Flag power gains and power losses against the comparator

Adds the power difference, its interval, and the gain and loss flags derived from that single interval. Deciding both from the same interval keeps gains and losses at the same stringency; they were previously detected with two different rules – a z-test built from interval widths on one side, and non-overlap of two intervals on the other.

Usage

flag_power_differences(results_df, correlation = 0)

Arguments

results_df

A data frame of operating characteristics.

correlation

Correlation between the two estimates, passed to mover_difference_ci().

Value

The data frame with power_difference, power_difference_lower, power_difference_upper, power_gain and power_loss columns added.