FYI beginning with SAS9.4 MEDIAN is also an aggregate function and can be used with proc sql, see sample below. See also doc here SAS(R) 9.4 SQL Procedure User's Guide proc sql; create table sample as select make , model , invoice from sashelp.cars group by make having median(invoice) = invoice ; quit;
... View more
With both GPLOT and SGPLOT you can group your data, see sample below: proc gplot data=sashelp.class; plot age*height = sex ; run; quit; proc sgplot data=sashelp.class; series x=age y=height / group=sex ; run;
... View more
You can make use of BY group processing and the SUM statement. See sample below: data have; infile cards dlm=","; input code : $8. year : 4. count : 8. sum_count : 8. ; cards; 10,2000,1,1 10,2001,.,. 10,2002,1,1 10,2002,1,2 10,2003,1,1 10,2003,1,2 10,2003,1,3 20,2000,.,. 20,2000,1,1 ; data want; set have; by code year; if first.year = 1 then do; sumCount2 = 0; end; sumCount2 + count; if missing(count) = 1 then do; call missing(sumCount2); end; run;
... View more
Hi Assuming the input data is sorted by var_1, then the following code should do what you want: data want; set have; by Var_1; if first.var_1 = 1 then do; retain checkForA; checkForA = 1; end; if checkForA = 1 and var_2 = "A" then do; var_3_sas = 1; checkForA = 0; end; var_3_sas = coalesce(var_3_sas, 0); run;
... View more
Just in case the two lines eventDate = datetime(); format eventDate datetime22.3; are only for testing purposes, they are not needed in your code
... View more
This is a good case for the PRXMATCH function, see example below. You might also try the ANYDIGIT function data have; attrib x format=$200.; x='<skl>asffaafs 15178</skl>'; output; x='<skl>30653</skl>'; output; x='<skl>353</skl>'; output; x='353'; output; run; data want; set have; d5match = prxmatch("/\d{5}/", x); if d5match > 0 then do; d5 = substr(x, d5match, 5); end; run;
... View more
There are various ways, using DATEPART and YEAR function or DTYEAR format, see below data have; eventDate = datetime(); format eventDate datetime22.3; eventYear = year( datepart(eventDate) ); eventYear_c = put(eventDate, dtyear4.); run;
... View more
You can use Proc REPORT with the NOHEADER option or depending on the ODS desination make use of the ODS text="", see sample below. The TITLE statement will not produce output without some Proc printing something. data title; title = "This is my title Proc Report"; run; proc report data=title nowindows noheader nocenter style(column) = { background=yellow }; run; ods text="this is my title ods text=";
... View more
Try to use the dummy var only to color the complete line. Another approach would be the use of arrays in the compute block. Using this technique it is quite simple to compare a 100 value pairs, see example below. I guess part of the code will be created with a macro if you have a 100 value pairs to compare. data have; id_actual = 1; id_expected = 1; Var1_actual = 10; Var1_expected = 10; Var2_actual = 10; Var2_expected = 12; Var3_actual = 11; Var3_expected = 11; output; id_actual = 2; id_expected = 3; Var1_actual = 10; Var1_expected = 12; Var2_actual = 10; Var2_expected = 10; Var3_actual = 13; Var3_expected = 14; output; run; proc report data=have nowindows; column id_actual id_expected var1_actual var1_expected var2_actual var2_expected var3_actual var3_expected _dummy ; define id_actual / display; define id_expected / display; define var1_actual / display; define var1_expected / display; define var2_actual / display; define var2_expected / display; define var3_actual / display; define var3_expected / display; define _dummy / noprint; compute _dummy; array xAct{*} var1_actual var2_actual var2_actual; array xExp{*} var1_expected var2_expected var2_expected; if id_actual ne id_expected then do; call define(_row_, "style", "style={background=green}"); end; else do; do i = 1 to dim(xAct); if xAct{i} ne xExp{i} then do; call define(vname(xAct{i}), "STYLE", "Style={background=yellow}"); call define(vname(xExp{i}), "STYLE", "Style={background=yellow}"); end; end; end; endcomp; run;
... View more
Since all numeric variables are by default ANALYSIS variables, you can not use "age" to access the value but rather "age.sum". You can also define age to be a DISPLAY variable and then your code will work. See example below: proc report data=sashelp.class nowd; column name age sex weight; define age / display; compute age; if age in (13,15) then call define("age","style","Style={background=yellow}"); endcomp; compute sex; if sex='M' then call define(_row_,"style","style={background=green}"); endcomp; run; You can also add a dummy column and do all the processing in just one compute block, this will make it easier if the logic gets more complicated, see also an example below: proc report data=sashelp.class nowd; column name age sex weight _dummy; define _dummy / noprint; compute _dummy; if sex='M' then do; call define(_row_,"style","style={background=green}"); end; if age.sum in (13, 15) then do; call define("age.sum","STYLE","Style={background=yellow}"); end; endcomp; run;
... View more
Since you only change the SCL List contents the widget is not aware of any changes. You can you the _refresh, or _needRefresh method, so that the widget redraws it s content. Something like: rc = insertc(textPad.text, tempString, -1); textPad._needRefresh();
... View more
Have a look at the code below, you need to first convert from numeric to char and then use appropriate informat to make a SAS date from the char variable data want; infile cards dlm=","; input @1 date_n 8. ; * convert numeric to char left aligned ; date_c = put(date_n, 8. -L); * convert from char to SAS date ; date_s = input( date_c, ddmmyy8.); format date_s date9.; cards; 10413 31122013 ;
... View more
the Note NOTE: Character values have been converted to numeric values at the places given by: (Line):(Column). 19:16 19:21 NOTE: Invalid numeric data, COL5=' 1* 6' , at line 19 column 21. Indicates that char values have been converted to numeric values, they are not coming from the CATX function
... View more
Hi as Amir mentioned you can make use of variable lists in the CATX function, see also sample below using an array:
data want;
length colString $ 120;
array myVars{30} $ 4 col1 - col30;
nChar = rank("A");
do i = 1 to dim(myVars);
myVars{i} = byte(nChar);
nChar + 1;
end;
colString = catx(",", OF myVars{*});
run;
... View more
