filter (var* Data, int Length, var Kernel[6]) : var*
Applies a 5-element convolution filter
with the given Kernel to the Data array. Each
Data element is added to its local neighbors, weighted by the
Kernel, using this formula: Data[i] =
Kernel[0]*Data[i-2] + Kernel[1]*Data[i-1] + Kernel[2]*Data[i] +
Kernel[3]*Data[i+1] + Kernel[4]*Data[i+2] + Kernel[5]. This function
can be used for data smoothing, summing up, subtraction, sharpening, edge
enhancement, or similar data array manipulation operations.
Parameters:
| Data |
Array or series to be filtered. |
| Length |
Number of elements to be filtered. |
| Kernel |
Vector of 5 weights plus constant to be
added. |
Returns
Modified Data
renorm (var* Data, int Length, var Sum) : var*
Modifies the Data array by multiplying all elements with a
factor so that they sum up to
Sum.
This function can be used to normalize a list of weights to a certain total.
Parameters:
| Data |
Array or series to be normalized. |
| Length |
Number of elements to be normalized. |
| Sum |
Resulting sum of elements. |
Returns
Modified Data
Remarks:
- Use the renorm function on the kernel
when the weight sum must be 1.
- For boundary handling, Data[0] and
Data[Length-1] are extended.
- Some examples of filter kernels:
| Identity |
{ 0, 0, 1, 0, 0, 0 } |
| Sharpen |
{ -1, -1, 4, -1, -1, 0 } |
| Smooth |
{ 0.2, 0.2, 0.2, 0.2, 0.2, 0 } |
| Gaussian |
{ 0.1, 0.2, 0.4, 0.2, 0.1, 0 } |
| Shift to the right |
{ 0, 1, 0, 0, 0, 0 } |
| Mean subtraction |
{ 0, 0, 1, 0, 0, -mean } |
| Fill with constant |
{ 0, 0, 0, 0, 0, constant } |
Example:
var* Filter = { 1,2,3,2,1,0 };
filter(Data,Length,renorm(Filter,5,1));
See also:
predict, advise, polyfit,
distribute
► latest
version online