I reread your post - I see what you mean. If time = . on the first record, then treat the second record as the first record? Depending where you are heading downstream of this program, I see two possibilities: 1. filter out records where time is missing in the set statement if you don't need them downstream of this step, then they won't be counted in the first. calculations. If you need them later, you could handle them separately and append them back in later. 2. you could have a variable that you retain to keep track of this situation - like firstnonmissingvalue ... View more

@ari wrote: @Kurt_Bremser: Thanks, what if the values need to be replaced with values other than 0. I have updated the text to represent such a case. Is there a way to identify and specifically modify the first row (of the two rows with same time and week values) Then we need to know from where to retrieve those replacement values. If you need to replace only the first line of a multiple of identical lines, I suggest - run a data step where you set a counter variable from _n_ - sort the dataset by descending counter (you reverse the order completely) - use by as already suggetse in the next data step, and set your values only when last. and not first. of the last by variable is met - recreate the original order by sorting by counter; if you use out= in the proc sort to create a new dataset, you can drop counter there. ... View more

@RW9: Perfect, thanks ... View more

I'm not sure I fully comprehend. Please show some data before and after using the new criterion. ... View more

Please run next code. No changes to step 1 - creating HAVE dataset. There are still few observations with serial NE cal_sn (calculated serial number); Check dataset CHECK. proc sort data=have out=haves; by id time dose c3; run; data count_cs(rename=(k1=id k2=time k3=dose k4=c3)); set haves end=eof; by id time dose c3; retain phase 0 /* 0 1st time only */ k1 k2 k3 k4 count_cs; /* count consecutive samples */ if phase=1 and dose=1 or last.id or eof then output; if dose=1 then do; phase=1; k1=id; k2=time; k3=dose; k4=c3; count_cs=0; end; if dose=0 then count_cs+1; keep k1 k2 k3 k4 count_cs; run; data want; merge haves count_cs(rename=(count_cs=count)); by id time dose c3; retain count_cs sn dose_time calc_sn; /* sn = serial number */ lag_time = lag(time); if first.id then sn=0; if dose=1 then do; dose_time=time; count_cs=count; calc_sn=.; output; if count_cs < 3 then calc_sn=1; else do; if sn=0 then sn=2; else sn+1; calc_sn=sn; end; end; else do; if time - dose_time > 192 then calc_sn=1; output; end; drop sn count; run; data check; /* should contain 0 obs */ set want; if serial ne calc_sn; run;   ... View more

@ari wrote: @data_null__. It works but difficult to understand to modify to copy the remaining variables as such in the dataset. My original dataset contains 50 variables and those be copied as such. Any pointers? Thanks   Did you do any reading in the SAS online documentation about the statements used in the program?   change   set run(keep=day); to set run(drop=id run); All variables in RUN except ID and RUN are added but their values are not carried forward.   The SET statement is to bring in all the variables that you do not want have values carried forward. ... View more

This warning indicates that your query didn't include a max or sum summary function. Please post the code that you ran.  ... View more

Use PROC COPY. libname in xlsx "my excel file.xlsx"; libname out xport "my v5 transport file.xpt"; proc copy inlib=in outlib=out; run; But if the sheetnames and column headers in your Excel file do not meet the 8 character name limits of a SAS V5 transport file you will have errors.  ... View more

thanks LinusH ... View more

data have; infile cards dlm='09'x truncover; input id a b c; cards; 1 2 2 1 2 3 1 0 3 1 0 0 4 1 2 3 5 3 2 1 ; run; data _null_; set have end=last; retain max; temp=max(of a--c); if temp gt max then max=temp; if last then call symputx('n',max); drop temp; run; data want; set have; array x{*} a--c; array g{*} g1-g&n; do i=1 to dim(g); if i in x then g{i}=1; else g{i}=0; end; drop i; run; ... View more

Thanks Kurt ... View more

Here's an example that demonstrates how it works.   ods powerpoint file='/folders/myfolders/out.pptx' style=meadow; title 'Class - Height by Weight'; proc sgplot data=sashelp.class; scatter x=height y=weight; run; title 'Cars - City vs Highway MPG'; proc sgplot data=sashelp.cars; scatter x=mpg_city y=mpg_highway; run; ods powerpoint close; ... View more

/*Create dataset*/ data create_data; infile datalines; input id t base value; datalines; 1 1 5.053078 14.61624 1 5 5.687529 36.09745 1 7 4.149219 45.9342 1 14 3.342423 72.15495 ; run; data want; set create_data end=last; by id t; /*Initialize variables*/ retain t_first 0 base1 0 t_first 0 t_last 0 value_last 0 t_diff 0 value_diff 0; /*Read base and t values for first observation*/ if first.id then do; base1=base; t_first=t; end; /*Read t and value values from last observation*/ if last.id then do; t_last=t; value_last=value; end; /*Perform calculation*/ t_diff=t_last - t_first; value_diff=value_last-base1; /*Subset the dataset for the last observation*/ if last then output; run; proc print data=want noobs; /*Displaying variables of interest*/ var t_first t_last t_diff base1 value_last value_diff; run; ... View more