 |
 |
Van Snyder asked > I think P1 should be > 11 12 13 14 21 22 23 24 31 32 33 34 41 42 43 44 > and P2 should be > 1 5 9 13 > 2 6 10 14 > 3 7 11 15 > 4 8 12 16 > > 1. What do you think the "right" answers are for P1 and P2? > 2. Do you have a compiler that does what you (or I) expect? I am extremely hesitant to question Van's thoughts but is there any reason why the output of P1 => res(reshape(t2,(/16/))) should not be the same as the output of reshape(t2,(/16/)) ? (Similarly P2 => res(reshape(t1,(/4,4/))) should look the same as reshape(t1,(/4,4/)) ). Consequently I think P1 should be 11 21 31 41 12 22 32 42 13 23 33 43 14 24 34 44 and (p2(i,:),i=1,4) 1 5 9 13 2 6 10 14 3 7 11 15 4 8 12 16 which is exactly what I get from DEC compliers on VMS and Tru64. (I compared print *, p1 with print *, t2 and print *, p2 with print *,t1 as a sanity check) However, with Fujitsu on Linux my output from Van's code is as follows. This is obviously a problem area. Has somebody already reported this to Fujitsu ? T1 = 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 T2 = 11 12 13 14 21 22 23 24 31 32 33 34 41 42 43 44 P1 => res(reshape(t2,(/16/))) = 0*************** 0 0 0 0 0 0 0 0 0 1 P2 => res(reshape(t1,(/4,4/))) = 0*** 0 0 ****** 0 0 *** 0 0 0 *** 0 0 1 SHAPE(P1) = 16 SHAPE(P2) = 4 4 Bob ([log in to unmask]) %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%