> Perhaps counts for specific stratum is missing and it should be zero? If so, how do I go about addressing this issue? Yes, you need to have an entry for each group if it's 0. Make sure that each X has every strata, you cannot have missing categories. You can use PROC FREQ and the sparse option. I think the following works: proc freq data=data_set; table age_group_5*gender / out=want sparse; weight pop_denominator; run; ... View more

If you need to automate you could simply run a Proc MEANS prior to running LOGISTIC and save the standard deviation to a MACRO variable. ... View more

@Jack_Smitherson wrote: Thank you for your response. I do want to run some proc sql later on, so simply formatting the variable is not the most optimal option.   In the first option you gave me "Convert it to a numerical variable, which is basically the year function" - how can I go about applying that to 12 months (i.e. 01Jan2000, 1Feb2000, 01Mar2000, etc.) with five years worth of data? (i.e. 2000-2004; I have a combination of 01Jan2000, 01Jan2001, 01Jan2002, etc.).    Thanks again!      That depends on your data structure.   Use the YEAR() function on the variable to obtain the YEAR.  ... View more

There is no "UNIT" statement in ARIMA.  I suspect you want to find how the response variable changes as the values of some input variable change.  Kind of scoring or what-if analysis.  In ARIMA there is no simple way to do it.  You could manually provide the desired input variable values in the forecast region (in the input data set) and reforecast.  ... View more

The data is from a cross-sectional survey. I'm attempting to calculate the age-standardized prevalence rate. I have counts for outcome of interest, age groups, and a standard population.  ... View more

proc sql; describe table dataset_name;          quit;   the log will show you the data structure; use quotes accordingly as noted by an earlier responder.   Ron Fehd  data structure! ... View more

@Jack_Smitherson wrote: When I run the code on SAS, this is the message that I get:  "WARNING: Multiple lengths were specified for the variable VERDATE by input data set(s). This can cause truncation of data." This tells us that you have multiple data sets and that the lengths of a variable are different in each. If this is the date variable you are manipulating then the description you started with may be inaccurate for some of the data due to truncation.   And it may be time to show the exact code you ran that did not work as needed. ... View more

Look up PROC STDRATE and specifically the documentation has fully worked examples.  ... View more

@Jack_Smitherson wrote: Hi SuryaKiran,   When I import the file anddo the proc summary procedure, I get this message on SAS: " Variable columnb in list does not match type prescribed for this list."   Not sure why I'm getting this message...  Proc summary will only allow actual numeric variables on the VAR statement. So you have an issue with how ever you imported the data the variable columnb came in as character. This is a common problem when starting with Excel especially using Proc Import as the procedure has to guess a variable type from the first few rows of data and if the values are all missing will default to character or if the data is mixture, not an infrequent occurence with data entered into Excel.   Just because it looks like a number when viewed in Excel it is not necessarily so. ... View more