Re: [R] Fast linear convolution with R

2014-02-27 Thread Martin Maechler
> "MM" == Mitchell Maltenfort > on Thu, 27 Feb 2014 10:34:00 -0500 writes: MM> "Convolve" uses the FFT so probably expects powers of 2. Not quite (but in the correct direction): ?fft contains The FFT is fastest when the length of the series being transformed is highl

Re: [R] Fast linear convolution with R

2014-02-27 Thread Mitchell Maltenfort
"Convolve" uses the FFT so probably expects powers of 2. You might want to look at using "filter" Ersatzistician and Chutzpahthologist I can answer any question. "I don't know" is an answer. "I don't know yet" is a better answer. On Thu, Feb 27, 2014 at 5:31 AM, Phi