find_Outliers: Function to detect outliers in the roughness data

In tzerk/roughness: A Collection of Functions Related to Soil Surface Roughness

Description

This function applies three different methods to detect outliers.

Usage

find_Outliers(data, prefilter = TRUE, method = c("iqr", "sd", "mad"),
  remove = "mad", window = FALSE, width = 20, step = 10,
  hist = TRUE, plot = TRUE, plot.single = FALSE, ...)

Arguments

data	data.frame (required): A log file imported with import_Data.
prefilter	logical (optional): TRUE to remove values that are +- 10 cm of the median
method	character (optional): Vector of methods applied to detect outliers. Currently implemented: "iqr" for Interquartile Range (see IQR), "sd" for Standard Deviation or "mad" for Median Absolut Deviation.
remove	character (optional): Returns the original data set with ourliers removed by the specified method (same options as for argument method).
window	logical (optional): If TRUE the data set is split in chunks of size specified by width before applying any of the detection methods.
width	integer (optional): Width of moving window.
step	integer (optional): Step size of moving window.
hist	logical (optional): Should histograms be plotted?
plot	logical (optional): Should the results be plotted?
plot.single	logical (optional): Should the plots be combined or individually presented?
...	Currently unused.

Value

The data set with outliers removed is returned. Optionally, a plot is created

Examples

# Load example data
file <- system.file("extdata/raw.log", package = "roughness")
data <- import_Data(file)

# find outliers
data_cleaned <- find_Outliers(data = data,
                              prefilter = TRUE,
                              method = c("iqr", "sd", "mad"),
                              remove = "mad",
                              window = FALSE,
                              width = 20,
                              step = 10, 
                              hist = TRUE,
                              plot = TRUE)

head(data_cleaned)

# find outliers with window technique
data_cleaned2 <- find_Outliers(data = data,
                               prefilter = TRUE,
                               method = "mad",
                               remove = "mad",
                               window = FALSE,
                               width = 30,
                               step = 15, 
                               hist = TRUE,
                               plot = TRUE)
                               
head(data_cleaned2)

tzerk/roughness documentation built on May 3, 2019, 2:05 p.m.