SAS信用风险建模（2） [推广有奖]

一篇关于SAS编程计算AUC的文章，童鞋们参考下吧！
其他文章，可以见：
http://www.pinggu.org/bbs/thread-889118-1-1.html

The SAS® Calculations of Areas Under the Curve (AUC) for Multiple Metabolic Readings


ABSTRACT

In biomedical studies, the computation of Area Under the Curve (AUC) is a convenient way to
combine multiple readings, such as some metabolic values (glucose and insulin) or blood (serum
or plasma) concentration within a specific time interval, into a single index in drug toxicology and
pharmacokinetics.  Thus, computation of a representative index AUC value will be an essential
and interesting task in our metabolic studies and SAS programming applications.

The SAS program I present here is a flexible and compact code for parameters adaptable input
and comprehensive AUC computations, which is suitable to be applied to the case of a discrete
curve over a specific time interval.  In the program, SAS macro, SQL and GPLOT procedures are
used to calculate and present sample data.  This code can be utilized to all three kinds of AUC
computations, the Total AUC (no baseline), Net Incremental AUC (with baseline) and Positive
Incremental AUC (with baseline).  The “negative” measured values after subtracting the baseline
value are taken into account and also within the scope of all possible readings.  Moreover, the
sample data can be measured at various time intervals (∆t), and the intervals do not need to be a
constant during the measurements.