>BEGIN RECORD Contrib > DIM pixel as Int > DIM weight as Double >END RECORD > >BEGIN RECORD CList > DIM n as int > DIM p as Contrib // max contributing pixels to each dest pixel, 50 is >example >END RECORD > >DIM myList(1200) as CList // width of image row/column, 1200 is example > >This works great. But... the 1200 for the Clist DIM is based on the >destination image size (CList) and the pixel contribution lists per dest >pixel for p can range from 1 to 120 or so, depending on the filter,etc. > >Bingo Question of the Day: How might I convert the above to an XREF >situation and load/recover the data I store in there now? > > >Load as is current: > >k = myList.n(i) >myList.p[k].pixel(i) = n >myList.p[k].weight(i) = weight Robert, What I would do to store an array of indeterminate length within a record is simply store a handle to the array. (In fact this is exactly what I am doing with Phil's arrays of next-words and usage counts in ZSpeak). BEGIN RECORD Contrib DIM pixel as Int DIM weight as Double END RECORD BEGIN RECORD CList DIM n as int DIM pHndl as handle // to XREF@ array of contrib END RECORD XREF@ p(999) as contrib DIM myList(1200) as CList // width of image row/column, 1200 is example k = myList.n(i) p = myList.pHndl(i) 'Set up contrib array in XREF p.pixel(k) = n p.weight(k) = weight (Not sure I got all the vars in correct places for your use, but you can work out the logic.) 0"0 =J= a y "