|Date: ||Thu, 29 Mar 2007 16:22:15 -0400|
|Reply-To: ||Samantha Huang <cheng_an@YAHOO.COM>|
|Sender: ||"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>|
|From: ||Samantha Huang <cheng_an@YAHOO.COM>|
|Subject: ||Percentile values from CDF curve (normal)|
I tried to obtain the percentile values of the normalized CDF curve (instead
of the empirical curve) - is anyway SAS can output the data?
OK, here are two things I have tried:
1. I used the proc capability cdfplot / normal to obtain the correct CDF
curve I need - but I don't know how I can output the X-axis values for all
the percentile (5% - 100%, every 5%).
2. I tried to use the proc univariate function (pctlpts = 5 to 100 by 5)
I got the percentile data - but they were empirical (observed) data - and
since I have very few data points, many percentiles have exactly the same
So my question is: is it a way to combine the two, that is, to output the
percentiles based on the normalized CDF instead of the empirical one?
Many thanks for your kindly help!!!!