Using IBM i? Need to create Excel, CSV, HTML, JSON, PDF, SPOOL reports? Learn more about the fastest and least expensive tool for the job: SQL iQuery.
If I understand correctly, you're going to need to keep track of it yourself. As you're paging through the table, you only see one page of records at a time in the subfile. When you page down or up, you need to read the subfile rows, and store any changes, including non-database fields such as an Opt field, then load the next or previous page.
I probably wouldn't write every subfile record to the temp storage area, only those that have changes. As you load each page, you'll want to check your storage area to see if your record is already there. Then either load form the temp storage, or go get the info from the database.
The form of the temp storage area is up to you. Could be a temp table, an array DS, or something else which you're comfortable with. Don't forget to include all subfile fields, including Opt columns and such, so that you can restore them if you happen to page back to that record.
As Dale suggested, my way is to use a Data Structure with the DIM(15000) keyword on it to set it to the number of entries you could have. I then map it to a point to a User Space--which gives me the 16MB size with auto-grow features. Then I use the SORTA %SUBARR( mySubFile : 1 : n ) where "n" is the current number of elements.
I normally use LIKEREC(mySFL:*ALL) when defining the DS to make sure its the same as the correspending Subfile.
Then as you page through it a page at a time, you simply increament the page number to the PGNBR * PGSIZE and you have the array element of the top-most subfile entry.
As they roll throught he subfile, you set the OPT (or OPTION) subfield to their option/choice and save it. F5=Refresh clears that entry.
One other method, would be to create an SQL Cursor as a SCROLLING CURSOR and then just roll back and forth through the records--but that option doesn't give you an integrated way to save the OPT/OPTION value.