RepARS.MIX {bear} | R Documentation |
Adjusted R squared (ARS) method: This method selects data points to estimate lambda(z) based on the maximun adjustedR squred values. It starts with the last three data points from the concentration-time course, performing log-linear regression to calculate the slope of that tail portion of the concentration-time curve. And then the last 4 data points, the last 5 data points, on and on until it excludes the data points of Cmax. Thus, this method may exclude the data point of (Tmax, Cmax). WNL v6. has the similar algoirthms like this.
Statistical analysis (lme, 90CI...): With a two-treatment, more than two period, two-sequence randomized crossover design, linear mixed-effects model may be used. The lme in nlme package is used to obtain estimates for the adjusted differences between treatment means and the standard error associated with these differences. Log-transformed BA measures will also be analyzed.