Previous: expfit
Up: References
Next: fiteval
Previous Page: expfit
Next Page: fiteval

filter

htmlonly

• name:
  filter
• synopsis:

  filter inputfile outputfile [-format=" format-string"] [-low_pass= start,end] [-high_pass= start,end] [-threshold= amplitude] [-dc_block] [-pad_with_zeroes] [-window] [-zero_pad_spectrum= length-multiplier] [-single_precision] [-derivative]

• description:
  Performs digital filtering of a data set by doing an FFT, applying the filter, then doing an inverse FFT. Supports low-pass and high-pass filters with linear roll-off, as well as Parzen windowing[4] of the data.
• files:
  inputfile contains the data to be filtered, usually consisting of values at equispaced abscissae. The result is placed in outputfile.
• options:

  format

  
  Specifies a C-style format string for printing the data to the output file.

  low_pass, high_pass

  
  These two options allow specification of idealized low- and high-pass filters to be applied to the data. The filters are implemented by multiplying the real and imaginary components of the FFT by a factor between 1 and 0. For low-pass (high-pass), all components below start are multiplied by 1 (0), all components above end are multiplied by 0 (1), and components between start and end are multiplied by a linearly decreasing function that begins with value 1 (0) and ends with value 0 (1).

  threshold

  
  Specifies a threshold magnitude below which all frequency components will be set to zero.

  dc_block

  
  Requests that the DC term be filtered out.

  pad_with_zeroes

  
  Requests zero-padding of the data set to the nearest power of 2, in the event that the number of data points is not a power of 2. The default behavior is truncation to the greatest power of 2 less than the number of points.

  zero_pad_spectrum

  
  Requests zero-padding of the frequency spectrum to give a spectrum that is length-multiplier times longer, where length-multiplier is a power of 2. This has the effect of interpolating at length-multiplier points between each original data point.

  single_precision

  
  Requests that operations be performed using single-precision arithmetic.

  derivative

  
  Requests that, instead of doing digital filtering, the program should use FFTs to take the numerical derivative of the input data set. This option is incompatible with any of the filtering options.

• see also:

  
  fft, interp, deriv.