robust_loess_fit: Fit a robust loess regression

In nategarton13/bulletcp: Automatic Groove Identification via Bayesian Changepoint Detection

Description

Internal function called by get_grooves_lassobasic and get_grooves_lassofull

Usage

robust_loess_fit(cc, iter)

Arguments

cc	data frame with columns x and value_std, representing the crosscut
iter	number of iterations

Examples

data("example_data")
head(example_data)
example_data <- example_data[seq(from = 1, to = nrow(example_data), by = 30),]
plot(example_data$x, example_data$y)

# set the minimum y-value to zero
check_min <- min(example_data$value[!is.na(example_data$value)])
example_data <- dplyr::mutate(example_data, value_std = value - check_min)

# remove global structure
rlo_fit <- robust_loess_fit(cc = example_data, iter = 20)
example_data$rlo_pred <- predict(rlo_fit, newdata = example_data)
example_data$rlo_resid <- example_data$value_std - example_data$rlo_pred

# define new data frame without the global structure
data <- data.frame("x" = example_data$x, "y" = example_data$rlo_resid)
plot(data$x, data$y)

nategarton13/bulletcp documentation built on Feb. 1, 2020, 10:03 a.m.