I have a long data set with 6 years of data for every state. There are 7 variables that have missing data. I want to replace data by state and year where if a variable is missing for a year, replace it with the average of the data value for the year directly above and below, else replace it with data from the nearest year. I have been working on this array for a long time and cannot make it work - with this code I am receiving the following error: ERROR: Array subscript out of range at line 2276 column 49.   data new; set old;   /* Set missing values to the average of the previous and next years */ array indicators[7] indicator1 indicator2 indicator3 indicator4 indicator5 indicator6 indicator7;   /* Handle missing values for the first year separately */ by state year;   if first.state then do; do i = 1 to dim(indicators); if missing(indicators[i]) and not missing(indicators[i+1]) then indicators[i] = indicators[i+1]; end; end; /* Handle missing values for the last year separately */ if last.state then do; do i = 1 to dim(indicators); if missing(indicators[i]) and not missing(indicators[i-1]) then indicators[i] = indicators[i-1]; end; end; /* /* Handle missing values for the intermediate years */ if not first.state and not last.state then do; do i = 1 to dim(indicators); if missing(indicators[i]) then do; /* Calculate the average of the previous and next years */ if not missing(indicators[i-1]) and not missing(indicators[i+1]) then indicators[i] = (indicators[i-1] + indicators[i+1]) / 2; /* Look for the closest non-missing year */ else if not missing(indicators[i-1]) then indicators[i] = indicators[i-1]; else if not missing(indicators[i+1]) then indicators[i] = indicators[i+1]; end; end; end; drop i; run;     Any help would be GREATLY appreciated! ... View more

I am trying to estimate a risk difference using GEE (proc genmod). The outcome is clustered count data modeled as a rate (deaths over births), so Poisson distribution is preferred but I can only seem to get this to work with the normal distribution.    Here's what works:  proc genmod data= a.pmss; class geoid1; model prmr= RS_pctNOHS/ link=identity dist=normal; repeated subject=geoid1; weight births; run; *RD: 1.6358 1.1015 to 2.1701   Any idea for code to get a similar output but based on the Poisson or Binomial? So far nothing will converge.    ... View more

I have a set of county-level indicators that I am trying to robustly scale for comparability across regression models. I've tried many things but can't seem to find the code to do this.  ... View more