spiro_smooth: Apply a smoothing filter to data from cardiopulmonary...

In spiro: Manage Data from Cardiopulmonary Exercise Testing

Apply a smoothing filter to data from cardiopulmonary exercise testing.

Description

Filter vectors and data frames with moving averages and digital filters. Provides the data filtering for spiro_max and spiro_plot.

Usage

spiro_smooth(data, smooth = 30, columns = NULL, quiet = FALSE)

Arguments

data	A data frame of the class spiro. Usually the output of the spiro function.
smooth	An integer or character string specifying the smoothing strategy and parameters. Default is 30, which means the applied filter is a 30-second moving average. See the section 'Filtering Methods' for more details.
columns	A character vector of the data columns that should be filtered. By default the filtering applies to all data column of data (besides load, time and step).
quiet	Whether warning message should be suppressed. Default is FALSE.

Details

Raw data from cardiopulmonary is usually noisy due to measurement error and biological breath-to-breath variability. When processing or visualizing the gas exchange data, it is often helpful to filter the raw data. This function provides different filtering methods (time average, breath average, digital filters).

Breath-based and digital filters will be applied on the raw breath-by-breath data. Time-based averages will be used on the interpolated data.

Value

A data frame

Filtering Methods

Time-Based Average (e.g. smooth = 30)

A (centered) moving average over a defined time span. The number can be given as an integer or as a character (e.g. smooth = "30") and defines the length of the calculation interval in seconds.

Breath-Based Average (e.g. smooth = "15b")

A (centered) moving average over a defined number of breaths. The integer before the letter 'b' defines the number of breaths for the calculation interval.

Butterworth filter (e.g. smooth = "0.04f3")

A digital Butterworth filter (with lag). The number before the letter 'f' defines the low-pass cut-off frequency, the number after the letter 'f' gives the order of the filter. See bw_filter for more details.

Zero-lag Butterworth filter (e.g. smooth = "0.04fz3")

A digital forwards-backwards Butterworth filter (without lag). The number before the letter 'f' defines the low-pass cut-off frequency, the number after gives the order of the filter. See bw_filter for more details.

Examples

# Get example data
file <- spiro_example("zan_gxt")
d <- spiro(file)

out <- spiro_smooth(d, 30)
head(out)

# filter only the VO2 column with a zero-phase Butterworth filter
out2 <- spiro_smooth(d, "0.04fz3", columns = "VO2")
head(out2)

spiro documentation built on Aug. 14, 2023, 5:07 p.m.