Re: [R] smooth function advice

2008-10-31 Thread Bert Gunter
10:04 AM To: [EMAIL PROTECTED] Cc: r-help@r-project.org Subject: Re: [R] smooth function advice Thanks to Mark Leeds for an implementation of exponentially weighted moving averages as follows which solves this problem. This also resmooths recursively. ewma<-function(x,lambda = .5, init = (1-l

Re: [R] smooth function advice

2008-10-31 Thread tolga . i . uzuner
Thanks to Mark Leeds for an implementation of exponentially weighted moving averages as follows which solves this problem. This also resmooths recursively. ewma<-function(x,lambda = .5, init = (1-lambda)*.raw[good.ind][1],order=1) { .raw <- unclass(coredata(x)) good.ind <- !is.na(.raw) # dete

Re: [R] smooth function advice

2008-10-30 Thread Dieter Menne
jpmorgan.com> writes: > I am looking for a smoothing function with the following characteristics > for a time series of data: > - at each date, should only use data up to that date (so, right aligned > and not centered) > - should return a smoothed series of length equal to the original time

[R] smooth function advice

2008-10-30 Thread tolga . i . uzuner
Dear R Users, I am looking for a smoothing function with the following characteristics for a time series of data: - at each date, should only use data up to that date (so, right aligned and not centered) - should return a smoothed series of length equal to the original time series: - fo