Skip to contents

Apply smoothing functions to data

Source: R/smooth_data.R
apply_smoothing.Rd

Applies the selected smoothing method to the specified data range.

Usage

apply_smoothing(data, y_var, method, index = NULL, k = 7)

Arguments

data

A data.frame with the data to smooth (must have the columns Index and y_var)

y_var

Character specifying the variable to apply smoothing to.

method

Character specifying the method to use for smoothing. Options include "rollmean", "rollmedian", "savgol", and "kalman". See details for details on the different methods.

index

Index values of the rows that need to be smoothed

k

A adjustment parameter for the selected method (see details).

Value

data with the values within range replaced with the smoothed values

Details

The following smoothing methods are currently supported:

  • rollmean: Based on the rollmean function from the zoo package. Replaces data with a rolling mean value. The k parameter is used to control the number of points to include in the mean calculation.

  • rollmedian: Based on the rollmedian function from the zoo package. Replaces data with a rolling median value. The k parameter is used to control the number of points to include in the median calculation, should be odd but will convert to an odd number with a warning.

  • savgol: Based on the savgol function from the pracma package. Applies a Savitzky–Golay filter which fits a fourth order polynomial to a sliding range of data. The k parameter is used controls the number of points in the window.

  • savgol: Based on the dlmSmooth function from the dlm package. Applies a Kalman filter which process model to the data. The k parameter is used controls the amount of smoothing.

Examples

smoothed <- apply_smoothing(example_data, "Temp_C", "rollmean", k=100)
ggplot2::ggplot(example_data, ggplot2::aes(x=DateTime_rd, y=Temp_C)) +
ggplot2::geom_line(color="black", na.rm=TRUE) +
ggplot2::geom_line(data=smoothed, color="darkred", na.rm=TRUE)