Detect outliers in a time series

The function fills in the existing column to hold outlier flags, and either overwrites the original file or outputs a data structure.

fdk_detect_outliers(df, iterations = 20, sigma = 2, plot = FALSE)

Arguments

df

dataframe with dates (date) and values (value) to smooth

iterations

number of iterations in order to detect outliers

sigma

number of deviations to exclude outliers at

plot

visualize the process, mostly for debugging (TRUE / FALSE = default)