iFit/conv

 c = conv(a,b) : computes the convolution of iFunc models

   @iFunc/conv function to compute the convolution of models (FFT based).
     A deconvolution mode is also possible.
     When used with a single scalar value, it is used as a width to build a 
       gaussian function.
     when one of the argument is a character string, it is used as-is in the 
     operator expression. 

     conv(a, 'tas') convolves the 'a' model with the TAS resolution function
       computed from ResLibCal. The dispersion is either a 2D model S(|q|,w)
       or a 4D model S(qh,qk,ql,w). The returned model can then further be used
       to fit its parameters to match a measurement scan.

 input:  a: object or array (iFunc or numeric)
         b: object or array (iFunc or numeric)
     shape: optional shape of the return value
          full         Returns the full two-dimensional convolution.
          same         Returns the central part of the convolution of the same size as a.
          valid        Returns only those parts of the convolution that are computed
                       without the zero-padded edges. Using this option, y has size
                       [ma-mb+1,na-nb+1] when all(size(a) >= size(b)).
          deconv       Performs an FFT deconvolution.
          pad          Pads the 'a' signal by replicating its starting/ending values
                       in order to minimize the convolution side effects
          center       Centers the 'b' filter so that convolution does not shift
                       the 'a' signal.
          normalize    Normalizes the 'b' filter so that the convolution does not
                       change the 'a' signal integral.
          background   Remove the background from the filter 'b' (subtracts the minimal value)
     Default shape is 'same'

 output: c: object or array (iFunc)
 ex:     c=conv(a,b); c=conv(a,b, 'same pad background center normalize');

 Version: Aug. 22, 2017
 See also iFunc, iFunc/convn, iFunc/xcorr, fconv, fconvn, fxcorr, ResLibCal