calc_rolling_stats: Calculate rolling window statistics

In cbuelo/tvsews: Time vs. Space Early Warning Statistics

Description

Calculate rolling window statistics for multiple combinations of statistics, variables, rolling window widths, and detrending options.

Usage

calc_rolling_stats(
  data,
  var_cols,
  id_cols = c("Lake", "Year"),
  time_col = "DOY",
  statistics = c("Mean", "SD", "SD_sd", "Ar1", "Ar1_sd"),
  vars_to_log = NULL,
  log_offset = 0.1,
  widths = c(21),
  detrend = FALSE,
  min_prop = 0.99
)

Arguments

data	data frame containing time series to calculate rolling window stats on
var_cols	character vector, variable names that are columns of *data* containing time series to calculate rolling window stats on
id_cols	character vector, two columns specifying "groups" to separate data into before calculating stats independently on, defaults are "Lake" and "Year"
time_col	character, column name that defines time step of data within the *id_cols* groupings
statistics	character vector, rolling window statistics to calculate, defaults to all available: c("Mean", "SD", "SD_sd", "Ar1", "Ar1_sd")
vars_to_log	character vector, optional variables names to log10 transform and create new columns before calculating rolling window stats
log_offset	positive numeric value, if any variables are being log transformed that contain negative values, all values will be shifted so that the minimum (before transformation) equals this value
widths	numeric vector, rolling window width(s) to be used, default is c(21)
detrend	TRUE or FALSE, should data within rolling windows be linearly detrended before calculating *statistics*
min_prop	numeric value between 0 and 1, proportion of data within a rolling window to calculate rolling window statistic, default 0.99

Value

a data frame with the results for the specified stats, widths, variables, etc.

Examples

rw_stats <- calc_rolling_stats(data = ts_data, var_cols = c("DO_Sat"))

cbuelo/tvsews documentation built on Jan. 21, 2022, 1:31 a.m.