divide by the total giving me a
number from 0 to 40 that is similar to the work point count commonly used to
evaluate bridge hands using 4, 3, 2, 1 for ace through jack for strength
evaluation. Now that I think about it I can simply multiply each of the
original numbers by 10 to eliminate that step.

I know that's more than you what to know about the numbers but you asked.
The traditional 4, 3, 2, 1 evaluation undervalues aces and kings. Different
numbers are used for suit play and notrump play because there is no chance
that a card will be trumped during play. A human would find it difficult to
use this method of evaluation but it is no problem for a computer. By the
way, I was just converting procedures I had written in Forth and Ruby to
Smalltalk. Now if I could just learn not to think in sequential terms.

Charlie

Charlie
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/beginners/attachments/20080905/659de1fe/attachment.htm