Initializing a 2D Array in C

Hey guys,

I am trying to initialize a 2D array in C. For example, I have the matrix:

0 1 0 0
0 1 1 0
1 0 1 1
0 1 1 1

Now I think this would work:

Code:

---

int A[4][4];

A[0][0] = 0;
A[0][1] = 1;
A[0][2] = 0;
A[0][3] = 0;
A[0][4] = 0;
A[1][1] = 0;
A[1][2] = 1;
... ETC....
A[4][4] = 1;

---

But this is a real pain. I will have larger matrices like 17x17 and doing this would be a real annoyance.

Is there a way to do something like:

Code:

---

int  A[4][4] = {0 1 0 0; 0 1 1 0, ... etc}

---

?

Thanks!

Code:

---

A[3][4] = { { 0, 0, 0, 0 }, { 0, 0, 0, 0 }, { 0, 0, 0, 0 } };

---

Or, for larger 2D arrays . . .

Code:

---

int array[100][100];
int x, y;
for(x = 0; x < 100; x ++) {
    for(y = 0; y < 100; y ++) array[x][y] = 0;
}

---

Better than a few thousand chars of static initializers ;)

Quote:

---

Originally Posted by Nightowl

Or, for larger 2D arrays . . .

Code:

---

int array[100][100];
int x, y;
for(x = 0; x < 100; x ++) {
    for(y = 0; y < 100; y ++) array[x][y] = 0;
}

---

Better than a few thousand chars of static initializers ;)

---

But if you are setting all elements to zero, just doing

Code:

---

int array[100][100] = { {0 } };

---

works just fine.

--
Mats

Good point. Though the for loop can be used at non-initialization times, at any point in the program. And, to me, the for loop is a bit clearer. ;)

You'll learn malloc() next and it'll shine a whole new light on this

Quote:

---

Originally Posted by shiryu3

You'll learn malloc() next and it'll shine a whole new light on this

---

shiryu, that is overly condescending. Also, you should specify who you are talking to . . .

And yeah, you are correct. That is the code I use for allocated arrays.

Thank you for the answers, guys.

However, for this code:

Quote:

---

Originally Posted by Nightowl

Or, for larger 2D arrays . . .

Code:

---

int array[100][100];
int x, y;
for(x = 0; x < 100; x ++) {
    for(y = 0; y < 100; y ++) array[x][y] = 0;
}

---

Better than a few thousand chars of static initializers ;)

---

This would just fill the matrix with zeroes. How can this be used to fill a matrix with specific values?

Also, I do know about malloc(). I don't know how I can use it to create a matrix with specified values, though. I also don't see what benefits it has over what brewbuck mentioned in his post.

I know there are plenty of ways to create a matrix filled with 0s, but I'd like to see how there can be a more efficient way to fill a matrix with specific values than what brewbuck posted.

Thanks again for the help.

> but I'd like to see how there can be a more efficient way to fill a matrix with specific values than what brewbuck posted.

Nope. Nothing more efficient than that, as far as initialisation/setting goes.

Quote:

---

Originally Posted by Cell

... but I'd like to see how there can be a more efficient way to fill a matrix with specific values than what brewbuck posted...

---

A more efficient way would be to use the pointer and offset version.

Quote:

---

Originally Posted by itCbitC

A more efficient way would be to use the pointer and offset version.

---

more effitient in what way?

Think on it. Having two loops rather than one? . . .

Code:

---

int array[100][100];
int x;
for(x = 0; x < sizeof(int)*100*100; x ++) *(array+x) = 0;

---

. . . should be faster than what I posted before.

Just more confusing, that's all . . .

Quote:

---

Originally Posted by Nightowl

Think on it. Having two loops rather than one? . . .

Code:

---

int array[100][100];
int x;
for(x = 0; x < sizeof(int)*100*100; x ++) *(array+x) = 0;

---

. . . should be faster than what I posted before.

Just more confusing, that's all . . .

---

things like that should not be "thinked" - or exemine asm code or - better - use profiler.

I suppose both loops will be replaces by memset and will have same time exactly

If using a good compiler such as gcc, with all optimizations turned on, certainly. However, I don't think that gcc would do that kind of level of optimization by default . . .

Besides, the second one is probably almost exactly how memset() is implemented . . .

Quote:

---

Originally Posted by Nightowl

If using a good compiler such as gcc, with all optimizations turned on, certainly. However, I don't think that gcc would do that kind of level of optimization by default . . .

Besides, the second one is probably almost exactly how memset() is implemented . . .

---

why to bother about performance is you do not turn optimization on?

if you debug something- you'd like to have most readable code, not the quickets, step-by-step has no gain from speed here.

and I doubt that memset is written with the loop like this - asm could move chunks of bytes longer than int in one instruction...

Point. And yes, the first example, the nested for loops are the most readable, IMO. Besides a call to memset(), as you point out so nicely. ;)

Note that optimizations really mess up GDB, however. If you're working on a project that has lots of calculations to go through before it reaches a breakpoint, speed is always nice to have.

Quote:

---

Originally Posted by Cell

Hey guys,

I am trying to initialize a 2D array in C. For example, I have the matrix:

0 1 0 0
0 1 1 0
1 0 1 1
0 1 1 1

Now I think this would work:

Code:

---

int A[4][4];

A[0][0] = 0;
A[0][1] = 1;
A[0][2] = 0;
A[0][3] = 0;
A[0][4] = 0;
A[1][1] = 0;
A[1][2] = 1;
... ETC....
A[4][4] = 1;

---

But this is a real pain. I will have larger matrices like 17x17 and doing this would be a real annoyance.

Is there a way to do something like:

Code:

---

int  A[4][4] = {0 1 0 0; 0 1 1 0, ... etc}

---

?

Thanks!

---

for decently small arrays(4x4 or 5x5) like the ones you'hv shown
a[4][4]={{1,2,3,4},{3,4,5,6}....} is appropriate enough...

however when you go for huge arrays like the SBOX of Triple-DES or AES algorithm then you'd find it painful to show SBOXes in the same fashion as shown above...

You'd be much better off passing all of them as single dimentional array...
such as,
sbox_des[8][4][16]={14,4,13,1,2,15,11,8,3,10,6,12,5,9,0,7,0,15...... ...}. the compiler would automatically parse the array grouping first 64 elements as sbox[0] and then 4 sets of 16 elements...
just make sure that you enter the values row-wise , one row at a time, before going to the next row...and you'll be fine...

this approch is not recommended if someone else would want to examine the array, however it is acceptable if the array is of know constants, whose values are never to be manipulated...such as the SBOXes...

Quote:

---

Originally Posted by vart

more effitient in what way?

---

It's easier on the compiler to translate a pointer and offset compared to array indexing.

Quote:

---

Originally Posted by itCbitC

It's easier on the compiler to translate a pointer and offset compared to array indexing.

---

I do not think it necessary to facilitate the work of the compiler, if it makes the code less readable and gives no other benefits, such as performance

Quote:

---

Originally Posted by vart

I do not think it necessary to facilitate the work of the compiler, if it makes the code less readable and gives no other benefits, such as performance

---

That would be true only for the uninitiated, besides it's not the work of the compiler but the choice of the programmer.

Quote:

---

Originally Posted by itCbitC

That would be true only for the uninitiated, besides it's not the work of the compiler but the choice of the programmer.

---

In my opinion, a programmer who chooses to use pointer notation with an offset over the equivalent array notation to iterate over an array is uninitiated. It is a different story if a programmer chooses to use pointers over array indices to iterate over an array. (It is still true that a modern optimising compiler would optimise away the difference, but for the latter case it is the choice between an iterator and an index, not merely a choice between pointer notation and array index notation.)