From: Kaushik Ghose <Kaushik_Ghose@hm...>  20080813 18:15:21

Hi, This is more a coding/algorithm question than a straight matplotlib question. I have x and y data based on a parameter k. x and y have noise, so that even though k increases monotonically neither x nor y are guaranteed to do so. x can have duplicates. I need to find the area under the curve x,y (You may have guessed, x is false alarms, y is hit rate, and this is the ROC) What I'm doing is rather cumbersome. I first sort x (shuffling y accordingly) Then I crawl through x discarding the duplicates. Only then does interp give me reasonable values which I then use for finding the area under the curve. Is there a function in matplotlib that handles cases like this? Thanks! K 