How to adapt Matlab's polyfit and array indexing into c? - c

Question

I'm trying to convert this line of Matlab code into c programming language:
coefs = polyfit(X(50:(length(X)-50)),Y(50:(length(Y)-50)),1);
I've tried different implementations, but none of them seem to give me the same coefs(1) value, i.e, slope. All I'm interested in is in the slope.
Note: all the code is probably over-complicated. As of know, I don't know any better. Still new to this.
Since c doesn't have dynamic arrays, I have two linked lists, X and Y
UPDATE: I changed everything to *float instead of lists. Below is the functions' code and the declarations.
float getSumF(float * head, int size){
int i;
float sum=0;
for (i=0;i<size;i++){
sum = sum + head[i];
}
return sum;
}
float coef (float* xx, float* yy, int size) {
int i=0;
float sumX2 = 0;
float sumXY = 0;
float sumX = 0;
float sumY = 0;
for (i=0;i<size;i++){
sumX = sumX + xx[i];
sumY = sumY + yy[i];
sumX2 = sumX2 + xx[i]*xx[i];
sumXY = sumXY + xx[i]*yy[i];
}
float slope = (size*sumXY-sumX*sumY)/(size*sumX2-sumX*sumX);
return slope;
}
int main(void) {
setbuf(stdout, NULL);
int threshold=20;
int k, i, j;
float slope, angle;
int width, height, channels;
float* X = 0;
float* Y = 0;
int arraysizeX = 0;
float* osc = 0;
int arraysizeOsc = 0;
float* coord = 0;
int arraysizeCoord = 0;
unsigned char *img;
The way I achieve the array indexing of Matlab is through this:
for(i=0;i<1440;i++){
if (coord[i] > 0){
X = realloc (X,++arraysizeX*sizeof(*X));
Y = realloc (Y,arraysizeX*sizeof(*Y));
X[arraysizeX-1]=i+1;
Y[arraysizeX-1]=coord[i];
}
}
if (arraysizeX>102){
slope=coef(X+50,Y+50,arraysizeX-100);
}
else{
if(arraysizeX==0){
free(X);
free(Y);
free(coord);
printf("Threshold too low. Exiting.\n");
return 3;
}
slope = coef(X,Y,arraysizeX);
}
I also coded the same linear regression algorithm to matlab, here:
sizeX=length(X);
sumX2 = 0;
sumXY = 0;
for i=1:sizeX
sumX2 = sumX2 + X(i)*X(i);
sumXY = sumXY + X(i)*Y(i);
endfor
slope=(sizeX*sumXY-sum(X)*sum(Y))/(sizeX*sumX2-sum(X)*sum(X));
slope_polyfit = 7.4615e-3
slope_matlab = 5.8421e-3
slope_c = -0.175331
Now.. is my mistake in the coef call/function? Have I done everything right?
UPDATE: The solution I provided (with guidance from the comments) is functioning properly. The error was in the matlab code, in the obtention of X and Y arrays.

Answer 1

As updated in the post: the solution I provided (with some tweaks from the comments) is functioning properly. The error was in the matlab code, in the obtention of X and Y arrays.