You don't need to "strip" any of the characters from the string to convert it into a datetime value.  To ignore the milliseconds just use an INFORMAT with a width of only 19 and it will only read the first 19 bytes of the string.     Of use a wider width and read in the milliseconds.  Then you can decide if you want to keep them in the value and set the format to not display them.  Or remove them by either truncating to an integer, or rounding to an integer. data want; set have; actual_datetime = input(start_datetime,anydtdtm19.); format actual_datetime datetime19.; run; Result Obs Start_Datetime actual_datetime 1 2022-06-26 19:43:00.358 26JUN2022:19:43:00 2 2022-06-26 21:33:13.403 26JUN2022:21:33:13 3 2022-08-30 07:26:04.178 30AUG2022:07:26:04 ... View more

See this note on overdispersion. Your Poisson model suggests overdispersion as evidenced by the Pearson or Deviance divided by the DF being considerably larger than 2. The negative binomial is often used for overdispersed data. If you do that, these statistics are only about 1.1 suggesting that overdispersion is not a problem with that model. Or, as suggested in the note, another common approach is to use a GEE model which seems to give a similar estimate on your primary predictor as the non-GEE negative binomial model (using either Poisson or negative binomial distribution in the GEE model). proc genmod data=forSAS namelen=65; class Female Location Income_thresh id; model Errors=LPA_Daily_min ST_win Female Location Income_thresh / dist=negbin; repeated subject=id; effectplot fit(x=LPA_Daily_min) / ilink; run; ... View more

Thank you for your PROC FREQ data aggregation solution (Method 1 SAS code below). I have also provided two additional solutions using PROC SUMMARY and PROC SQL (Methods 2 and 3 SAS code below). Below is the SAS Output for aggregated data. DUPERSID RXNDC RXNAME count ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ 2320018102 -15 ANTICOAGULANTS 2 2320038101 -15 ANTICOAGULANTS 3 2320105101 -15 INSULIN 4 2320222101 73070010011 INSULIN ASPA 8 Question: What DATA step code would I write to create the output data set (e.g., example observation below) based on the above data set as input? Any hints (or references) would be appreciated. DUPERSID RXNDC RXNAME count RXNDC1 RXNAME2 RXNDC1 RXNAME2 ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ 2320018102 -15 ANTICOAGULANTS 2 -15 ANTICOAGULANTS -15 ANTICOAGULANTS ******************************************************************** data testdata; infile datalines truncover ; input DUPERSID :$10. RXNDC :$char11. RXNAME & $50.; datalines; 2320018102 -15 ANTICOAGULANTS 2320018102 -15 ANTICOAGULANTS 2320038101 -15 ANTICOAGULANTS 2320038101 -15 ANTICOAGULANTS 2320038101 -15 ANTICOAGULANTS 2320105101 -15 INSULIN 2320105101 -15 INSULIN 2320105101 -15 INSULIN 2320105101 -15 INSULIN 2320222101 73070010011 INSULIN ASPA 2320222101 73070010011 INSULIN ASPA 2320222101 73070010011 INSULIN ASPA 2320222101 73070010011 INSULIN ASPA 2320222101 73070010011 INSULIN ASPA 2320222101 73070010011 INSULIN ASPA 2320222101 73070010011 INSULIN ASPA 2320222101 73070010011 INSULIN ASPA ; proc print data=testdata; run; title 'Method 1'; Proc freq data=work.testdata noprint; tables dupersid * rxndc * rxname / list nopercent out=FreqCount (drop=percent); run; proc print data=work.FreqCount; run; title 'Method 2'; proc summary data=testdata nway ; by dupersid rxndc rxname; output out=want (rename=(_freq_=count)drop=_type_) ; run; proc print data=work.want; run; title 'Method 3'; proc sql; select dupersid, rxndc, rxname, count(*) as count from testdata group by dupersid, rxndc, rxname; quit; ... View more

Yes, I meant the "treatment" and not covariates in this case. ... View more

Hi:   The good news is that any independent user can set up a SAS OnDemand for Academics account, which would give you access to both PROC SGPIE and PROC GCHART. That way you could compare the output. So, even if you do not have access to PROC SGPIE through your work or school account, you can set up a SAS OnDemand account for non-commercial learning purposes. PROC SGPIE was first available in SAS 9.4 Maintenance release 6, which is several years old now.   The syntax for PROC SGPIE would turn out to be something like this (untested): title "SGPIE Hours Watching Monthly"; proc sgpie data=work.DataforSAS2; pie Program / response=Hours_Watching_Monthly_rec datalabeldisplay=(category percent) datalabelloc=inside direction=clockwise otherpercent=5 sliceorder=unformatted; run; title; You could play with using STYLEATTRS or play with using an attibute table. You can set up a SAS OnDemand account by going to https://welcome.oda.sas.com and sign in with your SAS Profile. The first time you register for an account, you would need to specify a geographic region. Cynthia ... View more

Sounds like you want to break the records into periods by checking for gaps between START and the previous END.   Not clear what your criteria is for declaring a gap too large.  For example you might use 24 hours. data periods ; set duration; by ppt_id; gap = start - lag(end); if first.ppt_id then do ; gap=0; period=1; end; else if gap > '24:00't then period+1; format gap time12.; duration_h=duration_s/'01:00't ; run; proc summary data=periods ; by ppt_id period; output out=summary(drop=_type_) sum(duration_:)= min(start)= max(end)= ; run; data summary; set summary; days = 1+datepart(end)-datepart(start); hours_day = duration_h/days; run; Result     ... View more

/*Using GROUP= option*/ proc sgpanel data=sashelp.heart; panelby sex / rows=2 layout=rowlattice; histogram weight/group=sex; run;   ... View more

@confooseddesi89 wrote: Thanks - for some reason the "if then do" statements did not work on the "format" statements, so I applied "input" and if-then-do statements before transposing, and formatted each time/date variable after the data were transposed. Of course not.  There is nothing to execute for a FORMAT statement.  It just set one of the many attributes of the variable.  Just like a LABEL statement. ... View more

@Rick_SAS gives a great discussion on what the STDCOEF option does in GLIMMIX here.  The presence of a link function in most (not all) models that use GLIMMIX makes standardization of the response variable problematic. ... View more

Excellent! This worked perfectly.   To obtain the dashed/dotted lines instead of colors, and to get a legend, here is my modified code for the data myattrmap and proc sgplot: data myattrmap; length linecolor fillpatterns $ 20 value $ 20; input id $ value $ linecolor $ fillpatterns $ FILLTRANSPARENCY linepattern $; fillcolor=linecolor; LINETHICKNESS=3; output; datalines; myid DEP01 black X1 0.8 Solid myid DEP0 black L1 0.9 Dot myid DEP1 black R1 0.9 Dash ; run; proc print; run; proc sgplot data=all dattrmap=myattrmap; band X=_XCONT1 lower= _lclm upper= _uclm / group=group attrid=myid; series x = _XCONT1 y = _PREDICTED / group=group attrid=myid; xaxis values =(-2 to 2 by 1) label='LPA (difference from mean; hrs) ' LABELATTRS=(size=13 Weight=Bold) VALUEATTRS=(size=13); legenditem type=LINE name='DEP01' /label='Overall' lineattrs=(pattern=Solid); legenditem type=LINE name='DEP0' /label='No report of D/A' lineattrs=(pattern=Dot); legenditem type=LINE name='DEP1' /label='>=1 report of D/A' lineattrs=(pattern=Dash); keylegend 'DEP01' 'DEP0' 'DEP1' / title='Mental health status' VALUEATTRS=(size=12) TITLEATTRS=(Size=14 Weight=Bold) location=outside position=top; yaxis grid values=(0 to 10 by 1) label='Depression (score)' LABELATTRS=(size=13 Weight=Bold) VALUEATTRS=(size=13) GRIDATTRS=(color=GrayCA thickness=.0mm);run; ... View more

First a general comment about disaggregation (going from a group or range of values to a specific) : as you are finding out, not easy. Often it is better to ask for an estimate of duration and then if your respondents have a hard time offer ranges for a best fit. AFTER they have had problems with the initial question. And the response should be in a different variable.   I might be tempted to look at your data as a stacked bar per respondent totaling the time. This would mean transposing your data so that the values are all in the same variable and you have a variable to indicate which measure is involved. An example using the the SASHELP.CLASS data set if your aren't familiar with transposing: Proc transpose data=sashelp.class out=trans; by name; var height weight; run; proc sgplot data=trans; vbar name /response=col1 group=_name_; run; I would hope that you have a unique respondent ID for the collected values. You likely need to sort your data to use the BY statement in transpose. Your ID would replace the name and height and weight would be replaced by the numeric time you have assigned for the categories. You would look for individuals with unusual appearance in the charts. Likely those totals of 20+ hours will be of interest to see if the responses may sense in any form. Sometimes you just have to throw out respondents as their responses are unreliable. Or categorize individuals into groups based on these totals with a caution that the 21+ hours group may not be reliable. ... View more