Using a structure in a recursive function (referenced structure) - c

I'm having problems understanding how to write code that solves the following problem: I have a structure containing a 2D-array. Then I have a recursive function that take a pointer to the structure as an argument and I want the recursive function to be able to manipulate the structure sent, not a local copy.
The struct is initialized in the function initStruct, where memory for the 2D-array is allocated. The recursive function builds up an array and at a specific point calls a function to insert it into the structure's array.
The code:
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
int** spBasis(int);
void mpBasis(int**, int, int, int, int, int, int, int*, struct mpBasis *, int, int);
void initMpBasis(struct mpBasis *, int, int);
void insertMpState(struct mpBasis *, int *);
struct mpBasis {
int** basis;
int size;
int capacity;
};
int main() {
int a, b, c, d;
char maxE[256];
char noParticles[256];
char P[256];
char M[256];
FILE *fp;
int **spStates;
struct mpBasis *mp;
int mpState[6] = {0, 0, 0, 0, 0, 0};
printf("Input max e for sp states, no of particles, parity (1 for odd and 0 for even) and magnetic projection: ");
gets(maxE);
gets(noParticles);
gets(P);
gets(M);
spStates = spBasis(atoi(maxE));
fp = fopen("spStates.txt", "a+");
fprintf(fp, "E\tj\tl\tm\n");
for (a = 0; a < 330; a++) {
fprintf(fp, "State %d: ", a+1);
for (b = 0; b < 4; b++) {
fprintf(fp, "%d\t", spStates[a][b]);
}
fprintf(fp, "\n");
}
mp = malloc(sizeof(struct mpBasis));
initMpBasis(mp, 5449, 6);
for (c = 0; c < 5449; c++) {
for (d = 0; d < 6; d++) {
fprintf(fp, "%d: %d\t", c, mp->basis[c][d]);
}
fprintf(fp, "\n");
}
printf("%p\n", (void*) mp);
printf("hello 3");
mpBasis(spStates, 0, atoi(maxE), 0, atoi(M), 0, atoi(P), mpState, mp, 0, 0);
fclose(fp);
return 0;
}
int** spBasis(int maxE) {
int c;
int i, j, k, l;
int q = 0;
int** spStates;
spStates = (int**)malloc(330 * sizeof(int *));
for (c = 0; c < 330; c++) {
spStates[c] = malloc(4 * sizeof(int));
}
for (i = 0; i <= maxE; i++) {
for (j = i % 2; j <= i; j += 2) {
for (k = -(2 * j + 1); k <= (2 * j + 1); k += 2) {
spStates[q][0] = i;
spStates[q][1] = j;
spStates[q][2] = 2 * j + 1;
spStates[q][3] = k;
q += 1;
}
for (l = -(2 * j - 1); l <= (2 * j - 1); l += 2) {
spStates[q][0] = i;
spStates[q][1] = j;
spStates[q][2] = 2 * j - 1;
spStates[q][3] = l;
q += 1;
}
}
}
return spStates;
}
void mpBasis(int** spStates, int e, int maxE, int m, int M, int l,
int P, int * mpState, struct mpBasis *mpB, int position, int lastSpState) {
int i;
for (i = lastSpState; i < 330; i++) {
if (e > maxE) {
break;
} else if (position == 5) {
if (m == M && l % 2 == P) {
insertMpState(mpB, mpState);
break;
}
} else {
// add spState to mpState and make the recursive call for the next position
mpState[position] = i;
mpBasis(spStates, e + spStates[i][0], maxE, m + spStates[i][3], M,
l + spStates[i][1], P, mpState, mpB, position+1, i);
}
}
}
void initMpBasis(struct mpBasis *a, int initialSize, int sizeY) {
int c;
a->basis = (int **)malloc(initialSize * sizeof(int*));
for (c = 0; c < initialSize; c++) {
a->basis[c] = (int *) malloc(sizeY * sizeof(int));
}
a->size = 0;
a->capacity = initialSize;
}
void insertMpState(struct mpBasis *a, int* mpState) {
/*if (a->size == a->capacity) {
a->size *= 2;
a->basis = (int **)realloc(a->basis, a->size * sizeof(int));
}*/
a->basis[a->size++] = mpState;
}
Added all the code.
The problem is that after the recursive function has been called, the "basis" array in structure mpBasis still only contains random values, i.e. the mpBasis function hasn't done anything with it. Am I passing the mp argument by value here?
Thanks for your help!

The first step is to compile with warnings enabled. Eg if you are using GCC you can use option -Wall -Wextra.
EDIT:
(previous listing of >20 errors removed)
Ok, since you are using Visual Studio, enable warnings like this:
Open the project's Property Pages dialog box.
Select C/C++.
On the General property page, modify the Warning Level to /W4

Related

Pointer arrays allocated with malloc/calloc initializing with values other than 0

I have been given a school assignment in C to create a program that multiplies matrices. I will list assignment constraints below so people don't respond with questions as to why I am doing things this way.
Constraints from instructor:
Cannot use square brackets anywhere in code (use pointer notation instead)
Matrices A, B, C must be single integer pointer variables (int *A, *B, *C)
Can only use main function and those specified by header
Must compile with "gcc -ansi -Wall -o p2 p2.c"
I have not implemented the matrix multiplication function yet, as the issues I am having relate to either file reading or memory allocation.
The specific problem I am having is when I allocate space to the pointer matrix with either malloc OR calloc (tried both), the program inserts 33 in some places in the output instead of 0. I've tried everything at this point and am convinced my knowledge of pointers is fundamentally flawed.
p2.h (given by instructor)
#include <stdio.h>
#include <stdlib.h>
/* This function reads m, n, and p from the datafile.
It then allocates the correct amount of memory required for matrices
A, B, and C.
Then matrices A and B are filled from the datafile.
The values for m, n, and p are passed by reference, and are
thus filled in by this function
PARAMETERS in order are:
int ** matrix A
int ** matrix B
int ** matrix C
int * m The number of rows in matrix A
int * n The number of columns in matrix A and
The number of rows in matrix B
int * p The number of columns in matrix B
char * The name of the datafile, from the command line
*/
void read_matrices(int **, int **, int **, int *, int *, int *, char *);
/* This function prints a matrix. Rows and columns should be preserved.
PARAMETERS in order are:
int * The matrix to print
int The number of rows in the matrix
int The number of columns in the matrix
*/
void print_matrix(int *, int, int);
/* The two matrices A and B are multiplied, and matrix C contains the
result.
PARAMETERS in order are:
int * Matrix A
int * Matrix B
int * Matrix C
int m
int n
int p
*/
void mult_matrices(int *, int *, int *, int, int, int);
p2.c (sorry for the mess a lot of debugging went on)
#include <stdio.h>
#include <stdlib.h>
#include "./p2.h"
/* constants for testing */
#define cM 3
#define cN 2
#define cP 5
int main(int argc, char **argv) {
if (argc < 2) {
printf("Must include an argument.\n");
exit(1);
}
char *path = *(argv + 1);
int *m = (int *) malloc(sizeof(int));
int *n = (int *) malloc(sizeof(int));
int *p = (int *) malloc(sizeof(int));
*m = cM; *n = cN; *p = cP;
int i,j; /* loop counters */
/* allocate space for 2d pointer arrays */
int **A = NULL;
A = (int **) malloc(*m * sizeof(int *));
for (i = 0; i < *m; i++) {
*(A+i) = (int *) malloc(*n * sizeof(int));
}
int **B = NULL;
B = (int **) malloc(*n * sizeof(int *));
for (i = 0; i < *n; i++) {
*(B+i) = (int *) malloc(*p * sizeof(int));
}
int **C = NULL;
C = (int **) malloc(*m * sizeof(int *));
for (i = 0; i < *m; i++) {
*(C+i) = (int *) malloc(*p * sizeof(int));
}
/* write data to A */
for (i = 0; i < *m; i++) {
for (j = 0; j < *n; j++) {
*(*(A+i)+j) = 0;
}
}
/* testing a */
for (i = 0; i < *m; i++) {
for (j = 0; j < *n; j++) {
if (*(*(A+i)+j) != 0) {
printf("[x]");
} else {
printf("[0]");
}
}
}
printf("\n");
/* write data to B */
for (i = 0; i < *n; i++) {
for (j = 0; j < *p; j++) {
*(*(B+i)+j) = 0;
}
}
/* testing b */
for (i = 0; i < *n; i++) {
for (j = 0; j < *p; j++) {
if (*(*(B+i)+j) != 0) {
printf("[x]");
} else {
printf("[0]");
}
}
}
printf("\n");
/* write data to C */
for (i = 0; i < *m; i++) {
for (j = 0; j < *p; j++) {
*(*(C+i)+j) = 0;
}
}
/* testing c */
for (i = 0; i < *m; i++) {
for (j = 0; j < *p; j++) {
if (*(*(C+i)+j) != 0) {
printf("[x]");
} else {
printf("[0]");
}
}
}
printf("\n");
printf("Matrix A: \n");
print_matrix(*A, *m, *n);
printf("Matrix B: \n");
print_matrix(*B, *n, *p);
printf("Matrix C: \n");
print_matrix(*C, *m, *p);
return 0;
}
void read_matrices(int **A, int **B, int **C, int *m, int *n, int *p, char *path) {
FILE *fptr;
fptr = fopen(path, "r");
if (fptr == NULL) {
printf("Cannot open file: ./p2 [filename].txt\n");
exit(1);
}
/* get first 3 numbers from file, set m,n,p */
*m = fgetc(fptr);
fgetc(fptr);
*n = fgetc(fptr);
fgetc(fptr);
*p = fgetc(fptr);
fgetc(fptr);
/* read first matrix */
/* 1) calculate matrix size m x n
* 2) loop through malloc'ed matrix
* 3) each loop, insert char in loc
* 4) if next char NOT 10/32, add nextchar*10 to value in loc
*/
char cur;
while ( (cur = fgetc(fptr)) != EOF ) {
if (cur == 10 || cur == 32) {
/* do nothing :) */
} else {
*m = cur;
*n = cur;
*p = cur;
break;
}
}
printf("m: %c\n", *m);
printf("n: %c\n", *n);
printf("p: %c\n", *p);
printf("next: %c\n", fgetc(fptr));
fclose(fptr);
}
void print_matrix(int *X, int rows, int cols) {
int r, c;
int k = 0;
for (r = 0; r < rows; r++) {
for (c = 0; c < cols; c++) {
printf("\t%d", *(X+k));
k++;
}
printf("\n");
}
}
void mult_matrices(int *A, int *B, int *C, int m, int n, int p) {
}
d2.txt (data file)
3
2
4
1 2
3 4
5 6
7 8 9 10
11 12 13 14
Output: ./p2 d2.txt
[0][0][0][0][0][0]
[0][0][0][0][0][0][0][0][0][0]
[0][0][0][0][0][0][0][0][0][0][0][0][0][0][0]
Matrix A:
0 0
0 0
0 0
Matrix B:
0 0 0 0 0
0 33 0 0 0
Matrix C:
0 0 0 0 0
0 33 0 0 0
0 0 0 0 33
If you notice, I have some debug code that checks whether or not the current item in the array is 0. It seems to indicate that they are all 0, making me think it is a printing problem, but I am even more lost on what would be causing that. The ascii code for 33 is an exclamation point, but I am not sure what relevance it has.
Based on the function signatures you're supposed to use, you need to implement your 2D arrays as 1D with the correct index math. This will result in all memory being laid out contiguously, which is not at all guaranteed with the way you're allocating memory now (two calls to malloc for each matrix). For example:
#include <stdio.h>
#include <stdlib.h>
void print_matrix(int* A, int rows, int cols)
{
for (int r=0; r<rows; r++)
{
for (int c=0; c<cols; c++)
{
// If you want to treat A as a 2D matrix, this is where we have to do a bit of
// fancy index math to give you what double bracket notation [][] does for you
// r * cols gives you the index of the right row
// + c give you the column offset in that row
// add that offset to A then dereference
printf("%d\t", *(A + (r * cols + c)));
}
printf("\n");
}
}
int main(void)
{
// matrix A is supposed to be m by n
int* A;
// read these from file, or where ever they're supposed to come from
int m = 2;
int n = 10;
// Allocate the memory in one chunk. This makes the memory all contiguous, just the
// same as if you had done A[m][n]. However, the double call malloc for each int**
// matrix probably will not give you contiguous memory for the entire matrix. Each
// call to malloc is independent.
A = malloc(m * n * sizeof(int)); // or sizeof(*A) would be even better
if (A == NULL)
{
// handle error
}
// We can initialize values for A at this point, still not needing to care about
// rows or columns
for (int i=0; i<m*n; i++)
{
*(A + i) = i; // using i for a better visual when we print
}
print_matrix(A, m, n);
free(A);
return 0;
}
Demo
You are ovecomplicating simple things. Use pointers to arrays and allocate 2D array.
Use the correct type of your size variables.
Try to avoid side effects. Use parameters and function return values.
//this function is for the test purposes only
int writefile(const char *fn)
{
FILE *fo = fopen(fn, "w");
fprintf(fo,
"3\n"
"2\n"
"4\n"
"1 2\n"
"3 4\n"
"5 6\n"
"7 8 9 10\n"
"11 12 13 14\n");
fclose(fo);
}
void *allocIntMatrix(size_t rows, size_t cols)
{
int (*m)[cols] = malloc(rows * sizeof(*m));
return m;
}
void printIntMatrix(size_t rows, size_t cols, int (*m)[cols])
{
for(size_t row = 0; row < rows; row++)
{
for(size_t col = 0; col < cols; col++)
{
printf("[%5d] ", m[row][col]);
}
printf("\n");
}
}
int readData(FILE *fi, size_t rows, size_t cols, int (*m)[cols])
{
for(size_t row = 0; row < rows; row++)
{
for(size_t col = 0; col < cols; col++)
{
fscanf(fi, "%d", &m[row][col]);
}
}
return 0;
}
int main(int argc, char **argv)
{
size_t n,m,p;
writefile("a.aaa");
FILE *fi = fopen("a.aaa", "r");
fscanf(fi, "%zu", &m);
fscanf(fi, "%zu", &n);
fscanf(fi, "%zu", &p);
printf("n = %zu, m = %zu, p = %zu\n", n, m, p);
int (*A)[n] = allocIntMatrix(m, n);
int (*B)[p] = allocIntMatrix(n, p);
readData(fi, m, n, A);
readData(fi, n, p, B);
fclose(fi);
printIntMatrix(m, n, A);
printf("\n");
printIntMatrix(n, p, B);
return 0;
}
https://godbolt.org/z/adoEx1r4f
You need to check for errors (file, memory etc). I skipped it for the sake of simplicity of the example.

segmantation fault malloc pointers functions

hello guys this is my code :
#include <stdio.h>
#include <stdlib.h>
int power(int a, int b) {
int exponent = b, result = 1;
while (exponent != 0) {
result = result * a;
exponent--;
}
//printf("%d",result);
return result;
}
int fill_it(char ** p, int N, int fliptimes, int column2) {
if (N < 0) return 0;
int counter = 0, l;
char a = 'H';
for (l = 0; l < power(2, fliptimes); l++) {
p[l][column2] = a;
counter++;
if (counter == (power(2, N) / 2)) {
counter = 0;
if (a == 'H') a = 'T';
if (a == 'T') a = 'H';
}
}
fill_it(p, N--, fliptimes, column2++);
}
int main() {
int i, fores, j, l, m;
char ** p;
printf("how many times did you toss the coin?:");
scanf("%d", & fores);
p = (char ** ) malloc((power(2, fores)) * sizeof(char * ));
for (i = 0; i < fores; i++)
p[i] = (char * ) malloc(fores * sizeof(char));
fill_it(p, fores, fores, 0);
for (l = 0; l < power(2, fores); l++) {
for (m = 0; m < fores; m++) {
printf("%c", p[l][m]);
}
}
printf(",");
}
it does compile.But when i run the program it returns a "segmantation fault (core dumped)" error
i know it means that i tried to access memory,i dont have acces to but i dont understand which part of the program is defective
The problem is, you're not allocating enough memory. This line is fine
p = (char ** ) malloc((power(2, fores)) * sizeof(char * ));
but this loop is only allocating memory for part of the 2-dimensional array.
for (i = 0; i < fores; i++)
p[i] = (char * ) malloc(fores * sizeof(char));
The memory allocation should look more like this...
foresSquared = power(2, fores);
p = malloc(foresSquared*sizeof(char *));
for (i = 0; i < foresSquared; i++)
p[i] = malloc(fores);
Since the result of power is going to be consistent, it makes sense to store the value in a variable and use that rather than recalculating it. It'll make the code clearer too.
You also don't need to cast the return value of malloc as C handles that for you. And sizeof(char) isn't needed as it's guaranteed to always be 1.

How can I pass Pointer Matrix to Function in C?

I tried what is described in these links:
How can I pass character pointer reference to function and get affected value back?
Passing pointers (matrix) to a function in c
Passing an array as an argument to a function in C
but I can't. Dimension of my array is 3 and I initialized it like this:
char* trips[N][2] = {{"ANKARA", "10:00"}, {"BURSA", "11:00"}, {"IZMIR", "12:00"}, {"ISTANBUL", "13:00"}, {"ANTALYA", "14:00"}}
so how can I pass this matrix into function?
My code is here and it doesn't sort or effect my array.
#include <stdio.h>
#include <stdlib.h>
#define N 10
typedef enum {
SortByCity,
SortByHour
}SortType;
void merge(char**, int, int, int, SortType);
void mergeSort(char**, int, int, SortType);
void mergeSortByHour(char**, int, int, SortType);
int main(int argc, char *argv[]) {
int i;
char* trips[N][2] = {{"ANKARA", "10:00"}, {"BURSA", "11:00"}, {"IZMIR", "12:00"}, {"ISTANBUL", "13:00"}, {"ANTALYA", "14:00"}, {"ANKARA", "11:00"}, {"ISTANBUL", "13:00"}, {"ANKARA", "11:30"}, {"BURSA", "21:00"} , {"BURSA", "13:00"}};
mergeSort(**trips, 0, N-1, SortByCity);
for(i = 0; i < N; i++){
printf("%-10s %s\n", trips[i][0], trips[i][1]);
}
return 0;
}
void merge(char** T, int L, int M, int H, SortType S){
int i, j, k;
int N1 = M - L + 1;
int N2 = H - M;
char* LA[N1][2], RA[N2][2];
for (i = 0; i < N1; i++){
LA[i][S] = T[L+i][S];
LA[i][1-S] = T[L+i][1-S];
}
for (j = 0; j < N2; j++){
RA[j][S] = T[L+j][S];
RA[j][1-S] = T[L+j][1-S];
}
i = 0;
j = 0;
k = L;
while (i < N1 && j < N2)
{
if (strcmp(LA[i][S], RA[j][S]))
{
T[k][S] = RA[j][S];
T[k][1-S] = RA[j][1-S];
j++;
}
else
{
T[k][S] = LA[i][S];
T[k][1-S] = LA[i][1-S];
i++;
}
k++;
}
}
void mergeSort(char** T, int L, int H, SortType S){
if(L < H)
return;
int M = (L + H) / 2;
mergeSort(T, L, M, S);
mergeSort(T, M+1, H, S);
merge(T, L, M, H, S);
}
Please dont mark my question as duplicate because I dont understand the solutions explained in the site and can't solve my problem.

qsort structures on the basis of one element sorting [duplicate]

I'm not C expert and I've read through the forum, but I still need some advice regarding a sorting problem on C.
I have 4 dynamic arrays of doubles in C. All of them are the same size, and lets say n. What I want to do is to sort all of them using one of the arrays as first order and a second array as my second order. So if the arrays are *x, *y, *w and *z. I want to sort them according to the values of *x, then *y.
I must do this efficiently because the arrays are quite large.
Any help will be much appreciated.
The easy way to do this would be to map your four separate arrays onto a single array of a struct type like
struct rec {
double x;
double y;
double w;
double z;
};
struct rec *arr = malloc( sizeof *arr * N ); // where N is the number of
// elements in each array
if ( !arr )
// malloc failed, handle error somehow
for ( size_t i = 0; i < N; i++ )
{
arr[i].x = x[i];
arr[i].y = y[i];
arr[i].w = w[i];
arr[i].z = z[i];
}
and then create a comparison function to pass to qsort:
int cmpRec( const void *lhs, const void *rhs )
{
struct rec *l = lhs;
struct rec *r = rhs;
if ( l->x < r->x )
return -1;
else if ( l->x > r->x )
return 1;
else
{
if ( l->y < r->y )
return -1;
else if ( l->y > r->y )
return 1;
else
return 0;
}
return 0;
}
Now you can use the qsort library function to sort that array of struct:
qsort( arr, N, sizeof *arr, cmpRec );
Once that array is sorted, you can map the results back onto your four original arrays.
Clearly, sorting this using standard qsort() is not going to work; there isn't a mechanism for passing four arrays.
Equally clearly, if the data were structured as an array of structures, then using qsort() would be feasible.
Question 1: Is it feasible to create an array of structures, load it, sort it, and then unload back into the original arrays?
Question 2: Another option is to sort an array of integers:
int indexes[n];
for (int i = 0; i < n; i++)
indexes[i] = i;
qsort(indexes, n, sizeof(indexes[0]), comparator);
The comparator function would have to be able to access the x and y arrays as file scope variables:
int comparator(void const *v1, void const *v2)
{
int i1 = *(int *)v1;
int i2 = *(int *)v2;
extern double *x, *y;
if (x[i1] > x[i2])
return +1;
else if (x[i1] < x[i2])
return -1;
else if (y[i1] > y[i2])
return +1;
else if (y[i1] < y[i2])
return -1;
else
return 0;
}
You'd then be able to access the arrays using x[indexes[i]] etc to access the ith element in sorted order.
Is that acceptable?
If that is not convenient either, then you will end up writing your own sort; it isn't horribly painful, but will require some care.
I spent some time adapting an existing sort test framework to this scenario. The full code is quite large because it includes a lot of testing support code. The core function (compare, swap, partition and quicksort) are here (122 lines, including comment and blank lines):
/* SO 20271977 - sort arrays x, y, z, w (type double, size n) in parallel based on values in x and y */
/*
** To apply this to the real code, where there are 4 arrays to be sorted
** in parallel, you might write:
**
** Array4 a;
** a.x = x;
** a.y = y;
** a.z = z;
** a.w = w;
** a.n = n;
** quicksort_random(&a);
**
** Or even:
**
** quicksort_random((Array4){ .n = n, .x = x, .y = y, .z = z, .w = w });
**
** combining designated initializers and compound literals. Or you could write a
** trivial wrapper so that you can call:
**
** quicksort_random_wrapper(n, x, y, z, w);
*/
/* SOF so-20271977.h */
#include <stddef.h>
typedef struct Array4
{
size_t n;
double *x;
double *y;
double *z;
double *w;
} Array4;
extern void quicksort_random(Array4 *A);
/* EOF so-20271977.h */
#include <assert.h>
#include <stdlib.h> /* lrand48() */
/*
** Note that a more careful implementation would use nrand48() instead
** of lrand48() to prevent its random number generation from interfering
** with other uses of the x-rand48() functions.
*/
typedef size_t (*Part)(Array4 *A, size_t p, size_t r);
static void quicksort_partition(Array4 *A, size_t p, size_t r, Part partition);
static size_t partition_random(Array4 *A, size_t p, size_t r);
/* Quick Sort Wrapper function - specifying random partitioning */
void quicksort_random(Array4 *A)
{
quicksort_partition(A, 0, A->n - 1, partition_random);
}
/* Main Quick Sort function */
static void quicksort_partition(Array4 *A, size_t p, size_t r, Part partition)
{
if (p < r)
{
size_t q = (*partition)(A, p, r);
assert(p <= q && q <= r);
if (q > 0)
quicksort_partition(A, p, q-1, partition);
quicksort_partition(A, q+1, r, partition);
}
}
static inline int compare(Array4 const *A, size_t p, size_t r)
{
if (A->x[p] < A->x[r])
return -1;
else if (A->x[p] > A->x[r])
return +1;
if (A->y[p] < A->y[r])
return -1;
else if (A->y[p] > A->y[r])
return +1;
else
return 0;
}
static inline size_t random_int(size_t p, size_t r)
{
return(lrand48() % (r - p + 1) + p);
}
static inline void swap(Array4 *A, size_t i, size_t j)
{
double d;
d = A->x[i];
A->x[i] = A->x[j];
A->x[j] = d;
d = A->y[i];
A->y[i] = A->y[j];
A->y[j] = d;
d = A->z[i];
A->z[i] = A->z[j];
A->z[j] = d;
d = A->w[i];
A->w[i] = A->w[j];
A->w[j] = d;
}
static size_t partition_random(Array4 *A, size_t p, size_t r)
{
size_t pivot = random_int(p, r);
swap(A, pivot, r);
size_t i = p-1;
size_t j = p;
while (j <= r)
{
if (compare(A, j, r) <= 0)
swap(A, j, ++i);
j++;
}
return i;
}
The test framework (quite ridiculously elaborate if it weren't that I already had a variant of it on hand) is 369 lines including blank lines and comment lines — and all the code above:
#include <assert.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#define FLTFMT "%13.6f"
typedef struct Array4
{
size_t n;
double *x;
double *y;
double *z;
double *w;
} Array4;
static int trace = 0;
static void *xmalloc(size_t size)
{
void *space = malloc(size);
if (space == 0)
{
fprintf(stderr, "Out of memory (%zu)\n", size);
exit(1);
}
return space;
}
void quicksort_last(Array4 *A);
void quicksort_random(Array4 *A);
void selectionsort(Array4 *A);
static inline int compare(Array4 const *A, size_t p, size_t r)
{
if (A->x[p] < A->x[r])
return -1;
else if (A->x[p] > A->x[r])
return +1;
if (A->y[p] < A->y[r])
return -1;
else if (A->y[p] > A->y[r])
return +1;
else
return 0;
}
static void dump_array(char const *tag, Array4 const *A)
{
printf("%s [%zu..%zu]:\n", tag, (size_t)0, A->n-1);
for (size_t i = 0; i < A->n; i++)
printf("(" FLTFMT ", " FLTFMT ", " FLTFMT ", " FLTFMT ")\n",
A->x[i], A->y[i], A->z[i], A->w[i]);
}
static void chk_sort(Array4 const *A)
{
for (size_t i = 0; i < A->n - 1; i++)
{
//if (compare(A, i, i+1) > 0)
{
if (A->x[i] > A->x[i+1])
{
printf("Out of order: A.x[%zu] = " FLTFMT ", A.x[%zu] = " FLTFMT "\n",
i, A->x[i], i+1, A->x[i+1]);
}
else if ((A->x[i] == A->x[i+1] && A->y[i] > A->y[i+1]))
{
printf("Out of order: A.x[%zu] = " FLTFMT ", A.x[%zu] = " FLTFMT ", "
"A.y[%zu] = " FLTFMT ", A.y[%zu] = " FLTFMT "\n",
i, A->x[i], i+1, A->x[i+1], i, A->y[i], i+1, A->y[i+1]);
}
}
}
}
static inline void set(Array4 *A, size_t p, double d)
{
A->x[p] = d;
A->y[p] = d + drand48() - 0.5;
A->z[p] = d / 2.0;
A->w[p] = d * 2.0;
}
static void load_random(Array4 *A)
{
size_t size = A->n;
for (size_t i = 0; i < size; i++)
{
A->x[i] = drand48() * size;
A->y[i] = drand48() * size + drand48() - 0.5;
A->z[i] = drand48() * size / 2.0;
A->w[i] = drand48() * size * 2.0;
}
}
static void load_ascending(Array4 *A)
{
for (size_t i = 0; i < A->n; i++)
set(A, i, i);
}
static void load_descending(Array4 *A)
{
for (size_t i = 0; i < A->n; i++)
set(A, i, A->n - i);
}
static void load_uniform(Array4 *A)
{
for (size_t i = 0; i < A->n; i++)
set(A, i, A->n);
}
static void load_organpipe(Array4 *A)
{
for (size_t i = 0; i <= A->n / 2; i++)
set(A, i, i);
for (size_t i = A->n / 2 + 1; i < A->n; i++)
set(A, i, A->n - i);
}
static void load_invorganpipe(Array4 *A)
{
size_t range = A->n / 2;
for (size_t i = 0; i < A->n / 2; i++)
set(A, i, range - i);
for (size_t i = A->n / 2 + 1; i < A->n; i++)
set(A, i, i - range);
}
typedef void (*Load)(Array4 *A);
typedef void (*Sort)(Array4 *A);
typedef size_t (*Part)(Array4 *A, size_t p, size_t r);
static void test_one_sort(Array4 *A, Sort sort, char const *s_tag,
char const *l_tag, char const *z_tag)
{
if (trace)
{
printf("%s-%s-%s:", z_tag, l_tag, s_tag);
dump_array("Before", A);
}
clock_t start = clock();
(*sort)(A);
clock_t finish = clock();
double sec = (finish - start) / (double)CLOCKS_PER_SEC;
printf("%s-%s-%s: %13.6f\n", z_tag, l_tag, s_tag, sec);
chk_sort(A);
if (trace)
{
printf("%s-%s-%s:", z_tag, l_tag, s_tag);
dump_array("After", A);
}
fflush(stdout);
}
static Array4 *alloc_array(size_t size)
{
Array4 *A = xmalloc(sizeof(*A));
A->n = size;
A->x = xmalloc(size * sizeof(A->x[0]));
A->y = xmalloc(size * sizeof(A->y[0]));
A->z = xmalloc(size * sizeof(A->z[0]));
A->w = xmalloc(size * sizeof(A->w[0]));
return A;
}
static Array4 *dup_array(Array4 *A)
{
size_t size = A->n;
Array4 *B = alloc_array(size);
if (B != 0)
{
B->n = size;
memmove(B->x, A->x, size * sizeof(A->x[0]));
memmove(B->y, A->y, size * sizeof(A->y[0]));
memmove(B->z, A->z, size * sizeof(A->z[0]));
memmove(B->w, A->w, size * sizeof(A->w[0]));
}
return B;
}
static void free_array(Array4 *A)
{
free(A->x);
free(A->y);
free(A->z);
free(A->w);
free(A);
}
static void test_set_sorts(Array4 *A, char const *l_tag, char const *z_tag)
{
struct sorter
{
Sort function;
char const *tag;
} sort[] =
{
{ quicksort_last, "QS.L" },
{ quicksort_random, "QS.R" },
{ selectionsort, "SS.N" },
};
enum { NUM_SORTS = sizeof(sort) / sizeof(sort[0]) };
for (int i = 0; i < NUM_SORTS; i++)
{
Array4 *B = dup_array(A);
test_one_sort(B, sort[i].function, sort[i].tag, l_tag, z_tag);
free(B);
}
}
static void test_set_loads(size_t size, char const *z_tag)
{
struct loader
{
Load function;
char const *tag;
} load[] =
{
{ load_random, "R" },
{ load_ascending, "A" },
{ load_descending, "D" },
{ load_organpipe, "O" },
{ load_invorganpipe, "I" },
{ load_uniform, "U" },
};
enum { NUM_LOADS = sizeof(load) / sizeof(load[0]) };
Array4 *A = alloc_array(size);
for (int i = 0; i < NUM_LOADS; i++)
{
load[i].function(A);
test_set_sorts(A, load[i].tag, z_tag);
}
free_array(A);
}
/* Main Quick Sort function */
static void quicksort_partition(Array4 *A, size_t p, size_t r, Part partition)
{
if (p < r)
{
size_t q = (*partition)(A, p, r);
assert(p <= q && q <= r);
if (q > 0)
quicksort_partition(A, p, q-1, partition);
quicksort_partition(A, q+1, r, partition);
}
}
static size_t partition_random(Array4 *A, size_t p, size_t r);
static size_t partition_last(Array4 *A, size_t p, size_t r);
/* Quick Sort Wrapper function - specifying random partitioning */
void quicksort_random(Array4 *A)
{
quicksort_partition(A, 0, A->n - 1, partition_random);
}
/* Quick Sort Wrapper function - specifying partitioning about last element */
void quicksort_last(Array4 *A)
{
quicksort_partition(A, 0, A->n - 1, partition_last);
}
static inline size_t random_int(size_t p, size_t r)
{
return(lrand48() % (r - p + 1) + p);
}
static inline void swap(Array4 *A, size_t i, size_t j)
{
double d;
d = A->x[i];
A->x[i] = A->x[j];
A->x[j] = d;
d = A->y[i];
A->y[i] = A->y[j];
A->y[j] = d;
d = A->z[i];
A->z[i] = A->z[j];
A->z[j] = d;
d = A->w[i];
A->w[i] = A->w[j];
A->w[j] = d;
}
static size_t partition_random(Array4 *A, size_t p, size_t r)
{
size_t pivot = random_int(p, r);
swap(A, pivot, r);
size_t i = p-1;
size_t j = p;
while (j <= r)
{
if (compare(A, j, r) <= 0)
swap(A, j, ++i);
j++;
}
return i;
}
static size_t partition_last(Array4 *A, size_t p, size_t r)
{
size_t i = p-1;
size_t j = p;
while (j <= r)
{
if (compare(A, j, r) <= 0)
swap(A, j, ++i);
j++;
}
return i;
}
/* Selection Sort algorithm */
void selectionsort(Array4 *A)
{
size_t r = A->n;
for (size_t p = 0; p < r; p++)
{
for (size_t i = p; i < r; i++)
{
if (compare(A, p, i) > 0)
swap(A, p, i);
}
}
}
/*
** To apply this to the real code, where there are 4 arrays to be sorted
** in parallel, you might write:
**
** Array4 a;
** a.x = x;
** a.y = y;
** a.z = z;
** a.w = w;
** a.n = n;
** quicksort_random(&a);
**
** Or even:
**
** quicksort_random((Array4){ .n = n, .x = x, .y = y, .z = z, .w = w });
**
** combining designated initializers and compound literals. Or you could write a
** trivial wrapper so that you can call:
**
** quicksort_random_wrapper(n, x, y, z, w);
*/
int main(void)
{
srand48((long)time(0));
for (size_t i = 10; i <= 40; i += 10)
{
char buffer[10];
snprintf(buffer, sizeof(buffer), "%zuK", i);
test_set_loads(1000*i, buffer);
}
return 0;
}
If you can't use qsort with
typedef struct Point {
double x;
double y;
double w;
double z;
} Point;
Use qsort with
typedef struct UglyThing {
double x;
int i;
} UglyThing;
Create an array of size n, fill x with x values, i with index.
Call qsort. At the end, i will store the permutation order.
Swap the three other arrays according to the permutation order.
Then do the same with little arrays ("with same x") in the y direction.
If this ugly trick is not possible, then I don't see any other solution than reinventing the wheel.
(edit : I have just seen Andrew said something very close to this answer...sorry!)
Bye,
Francis

Array sorting in C

I'm not C expert and I've read through the forum, but I still need some advice regarding a sorting problem on C.
I have 4 dynamic arrays of doubles in C. All of them are the same size, and lets say n. What I want to do is to sort all of them using one of the arrays as first order and a second array as my second order. So if the arrays are *x, *y, *w and *z. I want to sort them according to the values of *x, then *y.
I must do this efficiently because the arrays are quite large.
Any help will be much appreciated.
The easy way to do this would be to map your four separate arrays onto a single array of a struct type like
struct rec {
double x;
double y;
double w;
double z;
};
struct rec *arr = malloc( sizeof *arr * N ); // where N is the number of
// elements in each array
if ( !arr )
// malloc failed, handle error somehow
for ( size_t i = 0; i < N; i++ )
{
arr[i].x = x[i];
arr[i].y = y[i];
arr[i].w = w[i];
arr[i].z = z[i];
}
and then create a comparison function to pass to qsort:
int cmpRec( const void *lhs, const void *rhs )
{
struct rec *l = lhs;
struct rec *r = rhs;
if ( l->x < r->x )
return -1;
else if ( l->x > r->x )
return 1;
else
{
if ( l->y < r->y )
return -1;
else if ( l->y > r->y )
return 1;
else
return 0;
}
return 0;
}
Now you can use the qsort library function to sort that array of struct:
qsort( arr, N, sizeof *arr, cmpRec );
Once that array is sorted, you can map the results back onto your four original arrays.
Clearly, sorting this using standard qsort() is not going to work; there isn't a mechanism for passing four arrays.
Equally clearly, if the data were structured as an array of structures, then using qsort() would be feasible.
Question 1: Is it feasible to create an array of structures, load it, sort it, and then unload back into the original arrays?
Question 2: Another option is to sort an array of integers:
int indexes[n];
for (int i = 0; i < n; i++)
indexes[i] = i;
qsort(indexes, n, sizeof(indexes[0]), comparator);
The comparator function would have to be able to access the x and y arrays as file scope variables:
int comparator(void const *v1, void const *v2)
{
int i1 = *(int *)v1;
int i2 = *(int *)v2;
extern double *x, *y;
if (x[i1] > x[i2])
return +1;
else if (x[i1] < x[i2])
return -1;
else if (y[i1] > y[i2])
return +1;
else if (y[i1] < y[i2])
return -1;
else
return 0;
}
You'd then be able to access the arrays using x[indexes[i]] etc to access the ith element in sorted order.
Is that acceptable?
If that is not convenient either, then you will end up writing your own sort; it isn't horribly painful, but will require some care.
I spent some time adapting an existing sort test framework to this scenario. The full code is quite large because it includes a lot of testing support code. The core function (compare, swap, partition and quicksort) are here (122 lines, including comment and blank lines):
/* SO 20271977 - sort arrays x, y, z, w (type double, size n) in parallel based on values in x and y */
/*
** To apply this to the real code, where there are 4 arrays to be sorted
** in parallel, you might write:
**
** Array4 a;
** a.x = x;
** a.y = y;
** a.z = z;
** a.w = w;
** a.n = n;
** quicksort_random(&a);
**
** Or even:
**
** quicksort_random((Array4){ .n = n, .x = x, .y = y, .z = z, .w = w });
**
** combining designated initializers and compound literals. Or you could write a
** trivial wrapper so that you can call:
**
** quicksort_random_wrapper(n, x, y, z, w);
*/
/* SOF so-20271977.h */
#include <stddef.h>
typedef struct Array4
{
size_t n;
double *x;
double *y;
double *z;
double *w;
} Array4;
extern void quicksort_random(Array4 *A);
/* EOF so-20271977.h */
#include <assert.h>
#include <stdlib.h> /* lrand48() */
/*
** Note that a more careful implementation would use nrand48() instead
** of lrand48() to prevent its random number generation from interfering
** with other uses of the x-rand48() functions.
*/
typedef size_t (*Part)(Array4 *A, size_t p, size_t r);
static void quicksort_partition(Array4 *A, size_t p, size_t r, Part partition);
static size_t partition_random(Array4 *A, size_t p, size_t r);
/* Quick Sort Wrapper function - specifying random partitioning */
void quicksort_random(Array4 *A)
{
quicksort_partition(A, 0, A->n - 1, partition_random);
}
/* Main Quick Sort function */
static void quicksort_partition(Array4 *A, size_t p, size_t r, Part partition)
{
if (p < r)
{
size_t q = (*partition)(A, p, r);
assert(p <= q && q <= r);
if (q > 0)
quicksort_partition(A, p, q-1, partition);
quicksort_partition(A, q+1, r, partition);
}
}
static inline int compare(Array4 const *A, size_t p, size_t r)
{
if (A->x[p] < A->x[r])
return -1;
else if (A->x[p] > A->x[r])
return +1;
if (A->y[p] < A->y[r])
return -1;
else if (A->y[p] > A->y[r])
return +1;
else
return 0;
}
static inline size_t random_int(size_t p, size_t r)
{
return(lrand48() % (r - p + 1) + p);
}
static inline void swap(Array4 *A, size_t i, size_t j)
{
double d;
d = A->x[i];
A->x[i] = A->x[j];
A->x[j] = d;
d = A->y[i];
A->y[i] = A->y[j];
A->y[j] = d;
d = A->z[i];
A->z[i] = A->z[j];
A->z[j] = d;
d = A->w[i];
A->w[i] = A->w[j];
A->w[j] = d;
}
static size_t partition_random(Array4 *A, size_t p, size_t r)
{
size_t pivot = random_int(p, r);
swap(A, pivot, r);
size_t i = p-1;
size_t j = p;
while (j <= r)
{
if (compare(A, j, r) <= 0)
swap(A, j, ++i);
j++;
}
return i;
}
The test framework (quite ridiculously elaborate if it weren't that I already had a variant of it on hand) is 369 lines including blank lines and comment lines — and all the code above:
#include <assert.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#define FLTFMT "%13.6f"
typedef struct Array4
{
size_t n;
double *x;
double *y;
double *z;
double *w;
} Array4;
static int trace = 0;
static void *xmalloc(size_t size)
{
void *space = malloc(size);
if (space == 0)
{
fprintf(stderr, "Out of memory (%zu)\n", size);
exit(1);
}
return space;
}
void quicksort_last(Array4 *A);
void quicksort_random(Array4 *A);
void selectionsort(Array4 *A);
static inline int compare(Array4 const *A, size_t p, size_t r)
{
if (A->x[p] < A->x[r])
return -1;
else if (A->x[p] > A->x[r])
return +1;
if (A->y[p] < A->y[r])
return -1;
else if (A->y[p] > A->y[r])
return +1;
else
return 0;
}
static void dump_array(char const *tag, Array4 const *A)
{
printf("%s [%zu..%zu]:\n", tag, (size_t)0, A->n-1);
for (size_t i = 0; i < A->n; i++)
printf("(" FLTFMT ", " FLTFMT ", " FLTFMT ", " FLTFMT ")\n",
A->x[i], A->y[i], A->z[i], A->w[i]);
}
static void chk_sort(Array4 const *A)
{
for (size_t i = 0; i < A->n - 1; i++)
{
//if (compare(A, i, i+1) > 0)
{
if (A->x[i] > A->x[i+1])
{
printf("Out of order: A.x[%zu] = " FLTFMT ", A.x[%zu] = " FLTFMT "\n",
i, A->x[i], i+1, A->x[i+1]);
}
else if ((A->x[i] == A->x[i+1] && A->y[i] > A->y[i+1]))
{
printf("Out of order: A.x[%zu] = " FLTFMT ", A.x[%zu] = " FLTFMT ", "
"A.y[%zu] = " FLTFMT ", A.y[%zu] = " FLTFMT "\n",
i, A->x[i], i+1, A->x[i+1], i, A->y[i], i+1, A->y[i+1]);
}
}
}
}
static inline void set(Array4 *A, size_t p, double d)
{
A->x[p] = d;
A->y[p] = d + drand48() - 0.5;
A->z[p] = d / 2.0;
A->w[p] = d * 2.0;
}
static void load_random(Array4 *A)
{
size_t size = A->n;
for (size_t i = 0; i < size; i++)
{
A->x[i] = drand48() * size;
A->y[i] = drand48() * size + drand48() - 0.5;
A->z[i] = drand48() * size / 2.0;
A->w[i] = drand48() * size * 2.0;
}
}
static void load_ascending(Array4 *A)
{
for (size_t i = 0; i < A->n; i++)
set(A, i, i);
}
static void load_descending(Array4 *A)
{
for (size_t i = 0; i < A->n; i++)
set(A, i, A->n - i);
}
static void load_uniform(Array4 *A)
{
for (size_t i = 0; i < A->n; i++)
set(A, i, A->n);
}
static void load_organpipe(Array4 *A)
{
for (size_t i = 0; i <= A->n / 2; i++)
set(A, i, i);
for (size_t i = A->n / 2 + 1; i < A->n; i++)
set(A, i, A->n - i);
}
static void load_invorganpipe(Array4 *A)
{
size_t range = A->n / 2;
for (size_t i = 0; i < A->n / 2; i++)
set(A, i, range - i);
for (size_t i = A->n / 2 + 1; i < A->n; i++)
set(A, i, i - range);
}
typedef void (*Load)(Array4 *A);
typedef void (*Sort)(Array4 *A);
typedef size_t (*Part)(Array4 *A, size_t p, size_t r);
static void test_one_sort(Array4 *A, Sort sort, char const *s_tag,
char const *l_tag, char const *z_tag)
{
if (trace)
{
printf("%s-%s-%s:", z_tag, l_tag, s_tag);
dump_array("Before", A);
}
clock_t start = clock();
(*sort)(A);
clock_t finish = clock();
double sec = (finish - start) / (double)CLOCKS_PER_SEC;
printf("%s-%s-%s: %13.6f\n", z_tag, l_tag, s_tag, sec);
chk_sort(A);
if (trace)
{
printf("%s-%s-%s:", z_tag, l_tag, s_tag);
dump_array("After", A);
}
fflush(stdout);
}
static Array4 *alloc_array(size_t size)
{
Array4 *A = xmalloc(sizeof(*A));
A->n = size;
A->x = xmalloc(size * sizeof(A->x[0]));
A->y = xmalloc(size * sizeof(A->y[0]));
A->z = xmalloc(size * sizeof(A->z[0]));
A->w = xmalloc(size * sizeof(A->w[0]));
return A;
}
static Array4 *dup_array(Array4 *A)
{
size_t size = A->n;
Array4 *B = alloc_array(size);
if (B != 0)
{
B->n = size;
memmove(B->x, A->x, size * sizeof(A->x[0]));
memmove(B->y, A->y, size * sizeof(A->y[0]));
memmove(B->z, A->z, size * sizeof(A->z[0]));
memmove(B->w, A->w, size * sizeof(A->w[0]));
}
return B;
}
static void free_array(Array4 *A)
{
free(A->x);
free(A->y);
free(A->z);
free(A->w);
free(A);
}
static void test_set_sorts(Array4 *A, char const *l_tag, char const *z_tag)
{
struct sorter
{
Sort function;
char const *tag;
} sort[] =
{
{ quicksort_last, "QS.L" },
{ quicksort_random, "QS.R" },
{ selectionsort, "SS.N" },
};
enum { NUM_SORTS = sizeof(sort) / sizeof(sort[0]) };
for (int i = 0; i < NUM_SORTS; i++)
{
Array4 *B = dup_array(A);
test_one_sort(B, sort[i].function, sort[i].tag, l_tag, z_tag);
free(B);
}
}
static void test_set_loads(size_t size, char const *z_tag)
{
struct loader
{
Load function;
char const *tag;
} load[] =
{
{ load_random, "R" },
{ load_ascending, "A" },
{ load_descending, "D" },
{ load_organpipe, "O" },
{ load_invorganpipe, "I" },
{ load_uniform, "U" },
};
enum { NUM_LOADS = sizeof(load) / sizeof(load[0]) };
Array4 *A = alloc_array(size);
for (int i = 0; i < NUM_LOADS; i++)
{
load[i].function(A);
test_set_sorts(A, load[i].tag, z_tag);
}
free_array(A);
}
/* Main Quick Sort function */
static void quicksort_partition(Array4 *A, size_t p, size_t r, Part partition)
{
if (p < r)
{
size_t q = (*partition)(A, p, r);
assert(p <= q && q <= r);
if (q > 0)
quicksort_partition(A, p, q-1, partition);
quicksort_partition(A, q+1, r, partition);
}
}
static size_t partition_random(Array4 *A, size_t p, size_t r);
static size_t partition_last(Array4 *A, size_t p, size_t r);
/* Quick Sort Wrapper function - specifying random partitioning */
void quicksort_random(Array4 *A)
{
quicksort_partition(A, 0, A->n - 1, partition_random);
}
/* Quick Sort Wrapper function - specifying partitioning about last element */
void quicksort_last(Array4 *A)
{
quicksort_partition(A, 0, A->n - 1, partition_last);
}
static inline size_t random_int(size_t p, size_t r)
{
return(lrand48() % (r - p + 1) + p);
}
static inline void swap(Array4 *A, size_t i, size_t j)
{
double d;
d = A->x[i];
A->x[i] = A->x[j];
A->x[j] = d;
d = A->y[i];
A->y[i] = A->y[j];
A->y[j] = d;
d = A->z[i];
A->z[i] = A->z[j];
A->z[j] = d;
d = A->w[i];
A->w[i] = A->w[j];
A->w[j] = d;
}
static size_t partition_random(Array4 *A, size_t p, size_t r)
{
size_t pivot = random_int(p, r);
swap(A, pivot, r);
size_t i = p-1;
size_t j = p;
while (j <= r)
{
if (compare(A, j, r) <= 0)
swap(A, j, ++i);
j++;
}
return i;
}
static size_t partition_last(Array4 *A, size_t p, size_t r)
{
size_t i = p-1;
size_t j = p;
while (j <= r)
{
if (compare(A, j, r) <= 0)
swap(A, j, ++i);
j++;
}
return i;
}
/* Selection Sort algorithm */
void selectionsort(Array4 *A)
{
size_t r = A->n;
for (size_t p = 0; p < r; p++)
{
for (size_t i = p; i < r; i++)
{
if (compare(A, p, i) > 0)
swap(A, p, i);
}
}
}
/*
** To apply this to the real code, where there are 4 arrays to be sorted
** in parallel, you might write:
**
** Array4 a;
** a.x = x;
** a.y = y;
** a.z = z;
** a.w = w;
** a.n = n;
** quicksort_random(&a);
**
** Or even:
**
** quicksort_random((Array4){ .n = n, .x = x, .y = y, .z = z, .w = w });
**
** combining designated initializers and compound literals. Or you could write a
** trivial wrapper so that you can call:
**
** quicksort_random_wrapper(n, x, y, z, w);
*/
int main(void)
{
srand48((long)time(0));
for (size_t i = 10; i <= 40; i += 10)
{
char buffer[10];
snprintf(buffer, sizeof(buffer), "%zuK", i);
test_set_loads(1000*i, buffer);
}
return 0;
}
If you can't use qsort with
typedef struct Point {
double x;
double y;
double w;
double z;
} Point;
Use qsort with
typedef struct UglyThing {
double x;
int i;
} UglyThing;
Create an array of size n, fill x with x values, i with index.
Call qsort. At the end, i will store the permutation order.
Swap the three other arrays according to the permutation order.
Then do the same with little arrays ("with same x") in the y direction.
If this ugly trick is not possible, then I don't see any other solution than reinventing the wheel.
(edit : I have just seen Andrew said something very close to this answer...sorry!)
Bye,
Francis

Resources