Calculate a lagged rolling mean

Returns a vector of the rolling mean of a vector calculated for a given window size.

rolling_mean(vec, window = 3)

Arguments

vec	A vector of numeric values
window	Size of rolling mean to compare against

Value

Vector of same length as vec with each element being the mean of the last window elements in vec (including current element).

Details

The rolling mean is calculated using the available data. For elements at start of the vector before a full window is able to fit, the mean is simply taken with respect to the available values. E.g. for a window size of 3 the first element will be the same as the source vector, the second will be the second elements distance from mean of the first and second elements of the source vector, and the remaining will be the mean of the element and the previous two elements.

This is used internally to compute elbow points in entropy calculations

See also

Other helpers: build_score_fn(), combination_indices(), get_sweep_pair_counts(), get_sweep_results(), print.sbm_network(), verify_model()

Examples

my_vec <- 1:25
rolled_mean <- rolling_mean(my_vec)
plot(rolled_mean)

# An increased window size
rolled_mean_window_5 <- rolling_mean(my_vec, window = 5)
plot(rolled_mean_window_5)