About BrunoMueller

BrunoMueller · ‎02-13-2017

Hi Have a look at this blog entry Tracking progress in your program with SAS Enterprise Guide: another trick by @ChrisHemedinger It shows how to use the DOSUBL functtion together with the SYSECHO statement. Together they will do what you need. Bruno

BrunoMueller · ‎02-13-2017

In my example I did not specify any data sources in the Metadata of the stored process. Did you have a look at the datasets you specified in the OUTPUTDATA statement where they created, what is the content? Which SAS Version are you using? The Version of EG should not matter. Bruno

BrunoMueller · ‎02-13-2017

Hi Have a look at Proc STP this allows you to call a Stored Process, run the code and return the data to the calling environment. Doc is available here http://support.sas.com/documentation/cdl/en/stpug/68399/HTML/default/viewer.htm#p0p6ncj6i4s8w5n1a8jxhfkurpl1.htm Please note: the Stored Process should not use the %stpbegin and %stpend macro if you are not creating any output in your stp, but rather just tables. The code is not actually run on the Stored Process Server, but rather the same process that runs your code the name given in the OUTPUTDATA statement like outputdata sugus_mvar=work.stp_mvars; is avialable within the Stored Process code as macro variable _TARGET_SUGUS_MVAR Here is the code of the Stored Process: %let carMake = %lowcase(&carMake); data &_TARGET_myCars; set sashelp.cars; where lowcase(make) = "&carMake"; run; proc sql; create table &_target_sugus_mvar as select * from dictionary.macros order by scope , name , offset ; quit; Here the code using Proc STP proc stp program='/someFolder/STP/createData' ; list; inputparam carMake="Audi"; outputdata myCars=work.stp_data; outputdata sugus_mvar=work.stp_mvars; run; Bruno

BrunoMueller · ‎02-13-2017

Hi I just looked at some other example I did some time ago, and saw that Proc REPORT has support for PCTN and PCTSUM statistics. So the code can be a lot easier, see example below. But it is always good to have an alternate way of doing things. proc report data=sashelp.shoes nowd; where region in ('Africa', 'Asia','Canada'); column Product Region, (sales sales=salespct) region=region2, sales=salesPct2 ; define product/group; define region/across; define sales/'sales in USD'; define salespct/ pctsum 'in %' format=percent9.2; define region2 / across; define salespct2 / pctsum 'in %' format=percent9.2; rbreak after/summarize; run; Bruno

BrunoMueller · ‎02-10-2017

Hi With Proc REPORT, you have to do the percentage calculation yourself, see code below. For a task like this one usually uses arrays, but they gave me some errors, so this code uses individual variable names. In the COMPUTE BEFORE the column totals are put into a variable, this variable is then used in each row to calculate the percentage. The reportItems _Cn_ refer to the column in the report. proc report data=sashelp.shoes nowd; where region in ('Africa', 'Asia','Canada'); column Product Region, sales region=region2, pctValue _dummy; define product/group; define region/across; define sales/'sales in USD'; define region2 / across; define pctValue / computed format=percent9.2; define _dummy / computed noprint; rbreak after/summarize; compute before; salesTotal1 = _c2_; salesTotal2 = _c3_; salesTotal3 = _c4_; endComp; compute _dummy / char length=32; _c5_ = _c2_ / salesTotal1; _c6_ = _c3_ / salesTotal1; _c7_ = _c4_ / salesTotal1; endcomp; run; Bruno

BrunoMueller · ‎02-10-2017

Hi Proc TABULATE can calculate the percentage values in a simple way, see code example below. With Proc REPORT, there is more coding involved to do this. Does it have to be Proc REPORT? In SAS Enterprise Guide, you can use the Summary Tables task an use point and click to put together the report. proc tabulate data=sashelp.shoes format=comma14.; where region in ('Africa', 'Asia','Canada'); class product region; var sales; table product all , region * sales region * sales * colpctsum ; table product all , (region all) * (sales*( sum colpctsum)) ; run; Bruno

BrunoMueller · ‎02-10-2017

Hi Try an expression like this: (int((age - 1) / 10) + 1) * 10 To test out your expression you can use a DATA Step like this: data want; do age = 1 to 100; age2 = round(age, 10); age3 = (int((age - 1) / 10) + 1) * 10; output; end; run; Bruno

BrunoMueller · ‎02-10-2017

Hi You can take this code data have; length filename $ 1024; input filename ; cards; c:\temp\anchor.pdf c:\temp\Book1.csv ; data a; set have; rcs = filename("fileref", filename); msgs = sysmsg(); fid=fopen('fileref'); Bytes=finfo(fid,'File Size (bytes)'); crdate=finfo(fid,'Create Time'); moddate=finfo(fid,'Last Modified'); rc = fclose(fid); rcc = filename("fileref"); run; proc print; run;

BrunoMueller · ‎02-10-2017

Hi Have a look here http://support.sas.com/kb/40/934.html Bruno

BrunoMueller · ‎02-09-2017

hi As re Reeza suggested I would go for Proc SGPLOT, find below some sample code that can get you started. data have; infile cards dlm=","; input x y; cards; 0.249377,5.175325 0.623441,10.3494 1.122195,15.30481 1.745636,20.17191 2.535328,25.06026 3.532835,30.11836 4.655029,35.0611 5.985037,40.02055 7.564422,45.03658 9.476309,50.06652 11.7207,55.02452 14.42228,60.06016 17.58105,65.04058 21.28013,70.03401 25.68579,75.03019 30.96426,80.03419 37.73899,85.0182 47.00748,90.01258 61.01413,95.01025 100,100 ; data have2; set have; if int(y) = 50 then do; x2 = x; y2 = y; group=1; end; if int(y) = 75 then do; x2 = x; y2 = y; group=2; end; if 50 <= int(y) <= 75 then do; x3 = x; y3 = y; end; run; proc sgplot data=have2 noautolegend; styleattrs datasymbols=(DiamondFilled) datacontrastcolors=(red green) ; series x=x y=y ; scatter x=x2 y=y2 / group=group markerattrs=(size=20) ; run; proc sgplot data=have2 noautolegend; series x=x y=y /; series x=x3 y=y3 / lineattrs=(THICKNESS=5); refline 50 75; run; Bruno

BrunoMueller · ‎02-08-2017

Hi How big are your numbers, the length of 12 for the comma format might be too short. Try with 16. The length specified in the format (comma16.) must be wide enough for the number plus the additional formatting chars like thousand seperator, decimal separator, sign, etc. Bruno

BrunoMueller · ‎02-07-2017

Hi Just a very wild guess, have a look at READ_ISOLATION_LEVEL option http://support.sas.com/documentation/cdl/en/acreldb/69580/HTML/default/viewer.htm#n0xth6czn6guxrn13z0b2vyer8dt.htm Bruno

BrunoMueller · ‎02-07-2017

Hi When you use an variable for ACROSS, then by default the number of rows will be printed. If you want to nest an analysis variable within the across use the syntax as follows: acrossVar, analysisVar or analysisVar, acrossVar what ever suits you better. See code sample below: data have; infile cards dlm=","; input Product : $32. Date : anydtdte. NetSales : 8. ; quarter = cats("Q", put(date, qtr.)); format date date9.; cards; Apples,20160222,400.00 Grapes,20160512,450.20 Oranges,20160812,250.30 Peaches,20160915,500.00 Plums,20161225,350.80 Bananas,20160119,160.60 ; proc report data=have spanrows missing nowd; column product quarter netsales; define product/group; define quarter/across; define netsales/sum f=comma12.; rbreak after/summarize ul ol; run; proc report data=have spanrows missing nowd; column product quarter, netsales netsales=netSales2; define product/group; define quarter/across; define netsales / sum f=comma12.; define netsales2 /sum f=comma12.; rbreak after/summarize ul ol; run; Bruno

BrunoMueller · ‎02-03-2017

Hi As an alternate method you could make use of the "COMPUTE AFTER" block. Using this method the text is written as part of the table, and is therefor part of the table measuring, see sample code below title; ods rtf file="C:\temp\rtf111.rtf"; ods escapechar="^"; proc report data=sashelp.cars (obs=36) nocenter; column rowid Make Model; define rowid / computed; compute rowid; _rowId + 1; rowId = _rowId; endcomp; compute after / style={just=left}; length msgline $ 1024; msgline = "line1 ^n line2 ^n line3 ^n line 4 ^n line 5 ^n line 6"; l = length(msgline); line msgLine $varying1024. l; endcomp; run; /*ods text="^S={leftmargin=1in}line1 ^n line2 ^n line3 ^n line 4 ^n line 5 ^n line 6";*/ ods rtf close; Bruno

BrunoMueller · ‎02-02-2017

Hi You can make use of the SPDE engine but store the data in HDFS. An example LIBNAME would look like this libname spd_hdfs spde '/user/yourUser' hdfshost=default ; Check the doc for more options. Bruno

Online Status	Offline
Date Last Visited	a week ago