error: expected expression before ‘{’ token

get error on c program..

Code:

---

void cal_fft(x,y,l,mode)
    float *x[256],*y;
    int l,mode;
{
  int np,lmx,lo,lix,lm,li,j1,j2,nv2,npm1,i,j,k;
  float scl,arg,c,s,t1,t2;
 
//  --x;
//  --y;
 x = {
455,699,831.5,357.7,-156.7,-238.8,-243.5,-412.3,-455.7,-347.7,-436.2,-597.1,-516.3,-488.6,-601.1,
-304.2,344.2,685.1,843.8,1014.5,722.2,60.5,-385.9,-822.1,-1358.5,-1381,-864.9,-435.4,-42.2,525.6,
724,582.5,670.4,766,506.1,432.8,567.8,261.1,-271,-632.5,-1229.6,-1800.1,-1677.5,-1329.8,-873.2,
82.6,671.5,859.6,1173.8,666.5,-59.3,-506.5,-1125.1,-1478.1,-1315.5,-1007.6,-678.5,-140.3,318.5,
503,729.5,754.2,422.2,191.9,-63.2,-346.2,-297,-6.3,657.1,1712,2352.2,2548.8,2644.3,2288.2,1723.3,
1619.6,1776.1,2122,3101,3739.4,4246.8,4780.5,4628.5,4651.2,4518.3,3444.9,-41.2,-7537.9,-11870.6,-9759.9,
-8684,-7629.3,-1460.7,3680.4,4703,7312.7,8529.8,3245.5,-1701.8,-2031.6,-2455.7,-4492.4,-4684.7,
-3794.4,-4865.1,-5996.2,-4903.6,-4551.7,-5013.1,-1892.3,3361.7,5620.1,6537.8,7446.2,4774.1,197.3,
-2400.5,-5046.5,-8156,-7590.2,-4451.3,-2433.4,-157.2,2490.4,2986.4,2425,3183.2,3229.6,2024,1947.7,
2225.8,697.1,-888,-1956.2,-4110.4,-5690.2,-5344.3,-4965.9,-3516.8,-337.3,1699.7,2876.6,3640.4,
2367.7,118.7,-1780.8,-3350.2,-4233.3,-3462.1,-2093.9,-614,862.7,1544.2,1581.2,1572.5,1027.2,249.9,
-261.2,-814.8,-1182.2,-912,-266.6,868.4,2467.6,3807.2,4182.5,3961.8,3271.8,2405,1657.9,1448.1,1578.7,
2175.9,2994.8,3653.2,4328,4616.3,4509.9,4513.1,3988.7,2708.3,-1204.8,-7215.8,-9026.1,-7237,-6753.7,
-5006.4,-379.3,2228,3011.5,4804.9,4647.9,1273.9,-856.5,-827.9,-1434,-2396.9,-2198.4,-2052.1,-2773.7,
-2883.5,-2293.6,-2303.4,-2033.6,-242.7,1535.7,2228.9,2786.6,2756.3,1407.3,49.8,-719.3,-1766.1,-2503.3,
-2056.6,-1330.8,-830.3,-62.7,551.1,537.4,545.2,837.4,720.1,499.5,615.6,501.4,112.9,-128.6,-388.4,
-835.9,-1084.5,-1204.6,-1189.6,-678.3,-22.5,422.5,763.6,746.3,433.1,82.4,-272.4,-532.6,-562.6,
-426.4,-214.7,0,132.2,157.1,158,134.3,86.4,49.7,0.1,-53,-82.1,-61.5,                };
  y=0;
 
  /* ====== radix-2 fft ====== */
 
  np=(int)pow(2.0,(float)l);
  lmx=np;
  scl=(float)(6.283185303/(float)np);
  for(lo=1;lo<=l;++lo){
    lix=lmx;
    lmx=lmx/2.0;
    arg=0.0;
    for(lm=1;lm<=lmx;++lm){
      c=cos(arg);
      s=mode*sin(arg);
      arg=arg+scl;
      for (li=lix; lix<0 ? li>=np : li<=np; li+=lix){
        j1=li-lix+lm;
        j2=j1+lmx;
//        t1=x[j1]-x[j2];
        t2=y[j1]-y[j2];
//        x[j1]=x[j1]+x[j2];
        y[j1]=y[j1]+y[j2];
//        x[j2]=c*t1+s*t2;
        y[j2]=c*t2-s*t1;
      }
    }
    scl=2.0*scl;
  }
 
  /* ====== bit reversal ====== */
 
 
  j=1;
  nv2=(int)(np/2.0);
  npm1=np-1;
  for(i=1;i<=npm1;++i){
    if(i>=j)
      goto L30;
//    t1=x[j];
    t2=y[j];
//    x[j]=x[i];
    y[j]=y[i];
//    x[i]=t1;
    y[i]=t2;
  L30:
    k=nv2;
   
  L40:
    if(k>=j)
      goto L50;
   
    j-=k;
    k=(int)(k/2.0);
    goto L40;
  L50:
    j=j+k;
  }
}

---

please help..

Initialise the array x before calling the function, then pass x to the function. Incidentally, if x (outside of the function) is an array of pointers to float, trying to initialise it with an array of double (or float) will still not work.

Quote:

---

Originally Posted by laserlight

Initialise the array x before calling the function, then pass x to the function. Incidentally, if x (outside of the function) is an array of pointers to float, trying to initialise it with an array of double (or float) will still not work.

---

I've change with ..

Code:

---

void cal_fft(float *x, float *y,int l,int mode)
//    float *x[],*y;
//    int l,mode;
{
  int np,lmx,lo,lix,lm,li,j1,j2,nv2,npm1,i,j,k;
  float scl,arg,c,s,t1,t2;
 
  --x;
  --y;
float x = {
455,699,831.5,357.7,-156.7,-238.8,-243.5,-412.3,-455.7,-347.7,-436.2,-597.1,-516.3,-488.6,-601.1,
-304.2,344.2,685.1,843.8,1014.5,722.2,60.5,-385.9,-822.1,-1358.5,-1381,-864.9,-435.4,-42.2,525.6,
724,582.5,670.4,766,506.1,432.8,567.8,261.1,-271,-632.5,-1229.6,-1800.1,-1677.5,-1329.8,-873.2,
82.6,671.5,859.6,1173.8,666.5,-59.3,-506.5,-1125.1,-1478.1,-1315.5,-1007.6,-678.5,-140.3,318.5,
503,729.5,754.2,422.2,191.9,-63.2,-346.2,-297,-6.3,657.1,1712,2352.2,2548.8,2644.3,2288.2,1723.3,
1619.6,1776.1,2122,3101,3739.4,4246.8,4780.5,4628.5,4651.2,4518.3,3444.9,-41.2,-7537.9,-11870.6,-9759.9,
-8684,-7629.3,-1460.7,3680.4,4703,7312.7,8529.8,3245.5,-1701.8,-2031.6,-2455.7,-4492.4,-4684.7,
-3794.4,-4865.1,-5996.2,-4903.6,-4551.7,-5013.1,-1892.3,3361.7,5620.1,6537.8,7446.2,4774.1,197.3,
-2400.5,-5046.5,-8156,-7590.2,-4451.3,-2433.4,-157.2,2490.4,2986.4,2425,3183.2,3229.6,2024,1947.7,
2225.8,697.1,-888,-1956.2,-4110.4,-5690.2,-5344.3,-4965.9,-3516.8,-337.3,1699.7,2876.6,3640.4,
2367.7,118.7,-1780.8,-3350.2,-4233.3,-3462.1,-2093.9,-614,862.7,1544.2,1581.2,1572.5,1027.2,249.9,
-261.2,-814.8,-1182.2,-912,-266.6,868.4,2467.6,3807.2,4182.5,3961.8,3271.8,2405,1657.9,1448.1,1578.7,
2175.9,2994.8,3653.2,4328,4616.3,4509.9,4513.1,3988.7,2708.3,-1204.8,-7215.8,-9026.1,-7237,-6753.7,
-5006.4,-379.3,2228,3011.5,4804.9,4647.9,1273.9,-856.5,-827.9,-1434,-2396.9,-2198.4,-2052.1,-2773.7,
-2883.5,-2293.6,-2303.4,-2033.6,-242.7,1535.7,2228.9,2786.6,2756.3,1407.3,49.8,-719.3,-1766.1,-2503.3,
-2056.6,-1330.8,-830.3,-62.7,551.1,537.4,545.2,837.4,720.1,499.5,615.6,501.4,112.9,-128.6,-388.4,
-835.9,-1084.5,-1204.6,-1189.6,-678.3,-22.5,422.5,763.6,746.3,433.1,82.4,-272.4,-532.6,-562.6,
-426.4,-214.7,0,132.2,157.1,158,134.3,86.4,49.7,0.1,-53,-82.1,-61.5
};
float  yx=0;
 
  /* ====== radix-2 fft ====== */
 
  np=(int)pow(2.0,(float)l);
  lmx=np;
  scl=(float)(6.283185303/(float)np);
  for(lo=1;lo<=l;++lo){
    lix=lmx;
    lmx=lmx/2.0;
    arg=0.0;
    for(lm=1;lm<=lmx;++lm){
      c=cos(arg);
      s=mode*sin(arg);
      arg=arg+scl;
      for (li=lix; lix<0 ? li>=np : li<=np; li+=lix){
        j1=li-lix+lm;
        j2=j1+lmx;
//        t1=x[j1]-x[j2];
//        t2=y[j1]-y[j2];
//        x[j1]=x[j1]+x[j2];
//        y[j1]=y[j1]+y[j2];
//        x[j2]=c*t1+s*t2;
//        y[j2]=c*t2-s*t1;
      }
    }
    scl=2.0*scl;
  }
 
  /* ====== bit reversal ====== */
 
 
  j=1;
  nv2=(int)(np/2.0);
  npm1=np-1;
  for(i=1;i<=npm1;++i){
    if(i>=j)
      goto L30;
//    t1=x[j];
//    t2=y[j];
//    x[j]=x[i];
//    y[j]=y[i];
//    x[i]=t1;
//    y[i]=t2;
  L30:
    k=nv2;
   
  L40:
    if(k>=j)
      goto L50;
   
    j-=k;
    k=(int)(k/2.0);
    goto L40;
  L50:
    j=j+k;
  }
}

---

I got alot warning such as

Code:

---

fourier-asli.c:87: warning: excess elements in scalar initializer
fourier-asli.c:87: warning: (near initialization for ‘x’)

---

It's so many warning on code float x. I think if warning it's no problem. But the output is none, I think I've some error because so many warning I can't scroll again..

help !!

If you declare "float x = " you are declaring a new variable that has nothing to do with the array of floats that were passed in to your function (which is therefore no longer accessible to you). It is not possible to use initializer notation to set values of a passed-in array.

x and y are pointers. Are you sure the --x, --y is what you want? You are changing an address which is a bad idea.
As tabstop said this kind of initialization won't work. But let me ask you this. Why would you want to initialize a string like that inside the function. It is something fixed. Which means that the array x will be intialized every time with the same values. Which also means that x must have a specific size. Which means that x is actually something standard, something constant.
So what you should do is this. You should declare a const float x[] and initialize it as you did. Then just copy if you want the values of x to another float xx[] (not constant) and pass it to the function and do your job.

My point is this:
1) You won't have to rewrite the whole initialization. The other way is to write x[0] = 43242, x[1] = 3213432 etc etc. But it not that nice as a code. And you don't really get a better performance
2) If you have something constant is better to declare it with the ease of the initialization C offers you. Then you can do whatever you want with that data.
3) A good idea is to give as a paramter floax x[number_of_elements_initialized] since you want to initialize an array of a specific number of elements, thus the compiler will warn you of any possible error. Giving float *x as a parameter means that any array of float can be passed

Quote:

---

Originally Posted by tabstop

If you declare "float x = " you are declaring a new variable that has nothing to do with the array of floats that were passed in to your function (which is therefore no longer accessible to you). It is not possible to use initializer notation to set values of a passed-in array.

---

Quote:

---

It is not possible to use initializer notation to set values of a passed-in array

---

Did you mean like

Code:

---

x[256] = { ... }

---

I've try but same with topic error: expected expression before ‘{’ token

Quote:

---

Originally Posted by C_ntua

3) A good idea is to give as a paramter floax x[number_of_elements_initialized] since you want to initialize an array of a specific number of elements, thus the compiler will warn you of any possible error. Giving float *x as a parameter means that any array of float can be passed

---

Actually, the number of elements in brackets is ignored when part of a parameter. An array is converted to a pointer to its first element when passed as an argument, so it really is just a pointer.

Quote:

---

Originally Posted by nendariani

I've try but same with topic error: expected expression before ‘{’ token

---

The initializer notation is the { ... } part. By the time x is in the function, it has been initialised, so you simply cannot initialise it any more. You can only assign to its elements.

I've tried

Code:

---

float x[256] = ...

---

and successful
thank you all