Hi,   Check - The filename case of your SAS programs containg/storing the stored process logic - Linux is case sensetive unlike Window. make sure to store all filenames in lower case on Linux.   - The directory/file permissions of your SAS programs containg/storing the stored process logic - Depending how you running your Stored Process on Linux, Ensure the Account ID interacting with your stored process, has access (Directory: r-w, File:r-- ) to all the stored process sas programs.   Hope this helps, Ahmed  ... View more

Hi Kroken,   Here is my personal opinion, and I'm by no means advanced user of this tools, but I wish I was.   I can not disagree with you with regards to, the lack of functionalities, but then again, which tool/software doesn't suffer from this? and that's why, R&D and software developers keep comping up with updates and enhancements in every new release. That's just the nature of software applications. This is by no means a justification, it's a fact, and that's why you see customers favor one Vendor over another, because of what the tools can offer out of the box.   When it comes to Report Design & Development, All BI tools vendors will agree on, having pre-aggregated data saves resources and speeds delivery of results and visuals to the end user. Their tools might have the ability to calculate these custom aggregates on the fly within the report, via custom calculations and what have you, but when start overloading/repeating these custom calculations over multiple reports and data sources, the performance starts taking a hit, and users start to notice delayes in response times and that leads to frustrations, despite all the Query Caching, and Parallel query capability thay may have put inplace!   When you watch or listen to their experts taking about techiques to enhance query/report performance, they all will advice on pushing the processing down to the database or utilize pre-aggregated data to minimize run-time processing. The other thing to consider, most / if not all of these BI Tools, provide a data preparation tool of some sort for that purpose.   As a report designer, you would probably start with a scetch of the report on a peice of paper, then you start putting/building the visual objects together and assembling the final report. During that process, you would discover  - Report Objective / Focus Area  - Exact data requirements (details, aggregates, ratios, differences, ....etc) - Visual Objects (native, custom built) - Annotations - Links   All I can say, don't get discouraged, no report happens in just few seconds, even the one you see in demos, they are all rehersed 😉   Check this blog for reference: How to design an infographic about volcanic eruptions using SAS Visual Analytics For additional related blogs, Just go to http://blogs.sas.com/content/ and serch for "Visual Analytics"   Hope this helps, Ahmed ... View more

Hey Kroken,   Her is an alternative approach for your consideration: I fully understand the benifit of loading data in Memory and what have you, but sometimes, working with purpose built data, can yield lots more benifits without demanding tones of resources.   Build your base percent frequency aggregated table first, then loaded into the LASR server, from that point, your calculated columns should be more straight forward.   Here is sample code to build the Base percent frequency Aggregated table:   DATA have;     LENGTH    Date 8                     ID 4                     Age_group $5;     INPUT Date :yymmdd10. ID Age_group;     FORMAT    date yymmdd10.; DATALINES; 2016-01-01 1 0-10 2016-01-01 2 0-10 2016-01-01 3 0-10 2016-01-01 4 11-20 2016-01-01 5 11-20 2016-01-01 6 21-30 2016-01-01 7 21-30 2016-01-01 8 31- 2016-01-01 9 31- 2016-01-01 10 31- 2016-06-01 1 0-10 2016-06-01 2 0-10 2016-06-01 3 11-20 2016-06-01 4 11-20 2016-06-01 5 11-20 2016-06-01 6 21-30 2016-06-01 7 21-30 2016-06-01 8 21-30 2016-06-01 9 31- 2016-06-01 10 31- 2017-01-01 1 0-10 2017-01-01 2 0-10 2017-01-01 3 11-20 2017-01-01 4 11-20 2017-01-01 5 21-30 2017-01-01 6 21-30 2017-01-01 7 21-30 2017-01-01 8 21-30 2017-01-01 9 31- 2017-01-01 10 31- ; RUN; PROC TABULATE DATA=WORK.HAVE OUT=WORK.summary;          CLASS Date /    ORDER=UNFORMATTED MISSING;     CLASS Age_group /    ORDER=UNFORMATTED MISSING;     TABLE         /* Row Dimension */         Age_group={LABEL=''},         /* Column Dimension */         Date={LABEL=''}*ColPctN={LABEL=''}*F=BEST4.         /*Table Options */         / BOX={LABEL='Age Group'} ;         ; RUN; PROC TRANSPOSE DATA=WORK.SUMMARY OUT=WORK.WANT(DROP=_NAME_);     BY    Age_group;     ID        Date;     IDLABEL    Date;     VAR    PctN_10; RUN;     Hope this helps, Ahmed ... View more

Hi,   I think this new version of the code, will give you what you are looking for,     DATA have;     Length         state $2         city $10         fmtname $7         start 8         label $5         hour 4         minute 4;     INPUT state city fmtname start :date9. label hour minute;     FORMAT start date9. label $5.; DATALINES; MD Baltimore sunrise 17Jan2017 07:24 7 24 MD Baltimore sunset 17Jan2017 17:10 17 10 DE Dover sunrise 13JUL2017 04:47 4 47 DE Dover sunset 13JUL2017 19:28 19 28 ; RUN; /* Sort the Data in preparation for consalidation into single line per state/city/date */ PROC SORT data=have out=have_srtd;     BY state city start label hour; RUN; /* --------------------------------------------------------- */ /* The step below will:                                      */ /* - Consolidate data into single record per state/city/date */ /* - Collect numbers & calculate factor for sunrise & sunset */ /* - Output a record for sunrise and record for sunset.      */ /* --------------------------------------------------------- */ DATA want(KEEP=state city fmtname start label hour_:);     ARRAY init_hrs [24] _temporary_ (1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1);     LENGTH    state $2             city $10             fmtname $7             start 8             label $5             hour 4             minute 4;     LENGTH     sunrise $5             sunset $5             rise_hr 4             set_hr 4             rise_min 4             set_min 4             rise_factor 8             set_factor 8;     ARRAY hours {24} 4 hour_0-hour_23;     FORMAT hour_: BEST8.4;     RETAIN init_rb8_str hr1_addr sunrise rise_hr set_hr rise_min rise_factor;     IF (_n_ = 1) THEN     DO;         /* Get the String representation of the array elements */         init_rb8_str = put (peekclong (addrlong(init_hrs[1]), 192), $192.);         /* Get the memory address of the first element of the hours array */         hr1_addr = addrlong(hours[1]);     END;     /* - Consolidate data into single record per state/city/date */     /* - Collect numbers & calculate factor for sunrise & sunset */     DO UNTIL (last.start);         SET have_srtd;         BY state city start label hour;         if (lowcase(fmtname)='sunrise') then         do;             /* Process SunRise Time */             sunrise = label;             rise_hr  = hour;             rise_min = minute;             rise_factor = rise_min/60;         end;         else         do;             /* Process SunSet Time */             sunset = label;             set_hr  = hour;             set_min = minute;             set_factor = set_min/60;         end;     END;     /* ---------------------------------------------------- */     /* - Output a record for sunrise and record for sunset. */     /* ---------------------------------------------------- */     * --- Output the SunRise Record --- ;     fmtname = 'sunrise';     label   = sunrise;     /* Set all Hours to 1 in a single command */     CALL POKELONG (init_rb8_str, hr1_addr, 192);     /* Assign the SunRise factor hours */     do i=(rise_hr+2) to (set_hr+1);             hours[i]=rise_factor;     end;     OUTPUT;     * --- Output the SunSet Record --- ;     fmtname = 'sunset';     label   = sunset;     /* Reset all Hours to 1 in a single command */     CALL POKELONG (init_rb8_str, hr1_addr, 192);     /* Deals with the hours prior to sunRise */     do i=1 to (rise_hr+1);         hours[i]=set_factor;     end;     /* Deals with the hours after sunSet */     do i=(set_hr+2) to dim(hours);         hours[i]=set_factor;     end;     OUTPUT; RUN;     Obs state city fmtname start label hour_0 hour_1 hour_2 hour_3 hour_4 hour_5 hour_6 hour_7 hour_8 hour_9 hour_10 hour_11 hour_12 hour_13 hour_14 hour_15 hour_16 hour_17 hour_18 hour_19 hour_20 hour_21 hour_22 hour_23 1 DE Dover sunrise 13JUL2017 04:47 1 1 1 1 1 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 0.783333 1 1 1 1 2 DE Dover sunset 13JUL2017 19:28 0.466666 0.466666 0.466666 0.466666 0.466666 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0.466666 0.466666 0.466666 0.466666 3 MD Baltimore sunrise 17JAN2017 07:24 1 1 1 1 1 1 1 1 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 1 1 1 1 1 1 4 MD Baltimore sunset 17JAN2017 17:10 0.166667 0.166667 0.166667 0.166667 0.166667 0.166667 0.166667 0.166667 1 1 1 1 1 1 1 1 1 1 0.166667 0.166667 0.166667 0.166667 0.166667 0.166667 ... View more

Hi,   If you are using SAS/ACCESS to ODBC to interact with MS SQL Server on Azure, check out the BULKLOAD= option for you ODBC libname statemant.   and while you are at it, I would recommend enabling SAS/ACCESS tracing/debugging via the following options   /*Turn on SAS tracing */ options sastrace=”,,t,” sastraceloc=saslog nostsuffix;   For additional details and usage examples, Check this SGF paper "SASTRACE: Your Key to RDBMS Empowerment"   Hope this helps, Ahmed       ... View more

You can run the following Proc product_status; run; Proc product_status: Is an undocumented (9.2) procedure that lists the the release level (or version number) of each installed component of SAS in the system in the SAS log ... View more