Hey @devi001! While you cannot password-protect your raw code file since it is just text, you can store and compile DATA Step programs with the Source option to save the source code with it. If you use the source=encrypt and pw options, you can save and encrypt the source code, then retrieve it later. For more information, see: Creating A Stored Compiled DATA Step Program. The next best option would be to save all of your programs into an encrypted ZIP file.   For example, this will create a DATA Step program and save it to sasuser as foo_compiled.sas7bpgm: data foo / pgm=sasuser.foo_compiled(source=encrypt pw=mypw); bar = 'Hello, World!'; run;   If you attempt to view the program source without a password, you will get an error: data pgm=sasuser.foo_compiled; describe; run; ERROR: Invalid or missing ALTER password on member SASUSER.FOO_COMPILED.PROGRAM. ERROR: Unable to load DATA STEP program SASUSER.FOO_COMPILED.   Supplying the password will reveal the source code in the log: data pgm=sasuser.foo_compiled(pw=mypw); describe; run; NOTE: DATA step stored program SASUSER.FOO_COMPILED is defined as: data foo / pgm=sasuser.foo_compiled(source=encrypt pw=XXXX); bar = 'Hello, World!'; run;      To call a stored compiled program and run it, simply omit the describe statement. data pgm=sasuser.foo_compiled(pw=mypw); run;   ... View more

Hey @Rian0126! If your client has a local Esri or OpenStreetMap server, you can specify them in Environment Manager for their basemaps. In addition, you can pull in Esri feature services for rendering polygons, or you can pull shapefiles into CAS and render them in Visual Analytics.   For more information on configuring local Esri or custom OpenStreetMap servers, see: Configuration Properties - Maps Service. The customOSM.servers and localEsriServicesUrl properties are the configuration options of interest. ... View more

Hey @Batman! This is one of those situations where efficiency is probably going to be minsiscule depending on which way you go. You can use whichn as @PaigeMiller showed, or you can add them to gether and only return rows where the sum is > 0.  So, let's go ahead and test this. We'll create a decent-sized dataset and compare methods so we can gauge how it scales.   This dataset will have 500k rows with exactly 100 rows where all values of x are 1. data have; array x[100]; do i = 1 to 500000; do j = 1 to dim(x); if(50000 LE i LE 50100) then x[j]=1; else x[j] = 0; end; output; end; drop i j; run; /* Helper function so we don't need to type x1, x2, x3, ... etc. */ %macro delimx(d); %local i; x1 %do i = 2 %to 100; &d x&i %end; %mend; %macro timeit; %do i = 1 %to 100; /* Test 1: Whichn */ %let start=%sysfunc(datetime()); proc sql noprint; select * from have where whichn(1, %delimx(%str(,))) ; quit; %let time1 = %sysevalf(%sysfunc(datetime())-&start); /* Test 2: Sum function */ %let start=%sysfunc(datetime()); proc sql noprint; select * from have where sum(%delimx(%str(,))) > 0 ; quit; %let time2 = %sysevalf(%sysfunc(datetime())-&start); /* Test 3: Addition */ %let start=%sysfunc(datetime()); proc sql noprint; select * from have where %delimx(+) > 0 ; quit; %let time3 = %sysevalf(%sysfunc(datetime())-&start); data time; time1 = &time1; time2 = &time2; time3 = &time3; run; proc append base=times data=time; run; %end; proc sql; select mean(time1) as avg_time1 label='Avg: Whichn' , std(time1) as std_time1 label='Std: Whichn' , mean(time2) as avg_time2 label='Avg: Sum' , std(time2) as std_time2 label='Std: Sum' , mean(time3) as avg_time3 label='Avg: Add' , std(time3) as std_time3 label='Std: Add' from times; quit; proc datasets lib=work nolist; delete times; quit; %mend; %timeit;   On my machine, we can see that there is nearly no difference between using whichn and the sum function across 100 trials in a SAS dataset, but whichn had a higher standard deviation. Straight addition was the slowest. So, what's the most efficient way of these three when working with SAS datasets? Either whichn or the sum function - but depending on the size of your dataset, you may be looking at milliseconds of difference that you'll never notice. ... View more

Hey @sasgorilla! If you need to analyze it by year and month, you'll want to align the dates to the first of the month and then merge them. To keep it easy to understand from a visual perspective, you can use the MONYY format to display them. For simplicity, let's assume there is only one day per month   data a2; set a; month = intnx('month', date, 0, 'B'); format yearmonth monyy.; run; data b2; set b; month = input(put(yearmonth, $4.), yymmn4.); format yearmonth2 monyy.; run;   This gets you two datasets whose dates are aligned to the first of the month but formatted to display as a year/month: A2 event date month 1 2020-01-01 JAN20 2 2018-07-06 JUL18 3 2015-02-15 FEB15 B2 var yearmonth month 1 1801 JAN18 2 1911 NOV19 3 2202 FEB22 If there are multiple values per month then you will need to aggregate them so that there is one value per month using SQL, PROC TIMESERIES, etc. ... View more

Hey @10134458! Here's how you can do it: Look up the county ID from mapsgfk.us_counties_attr and attach it to your dataset Add a hover-over HTML variable in the format needed by PROC GMAP Create your PROC GMAP code and use the HTML option in the CHORO statement. You can get the mapsgfk county IDs by merging your county data with it using DATA Step merge or SQL join, but in this case I'll choose to use a hash lookup.   At the bottom of this post is your data in CSV format.   data hospital_count2; length id $10.; set hospital_count; /* Create a lookup table for the mapsgfk County ID for Ohio counties */ if(_N_ = 1) then do; dcl hash lookup(dataset: "mapsgfk.us_counties_attr(where=(statecode='OH'))"); lookup.defineKey('county'); lookup.defineData('id'); lookup.defineDone(); end; /* Get the mapsgfk county ID based on the state's county ID */ rc = lookup.Find(); /* Create a hover-over of info with a line break, '0D'x */ hover_over=cat('title="County: ', strip(county_name), '0D'x, 'Count: ', count, '"'); drop rc; run; proc gmap data = hospital_count2 map = mapsgfk.us_counties; id id; choro count / html=hover_over; run;   Data:   county,county_name,Count 1,Adams,2 3,Allen,5 5,Ashland,1 7,Ashtabula,4 9,Athens,2 11,Auglaize,1 13,Belmont,3 15,Brown,1 17,Butler,5 19,Carroll,0 21,Champaign,1 23,Clark,3 25,Clermont,1 27,Clinton,1 29,Columbiana,2 31,Coshocton,1 33,Crawford,2 35,Cuyahoga,23 37,Darke,1 39,Defiance,3 41,Delaware,2 43,Erie,1 45,Fairfield,2 47,Fayette,2 49,Franklin,21 51,Fulton,1 53,Gallia,1 55,Geauga,1 57,Greene,2 59,Guernsey,2 61,Hamilton,18 63,Hancock,2 65,Hardin,1 67,Harrison,1 69,Henry,1 71,Highland,2 73,Hocking,1 75,Holmes,1 77,Huron,2 79,Jackson,1 81,Jefferson,2 83,Knox,1 85,Lake,2 87,Lawrence,1 89,Licking,1 91,Logan,1 93,Lorain,8 95,Lucas,16 97,Madison,1 99,Mahoning,6 101,Marion,1 103,Medina,2 105,Meigs,0 107,Mercer,1 109,Miami,2 111,Monroe,0 113,Montgomery,11 115,Morgan,0 117,Morrow,1 119,Muskingum,2 121,Noble,0 123,Ottawa,1 125,Paulding,1 127,Perry,0 129,Pickaway,2 131,Pike,1 133,Portage,1 135,Preble,0 137,Putnam,0 139,Richland,3 141,Ross,1 143,Sandusky,2 145,Scioto,2 147,Seneca,2 149,Shelby,1 151,Stark,6 153,Summit,11 155,Trumbull,5 157,Tuscarawas,2 159,Union,1 161,Van Wert,2 163,Vinton,0 165,Warren,2 167,Washington,2 169,Wayne,2 171,Williams,2 173,Wood,1 175,Wyandot,1       ... View more

The DATA Step here is trying to load a SAS dataset:   DATA OUT1.NEW_MULTI_CCS (DROP = i); SET IN1.&CORE; ... RUN; &CORE is defined as "~//PATH/officerlong", which is not a valid SAS dataset name. This looks like it is a path to a SAS dataset, which the library IN1 already is pointing to. You need to specify an unquoted SAS dataset name that is within the library IN1.   For example, suppose your SAS dataset was called officerlong and it is within the IN1 library. You should make the following change to the program: %LET CORE = officerlong;   ... View more

The reason for this is because I used the Sales variable within the key value rather than the calculated items, "Current Year Sales" and "Prior Year Sales." Since the key value is using a paramaterized filter with a lattice group of "Year", we do not need to use the calculated items to identify the selected year and the prior year. ... View more

One way to do this is by converting your year/month into a SAS date and run it through PROC TIMESERIES to fill in the end of the series.   data temp2; set temp end=eof; month2 = mdy(month, 1, year); if(eof) then call symputx('end', put(mdy(12, 1, year), date9.)); format month2 monyy.; run; proc timeseries data=temp2 out=temp3; by col1 col2 col3; id month2 interval=month end="&end"d; var value / setmissing=0; run; data want; set temp3; year = year(month2); month = month(month2); drop month2; run;   ... View more

Hey @shubha1babu! You can achieve this using a combination of a parameter and an expression. Since SAS dates are the number of days since Jan 1st 1960, you can subtract two dates from each other to get the number of days between them. In Visual Analytics, this works the exact same way. Let's get started.   Create a new text input control and add it to your report Create a new character parameter and name it "Date (DD/MM/YYYY)". This will give users a nice queue on what they should enter in the text box. You'll see how this comes into play in the next step. Assign the character parameter to the text input control. Note it says "Enter Date (DD/MM/YYYY)". Create a new expression and call it "Date Diff", then assign it a format of Float with 0 decimal places.   Parse('Date (DD/MM/YYYY)'p, 'DDMMYY10') - 'Date'n This expression says to read the value the user entered as a date in DDMMYY10 format, then subtract it from the date in the table Add it to your table and you're done! If for some reason you run into a problem subtracting the date, try using "TreatAs(_number_, 'Date'n)" instead.   ... View more

Can you change its format to Float12.2, then post a screenshot of the results? Visual Analytics uses the BEST12. format by default which may not necessarily include decimals.     ... View more