Actually this highlights that using either 10 o. or | will both cause problems
for non-complex data that includes negatives. The following remedies this
rescale=:3 :0
1 rescale y
:
'data xtrma'=. 2{. boxopen y
'minX magX minY magY'=. (2{.x) ,&(-~/\@(<./ , >./)) |^:(16=3!:0) xtrma,,data
minX+magX*(data-minY)%magY
)
However I think I'd probably prefer to format it like this for
readability/maintainability.
rescale=:3 :0
1 rescale y
:
'data xtrmY'=. 2{. boxopen y
minmag=. -~/\@(<./ , >./)
'minX magX'=. minmag 2{. x
'minY magY'=. minmag |^:(16=3!:0) xtrmY,,data
minx + magX * (data - minY) % magY
)
_5 10 rescale 4 1 9 5 3.5 7;0 10 NB. "out-of-10" to "from _5 to 10"
1 _3.5 8.5 2.5 0.25 5.5
0 10 rescale 1 _3.5 8.5 2.5 0.25 5.5;_5 10 NB. back to "out-of-10"
4 1 9 5 3.5 7
> From: Marshall Lochbaum
> Sent: Wednesday, 22 September 2010 00:27
>
> Is there any reason you are using 10 o. instead of | ?
>
> From: Sherlock, Ric
> Sent: Tuesday, September 21, 2010 5:16 AM
>
> Thanks guys,
> The following combines the functionality/brevity of versions to date:
>
> rescale=:3 :0
> 1 rescale y
> :
> 'data xtrm'=. 2{.boxopen y
> 'lowX magX lowY magY'=. (2{.x) ,&(-~/\@(<./,>./)) 10 o. xtrm,,data
> lowX+magX*(data-lowY)%magY
> )
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
