fullfilter: Filter functions for 'NetCDF' objects

In jonasbhend/geoutils: package to read in and analyse climate model output in NetCDF format

Description

fullfilter applies a moving average filter to the full time series (no missing values, decreasing filter length at start and end). applyfilter applies fullfilter to the time dimension of an object of class 'NetCDF'. make_filter helps you set up a low- band- or highpass filter with the desired response specified in desired_response. get_filter_weights provides the weights given a cutoff frequency. get_response can be used to illustrate and analyse the filter response.

Usage

fullfilter(x, h)

applyfilter(x, h)

get_response(h, accuracy = 1e-05)

desired_response(fr, type = "lowpass", accuracy = 1e-05)

get_filter_weights(n, fr, accuracy = 1e-05)

make_filter(n, fr, type = "lowpass", ...)

filter9p(x)

smoother(x, df = 5)

Arguments

x	time series or 'NetCDF' object
h	filter weights
accuracy	numerical accuracy to which the response should be computed
fr	cutoff frequencies
type	one of lowpass, bandpass, or highpass
n	number of time steps used in digital filter
...	additional arguments passed to get_filter_weights
df	degrees of freedom for spline smoothing

Examples

xx <- rnorm(100) + sin(seq(0,4*pi, length=100))
plot(xx, type='l')
hh <- make_filter(31, 10, type='lowpass')
lines(filter(xx, hh), lwd=2)
plot(get_response(hh), type='l')
abline(v=1/10, lty=2, lwd=2)

jonasbhend/geoutils documentation built on May 19, 2019, 7:27 p.m.