I work with at form thats use dynamic sas in a front end excel solution.  When tagattr formula i used with a if clause excel puts #Navn (danish) #NA in a correct code. See code exampel. If use of F2 without any changed correct formula has been used.  The code is in danish because we use excel2016 in danish - and because autotranslation in xml with excel function length or upper return error i xml to excel.  The english is   [tagattr="format:#,##0 formula:if(length(RC[-4])>0,RC[-2]*RC[-3],0)"] ; Full danish code with tabulate= ;  (danish excel syntax) gives error. It doesn't matter if tagattr is written in define or compute.  The full code i want to use is tagattr="formula:hvis(længde(RC[-1])>0,j5*(Satser!B4/100),0)" so that the user make input (here cell=j5) and return the right values. "Sater!b4" is a reference to another worksheet. The code is correct, but in excel is #Navn/#NA shown. How can i show the results in var=testhvis with out the #Navn error?     ods excel file="&udadhoc.\test_tagattr.xlsx"; proc report data=sashelp.class out=testcase ; column sex height weight testudenhvis testhvis ; define sex / group; define height / analysis sum; define weight / analysis sum; define testudenhvis / computed style(column)=[tagattr="format:#,##0 formula:RC[-1]*RC[-2]"] ; define testhvis / computed style(column)=[tagattr="format:#,##0 formula:hvis(længde(RC[-4])>0,RC[-2]*RC[-3],0)"] ; rbreak after / summarize ; run; ods excel close; Sex Height Weight testudenhvis testhvis F 545,3 811 442.238 #NAVN? M 639,1 1089,5 696.299 #NAVN?   1184,4 1900,5 2.250.952 #NAVN? ... View more

I'm looking for a denominator syntax for subgroups total in proc tabulate. My "product" is a report with 2 across groups. Rowgroups are dynamic multible. Case with sashelp.class below. The problem is that subgroup total for F and M is not 100 but the same as rowpctsum. Other calculations er ok. Normaly solved in proc report, but report make a column for every combination (se age 11 and 15). To remove these we have to made both data step and macro loop in compute statement. Solution with a subgroup total in dataset with sql give a problem in the "all" statement. Any idees? data class; set sashelp.class; firstl = substr(name,1,1); run; proc tabulate data=class; where firstl in ("J","A"); class firstl sex age; var height; table (all="Total" firstl="") * (height="SUM"*sum="" height="PCT totalrække"*rowpctsum="" height="PCT subgroup"*pctsum<sex*all age all>="" ), sex=" " * (age="" all="I alt") all="Total" / box="Udfordring: PCT subgroup f*ialt=100% og m*ialt=100 "; run; proc report data=class nowd; title 'I report medtages alle kolonnekombinationer i _cn_ -> fjernes ved noprint og macro loop i compute'; where firstl in ("J","A"); column firstl n sex, age ; define firstl / group; define sex / across; define age / across ; rbreak after /summarize; run; ... View more

I'm looking for a way to integreate the possibility for controlling color and size of the border lines in proc gmap with combined area and block (se below). Annotation with poly/polycont as described in difference whitepapers give error when use of block and there is no options to control outlines in area statement. The reason is, that I want to illustrate two results for different populations on a subgroup. If you only show one result you can annotate the borders in choro statement and let the bordercolors illustrate the subgroup. I'm looking for at way to compine the border annotation in choro with area/block. Do someone have an idea. proc gmap data=analysevis map=mapout annotate=annolegend all; id boligtype navn afd; area b_gnsb / midpoints=(1 to 6) missing nolegend; * error if annotate ni area; block a_gnsb / midpoints=(1 to 6) cempty=black coutline=grey legend=legend1 BLOCKSIZE=1 WOUTLINE=1 /*annotate=annoborder doesnt work with block statement*/ name="&gnavn._b"; format a_gnsb b_gnsb borger.; run; quit; ... View more