Yes, odd handles the odd tricks and after I posted it I realized that
for readability I should have split it into above-the-line and
below-the-line as
'below above' =. (2^dbl) * (10 0 * suit=4) + (level,y) * suit { 2 3 # 20 30
and then {.odd and {:odd could be replaced.
Hey, wait, I didn't include the partscore bonus! I'll append a fixed
version.
To handle rubber bridge or Chicago, you have other considerations:
0. most games still allow honors;
1. You need to be able to specify the partscore bonus: 0 except for the
last hand, then 300 (or whatever local custom calls for);
2. You might need to consider that some old-fashioned players don't
accept the scoring changes of 1987; for them doubled nonvulnerable
undertricks go 1-3-5-7... , and 50 for the insult is not changed to 100
when there is a redouble.
3. As you say, you would need to pass in any existing partscore. Maybe
you would add on to x, making it
'level suit vul dbl leg psbonus' =. 6 {. x
Second version below
Henry Rich
NB. score bridge (duplicate and Chicago)
NB. x is contract: level suit vul dbl [leg partscorebonus]
NB. suit is 0=C, 4=NT
NB. dbl is 0, 1, or 2
NB. y is result, neg=down, nonneg=made
NB. result is score for contracting side
NB. Does not support honors
bridgescore =: 4 : 0"1 0
'level suit vul dbl leg psb' =. x , (#x) }. 0 0 0 0 0 50
if. y < 0 do.
if. dbl do.
dbl * 100 + (y * 300) - 100 * ((y * -.vul) >. _2)
else.
y * vul { 50 100
end.
else.
'b a' =. (2^dbl) * (10 0 * suit=4) + (level,y) * suit { 2 3 # 20 30
pts =. (dbl * 50) + b + (100 <: leg+b) { psb , vul { 300 500
pts =. pts + (<vul,level) { _8 {."1 ] 500 1000 ,: 750 1500
pts =. pts + dbl { a , (>:vul) * y * 100 200
end.
)
Brian Schott wrote:
> Henry,
>
> Yes, I like your crisp code very much. I want to alter it to
> accommodate existing part scores and splitting the result into above-
> and below-the-line portions.
>
> It seems that the temporary noun `odd` produces the below-the-line
> portion as its first element, right? And that could be partitioned
> from the final result to produce the two parts?
>
> On the other hand if the existing part scores were appended to the y
> argument, making it a vector argument, wouldn't that be a way to
> produce a multipart `result` that would resemble the resulting score,
> at least for winning contracts?
>
> Thank you so much for sharing this code.
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
