```Date: Tue, 30 Oct 2001 13:38:44 -0800 Reply-To: Dale McLerran Sender: "SAS(r) Discussion" From: Dale McLerran Subject: Re: Area Under Curve (AUC) Comments: To: mbarish@progenics.com Content-Type: text/plain Melanie, You could consider an alternative method in which you generate and operate on a translated curve as follows: 1) determine the minimum value anywhere along the curve 2) subtract the minimum from all values. This subtracts a rectangular area of (minimum*length of X axis) from the total area. 3) compute the area under the curve for the translated curve using the trapezoidal rule 4) add back the rectangular area of (minimum*length of X axis) If the minimum is a negative value, then subtraction of a constant negative value in step 2 was equivalent to adding rectangular area equal to (|minimum|*total time length), and the operation in step 4 results in subtracting this rectangular area. This should be an easier approach to computing the area than the approach which you mention below. Dale >Date: Tue, 30 Oct 2001 15:33:31 -0500 >Reply-To: Melanie Barish > Melanie Barish Area Under Curve (AUC) SAS-L@LISTSERV.UGA.EDU >Hello everyone. I've been asked to calculate area under the curve for a >graph of percent change from baseline versus time at blood draw. I've >been doing research in the listserv archives, and I have found some very >helpful info regarding the trapezoidal method of calculating AUC. >However, in this case, because my graph is of percent change from >baseline, there are datapoints below the zero line. I need to calculate >area under the curve for the parts of the graph above zero, and then >subtract area under the curve for the parts below zero. Does this sound >feasible via SAS, and if so, is there another method besides the >trapezoidal that I can use? > >Thanks so much for your help. > >Melanie --------------------------------------- Dale McLerran Fred Hutchinson Cancer Research Center mailto: dmclerra@fhcrc.org Ph: (206) 667-2926 Fax: (206) 667-5977 --------------------------------------- ------------------------------------------------------------ --== Sent via Deja.com ==-- http://www.deja.com/ ```

Back to: Top of message | Previous page | Main SAS-L page