Hi ballardw!  Apologies for being unclear. For clarification, the sample code that I provided generates the PDF file but without  images1, image2, image3 etc.   Below is the SAS log that I generated but with only image 1 and 2. (I had the same output results when I tried to run the SAS script with Image 1 to 13 but wanted something trimmed down for this post)   ... View more

Hi @Reeza  and @ballardw , Thank you both for your feedback. What I am trying to do is calculate patient days during the patients admission for each month on each unit they stay on.  date1 would be the patients admission date and date2 would be the discharge date.  Consider State as the location of the patient. I tried the codes you suggested but added an additional row to the "HAVE" table data step before running your suggestions. data have; infile datalines ; format date1 date2 date7.; input date1 :date7. State:$ date2 :date7. ID ; datalines; 01Jan23 NY 04Feb23 1 01Jan23 NY 04Feb23 1 01Jan23 NY 04Feb23 1 01Jan23 NY 04Feb23 1 02Jul23 CA 04Jul23 2 02Jul23 CA 04Jul23 2 02Jul23 CA 04Jul23 2 20Sep23 CA 30Sep23 3 ; run;  When I try @Reeza's suggestion, obs 8 should be 11 (sep 20 to sep 30) but the output is 30 instead. When I tried @ballardw 's suggestion was missing the Year_month that i want the number of days to be attributed to. Is there a way to dedup to have just the below obs/row (obs 1, 2, 9, 12) along with the year_month the days are being attributed to?   ... View more

state yr_month days NY 2023_01 31 NY 2023_02 4 CA 2023_07 4 The number of months would depend on how many months are between date1 and date1 ... View more

that only gives number of days between date1 and date2. I want to get days by month by state. ... View more

Thank you so much @DanH_sas !!!!!  ... View more

@ballardw  I added the  DISCRETEATTRVAR statement you had suggested and removed "fillattrs=(transparency=0.4 color=Color);" from the line of code but the color logic isnt applied. Did I miss something?    I may be approaching this wrong in the code...what i want to do is have each Bar (Total_obs2) be a specific color base on the HH_rate variable. Is that possible at all?    This is the code I used.   proc template; define statgraph barline; begingraph; entrytitle "Have - HH Rate and Total HH Observations "; /* Define color mapping */ discreteattrmap name="color_map" / trimleading=true ignorecase=true ; value '>=95' / fillattrs=(color=Blue transparency=0.4); value '<95 & >85' / fillattrs=(color=Yellow transparency=0.4); value '<85' / fillattrs=(color=Red transparency=0.4); enddiscreteattrmap; discreteattrvar attrvar=groupmarkers var=color attrmap="color_map"; layout overlay/ xaxisopts=(display=(tickvalues)) yaxisopts=(label="# HH Observations" griddisplay=on offsetmin=0 linearopts=(viewmin=0 viewmax=500 thresholdmin=0 thresholdmax=1)) y2axisopts=(label="HH Rate %" griddisplay=on offsetmin=0 linearopts=(viewmin=0 viewmax=100 thresholdmin=0 thresholdmax=0)) ; innermargin / align=bottom; axistable x=Unit value=Total_Obs2 / display=(values); endinnermargin; barchartparm x=Unit y=Total_Obs2 /; /* fillattrs=(transparency=0.4 color=Color);*/ seriesplot x=Unit y=Compliance_Rate / yaxis=y2 name="series" lineattrs=(color=green thickness=3) markerattrs=(symbol=circlefilled size=100) datalabel=Compliance_Rate; discretelegend "series"; endlayout; endgraph; end; run;   ... View more

Hi @DanH_sas  and @ballardw , one more question >.< is it possible to add criteria to have the bars be a certain color if it meets it? I tried the below method but was not successful. Would you have any suggestions?   data have; input Unit $ Total_Obs2 Compliance_Rate; datalines; a 139 85.5 b 419 90.2 c 636 92.7 d 4854 88.3 e 3807 84.6 f 1903 93.8 ; run; /* Create a new variable for color */ data have; set have; if Compliance_Rate >= 95 then Color = 'Blue'; else if Compliance_Rate > 85 then Color = 'Yellow'; else Color = 'Red'; run; /* Define the modified graph template */ /* Define the modified graph template */ proc template; define statgraph barline_labels; begingraph; entrytitle "HAVE - Compliance Rate and Total Observations"; /* Define color mapping */ discreteattrmap name="color_map"; value '>=95' / fillattrs=(color=Blue transparency=0.4); value '<95 & >85' / fillattrs=(color=Yellow transparency=0.4); value '<85' / fillattrs=(color=Red transparency=0.4); enddiscreteattrmap; layout overlay/ xaxisopts=(display=(tickvalues)) yaxisopts=(label="# Observations" griddisplay=on offsetmin=0 linearopts=(viewmin=0 viewmax=500 thresholdmin=0 thresholdmax=1)); innermargin / align=bottom; axistable x=Unit value=Total_Obs2 / display=(values); endinnermargin; barchartparm x=Unit y=Total_Obs2 / fillattrsmap "color_map"; seriesplot x=Unit y=Compliance_Rate / yaxis=y2 name="series" lineattrs=(color=green thickness=3) markerattrs=(symbol=circlefilled size=100) datalabel=Compliance_Rate DATALABELPOSITION=bottom ; discretelegend "series"; endlayout; endgraph; end; run; /* Close the ODS HTML destination */ ods html close; /* Set the output directory and graph size */ ods listing gpath="C:file path\_Graphs"; ods graphics / reset imagename="HAVE" imagefmt=JPEG width=1200px height=700px; /* Generate the modified graph with data labels */ proc sgrender data=have template=barline_labels; format Total_Obs2 comma12.; run;       ... View more

Thank you!!! ... View more

Thank you  @FreelanceReinh !!!!!!   The word/acronym "Str" was just on top of my head at the moment so I just used it as an example variable name.    Posting the data step for anyone who needs it in the future!   data have; str="2023-02-03 15:49:06"; Want_str=input(str, yymmdd10.); format want_str mmddyy10.; run;       ... View more