| Rodionov {rshift} | R Documentation |
Rodionov (2004)'s STARS algorithm
Description
performs STARS analysis (Rodionov, 2004) on a dataset
Usage
Rodionov(data, col, time, l, prob = 0.05, startrow = 1, merge = FALSE)
Arguments
data |
The dataframe to be used. |
col |
The column we are measuring change on. |
time |
The column containing time units (e.g. age of a subsample) |
l |
The cut-off length of a regime; affects sensitivity (see Rodionov, 2004) |
prob |
The p-value for significance of a regime shift. Defaults to p = 0.05. |
startrow |
What row the analysis starts at. Defaults to 1. |
merge |
Sets the result to be either a regime-shift only table (if FALSE), or an addition to the original table (if TRUE) |
Value
If merge = FALSE (default), produces a 2-column table of time (the time value for each regime shift) and RSI (the regime shift index for each regime shift). If merge = TRUE, returns the original dataset with an extra RSI column, giving the regime shift index for each time unit - 0 for non-shift years.
Examples
Rodionov(lake_data, "DCA1", "Age", l=5)
Rodionov(lake_data, "DCA1", "Age", l=5, prob=0.01, startrow=2, merge=TRUE)