annotate man/man1/filt1.1 @ 0:5242703e91d3 tip

Initial checkin for AIM92 aimR8.2 (last updated May 1997).
author tomwalters
date Fri, 20 May 2011 15:19:45 +0100
parents
children
rev   line source
tomwalters@0 1 .TH FILT1 1 "1 September 1993"
tomwalters@0 2
tomwalters@0 3 .SH NAME
tomwalters@0 4 filt1 \- 1st order LP filter using an exponential smoother.
tomwalters@0 5
tomwalters@0 6 .SH SYNTAX
tomwalters@0 7 filt1 [options] [file]
tomwalters@0 8
tomwalters@0 9 .SH DESCRIPTION
tomwalters@0 10 A recursive "leaky integrator" 1st order low-pass filter which is a shifting
tomwalters@0 11 exponential window which decays into the past values.
tomwalters@0 12
tomwalters@0 13 .SH OPTIONS
tomwalters@0 14
tomwalters@0 15 1. tc
tomwalters@0 16
tomwalters@0 17 The time-constant parameter which controls the rate of decay and so the
tomwalters@0 18 size of the exponential window. For a given time constant of T seconds:
tomwalters@0 19
tomwalters@0 20 .nf
tomwalters@0 21 0.7T secs is the window half--life.
tomwalters@0 22 4.6T secs is 99% of the window's life.
tomwalters@0 23 .fi
tomwalters@0 24
tomwalters@0 25 When the time-constant parameter is given with time units (s or ms) then
tomwalters@0 26 it is converted to a decay constant using a = exp(-Ts/T),
tomwalters@0 27 where Ts is the sample interval in seconds from the given `samplerate' option,
tomwalters@0 28 and T is decay time-constant in seconds.
tomwalters@0 29 Otherwise the time-constant parameter is taken to be the decay constant directly.
tomwalters@0 30
tomwalters@0 31 With decay constant a=1 there is infinite memory because it does not decay.
tomwalters@0 32 With 0 < a < 1 there is a finite decaying memory.
tomwalters@0 33 The smaller `a', the smaller the memory (it decays faster).
tomwalters@0 34 With a = 0 the filter output is identical with its input.
tomwalters@0 35
tomwalters@0 36 2. de
tomwalters@0 37
tomwalters@0 38 Two optional difference equations are provided which repsectively compute
tomwalters@0 39 a recursive sum (when "de=sum") and mean (when "de=mean").
tomwalters@0 40 The n'th recursive update is respectively:
tomwalters@0 41
tomwalters@0 42 .nf
tomwalters@0 43 y[n] = a.y[n-1] + x[n] recursive sum
tomwalters@0 44 y[n] = a.y[n-1] + (1-a).x[n] recursive mean
tomwalters@0 45 .fi
tomwalters@0 46
tomwalters@0 47 where decay constant a = exp(-Ts/T),
tomwalters@0 48 Ts is the sample interval in seconds, and T is decay time-constant in seconds.
tomwalters@0 49
tomwalters@0 50 .SH "SEE ALSO"
tomwalters@0 51 options smooth
tomwalters@0 52
tomwalters@0 53 .SH COPYRIGHT
tomwalters@0 54 .LP
tomwalters@0 55 Copyright (c) Applied Psychology Unit, Medical Research Council, 1995
tomwalters@0 56 .LP
tomwalters@0 57 Permission to use, copy, modify, and distribute this software without fee
tomwalters@0 58 is hereby granted for research purposes, provided that this copyright
tomwalters@0 59 notice appears in all copies and in all supporting documentation, and that
tomwalters@0 60 the software is not redistributed for any fee (except for a nominal
tomwalters@0 61 shipping charge). Anyone wanting to incorporate all or part of this
tomwalters@0 62 software in a commercial product must obtain a license from the Medical
tomwalters@0 63 Research Council.
tomwalters@0 64 .LP
tomwalters@0 65 The MRC makes no representations about the suitability of this
tomwalters@0 66 software for any purpose. It is provided "as is" without express or
tomwalters@0 67 implied warranty.
tomwalters@0 68 .LP
tomwalters@0 69 THE MRC DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING
tomwalters@0 70 ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL
tomwalters@0 71 THE A.P.U. BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES
tomwalters@0 72 OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
tomwalters@0 73 WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,
tomwalters@0 74 ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
tomwalters@0 75 SOFTWARE.
tomwalters@0 76