Boxcar smooth and decimate a time series

Boxcar smooth and decimate a time series

smooth_decimate(x, filterlength, downsamplefactor = NULL, start = NULL,
  end = NULL, frequency = NULL, deltat = NULL, extend = FALSE,
  sides = 2)

Arguments

x

time series to smooth (ts or mts)
filterlength

Size of smoothing kernel (in points)
downsamplefactor

Factor to reduce number of points
start, end

Defined start and end to remove NAs after filtering (units of time)
frequency, deltat

New frequency can be specified by either (or both if they are consistent).
extend

Logical. If true, the start and end values are allowed to extend the series. If false, attempts to extend the series give a warning and are ignored.
sides

1=>convolution for past values, default 2=>centred on lag=0

Value

time series object with call attribute

Details

One of downsamplefactor, frequency or deltat must be specified. downsamplefactor will take priority when available.

See filter for details about sides argument. sides=1 (ie backwards) makes sense if you want to measure the start of a peak (latency). sides=2 (centred, the default) makes sense if you want to measure the mid-timepoint of the peak itself.

See also

window,filter

Examples

x=ts(rnorm(10000)+sin(1:10000/100),start=0,deltat=0.01)
# smoothed
xs=smooth_decimate(x,filterlength=100,downsamplefactor=10,start=1,end=99)
# smoothed with causal filter (past values only)
xsc=smooth_decimate(x,filterlength=100,downsamplefactor=10,start=1,end=99,sides=1)
plot(x)
lines(xs,col='magenta')
lines(xsc,col='green')