LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (December 1997, week 1)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:   Mon, 1 Dec 1997 10:18:46 +0000
Reply-To:   Paul Monaghan <pfm@SCMP.SCM.LIV.AC.UK>
Sender:   "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:   Paul Monaghan <pfm@SCMP.SCM.LIV.AC.UK>
Subject:   ICE macro (IML) for left/interval censored survival data
Content-Type:   text/plain; charset=US-ASCII

Many thanks to Alan Cantor and Dr. Hans-Peter Altenburg for their quick reply to my request for a SAS macro to compute nonparametric survival estimates when the data may be left, right, interval or uncensored. They suggested using the ICE macro in IML.

I have had problems implementing this algorithm when some of the data are uncensored. It appears that the problem is in the call to the INTERVAL function. As an example, take the example given by the ICE macro, augmented with two uncensored times at 1.5 and 1.7.

/* Define the data */ data ex2; input l r f; cards; 0 2 1 1 3 1 1.5 1.5 1 1.7 1.7 1 2 10 4 4 10 4 ; run; %ice(data=ex2,time=(l r),freq=f);

The macro runs until the following error message is returned:

ERROR: (execution) Invalid argument to function.

count : number of occurrences is 2 operation : LOG at line 5238 column 5 operands : XT

XT 6 rows 1 col (numeric)

0.6 0.8 0 0 0.4 0.2

statement : ASSIGN at line 5238 column 5 traceback : module EM at line 5238 column 5 module INTERVAL at line 5238 column 5 NOTE: Paused in module EM. Exiting IML.

The count line returns the number of uncensored observations (I have tried this for different data sets), and my guess is that the macro is trying to take log(0) somewhere.

I would be extremely grateful if anybody can help me with this problem, or if this is a known error in this macro.

Paul Monaghan, University of Liverpool.


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