Yes, but where things are located in that block of memory depends on how you calculate the location in the memory - particularly when dealing with 2D arrays. Since you are also COPY NY+1 elements more than your original code, I'd say that it is possible that it's going wrong for that reason.
--
Mats
Compilers can produce warnings - make the compiler programmers happy: Use them!
Please don't PM me for help - and no, I don't do help over instant messengers.
Hi again,
No, even with that fixed (allocating the coherent number of elements), I still obtian the same allocation. The problem is surely in the part of routine where I assign the variable:
and I don't know how to modify it correctlyCode:k=0; for(j = 0; j < NY; ++j){ for(i = 0; i < NX; ++i){ vars[0][k] = zonal[j][i]; k++; } }
That's what I meant by "copy" - you are writing NY * NX elements, not (NY-1) * (NX-1) elements to the new array. So the stride (step between each line) in the 1D layout is going to be different.
--
Mats
Compilers can produce warnings - make the compiler programmers happy: Use them!
Please don't PM me for help - and no, I don't do help over instant messengers.
Hi,
yes, I understand what you mean, but I correctly it accordingly. Is the loop conceptually correct or am I assigning the wrong part of memory then?
thank you again
The original loop from the original code produces:
Your new code produces:Code:1 2 3 4 1 2 3 4 1 2 3 4
Now, if you use the write method to extract this, you will get:Code:1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
written to disk. This assumes that I have understood things correctly. Does this make some sense compared to what you are seeing?Code:1 2 3 4 5 1 2 3 4 5 1 2
--
Mats
Compilers can produce warnings - make the compiler programmers happy: Use them!
Please don't PM me for help - and no, I don't do help over instant messengers.
Hi again Mat,
The error was exactly on that! Thank you so much! I did different tries to prove it against the original code, and yes, it was the NX-1, NY-1 where I was allocating a bigger memory!
I can't believe I still make certain mistakes after years of programmingThis dynamic allocation is very powerful, but delicate at the same time if not taken well care of
Thank you again for helping patiently
Best regards
CFD
Well, it ISN'T the allocation itself that is wrong - it is the loop that copies the zonal data into the variables[0][k] that goes wrong because you are incrementing k one more than the write loop does for each row.
--
Mats
Compilers can produce warnings - make the compiler programmers happy: Use them!
Please don't PM me for help - and no, I don't do help over instant messengers.
Correct, now it's all clear;
thank you again
Simon
