I won't be overly confident on order of events ,   data have; input person date : $20. product $ paid : dollar.; cards; 1 1/25/15 apple $50 1 1/25/15 apple $43 1 1/25/15 apple $-43 2 1/30/15 orange $50 3 1/31/15 apple $55 ; data want; ina=0;inb=0; merge have(where=(paid lt 0) in=ina) have(where=(paid gt 0) in=inb); by person date product; if not (ina and inb); run; ... View more

Took me some time to get rss into Gmail. Easy on mobile (feedly) and outlook (built in feature). Thanks a lot, Chris! ... View more

Can't skip this party     How about this non-orthodox solution: options missing=''; data test; input id m1 m2 m3 m4 m5 m6 m7 m8 m9 m10 m11 m12; datalines; 3424 0 1 0 0 0 0 1 1 1 0 1 1 5432 1 1 1 1 1 1 0 0 1 0 1 0 3930 0 1 0 1 1 1 1 0 0 1 1 1 6053 0 1 0 1 0 0 0 0 0 0 0 0 9493 1 1 1 1 1 0 0 0 0 0 1 1 4984 1 1 0 0 0 1 1 0 0 0 1 1 2039 0 1 1 1 1 0 0 0 0 1 0 0 2039 0 0 0 0 0 0 0 0 0 0 0 0 ; data want; set test; length _3 $12 _4 $12 _1 $12; _1=cats(of m:); do _i=1 by 1 until (missing(_3)); _3=scan(_1,_i,'0'); _4=max(_4,_3); end; continuity=lengthn(strip(_4)); drop _:; run; ... View more

This is a not bullet-proof approach, I need to leave office now so I may get back to you if this code needs patch. Anyway, hope this can get you started;   data youth; infile cards truncover; input (Y_ID age sex rank1 rank2 rank3 rank4) (:$8.); cards; 122 12 m B C 15 14 m B C 5 15 m B D C 666 16 f C D 561 16 m D B C 8 21 m A D C B 46 14 m D C B 5555 17 m A B D C 8484 13 f D C 48 5 m C ; data location; input Bed_ID $ Location $; cards; 1 A 2 B 3 C 4 D 5 D 6 D 7 A 8 C 9 B 10 D ; data want; if _n_=1 then do; if 0 then set location; declare hash h(dataset:'location', multidata:'y' ); h.definekey('location'); h.definedata(all: 'y'); h.definedone(); end; set youth; array rank rank:; do over rank; if h.find(key:rank) = 0 then do; output; rc=h.removedup(key:rank); leave; end; end; drop rc; run;  btw, I doubt SQL will support the feature you are looking for, even it could, it can be very complex.   ... View more

it is really hard to envision any pratical approaches before you show some sample data, both HAVE and WANT. For example, when you say 'not available' (which is a key concept in your description), which of the following do you really mean in data language: 1. The field is missing. 2. The field is not missing, but the value is not matching. 3. Others?   If it is No1. SAS has a straightforward function coalesce() worth looking up. Could be as simple as: proc sql; create table want as select * from h1 left join h2 on h1.location=coalesce(h2.rank1,h2.rank2, h3.rank3...down then line etc) ; run;   If it is No2 or 3, then it may involve some more complex coding. However, without seeing your data structure, this is all guessing. ... View more

Maybe we should get a ballot to reinstate these nice features? ... View more

Rename statement or Rename data set option supports batch renaming of your kind:   data want; set have; rename bd_dx_1-bd_dx_32=nad_dx_1-nad_dx_32; run; ... View more

For what it is worth, can also just be one-liner: abv=prxchange('s/(^[^_]+_[^_]+)(_[^_]+$|\s*$)/$1/', -1, "&dsn.");   ... View more

@Astounding, chance is very slim, yet possible:   data test; input num; char=put(num,8.); char_1=ifc(num>.,put(num,8.),''); char_2=ifc(not missing(num),put(num,8.),''); cards; 100 . ._ .A .Z ; ... View more