About rogerjdeangelis

rogerjdeangelis · ‎02-27-2017

It worked for me, see log below. The merge and the DOW (two set statements) had identical times. I ran 100,000,000. My $600 workstation is too fast for 10,000,000. data claims; input claim_number $ loss_code $; cards4; CML0001 27 CML0002 27C CML0002 27 CML0003 27Y CML0003 27C CML0003 27P CML0004 27O CML0004 27Y CML0004 27 CML0004 27P CML0005 27P CML0005 27O CML0005 27C CML0005 27 CML0006 27Y CML0006 27C CML0007 27O CML0007 27Y CML0008 27O CML0009 27P ;;;; run; data want(keep=claim_number); retain gotone 0; do until (last.claim_number); set claims; by claim_number; if loss_code='27' then gotone=1; end; do until (last.claim_number); set claims; by claim_number; if last.claim_number and gotone =0 then output; end; gotone=0; run;quit; proc print data=want; run;quit; 456 data claims; 457 input claim_number $ loss_code $; 458 cards4; NOTE: The data set WORK.CLAIMS has 20 observations and 2 variables. NOTE: DATA statement used (Total process time): real time 0.01 seconds cpu time 0.00 seconds 479 ;;;; 480 run; 481 data want(keep=claim_number); 482 retain gotone 0; 483 do until (last.claim_number); 484 set claims; 485 by claim_number; 486 if loss_code='27' then gotone=1; 487 end; 488 do until (last.claim_number); 489 set claims; 490 by claim_number; 491 if last.claim_number and gotone =0 then output; 492 end; 493 gotone=0; 494 run; NOTE: There were 20 observations read from the data set WORK.CLAIMS. NOTE: There were 20 observations read from the data set WORK.CLAIMS. NOTE: The data set WORK.WANT has 5 observations and 1 variables. NOTE: DATA statement used (Total process time): real time 0.00 seconds cpu time 0.01 seconds 494 ! quit; 495 proc print data=want; 496 run; NOTE: There were 5 observations read from the data set WORK.WANT. NOTE: PROCEDURE PRINT used (Total process time): real time 0.02 seconds cpu time 0.03 seconds 496 ! quit; Up to 40 obs from WORK.WANT total obs=5 CLAIM_ Obs NUMBER 1 CML0003 2 CML0006 3 CML0007 4 CML0008 5 CML0009 Benchmark data claims100mm; set claims; do rec=1 to 5000000; output; end; run;quit; data want; merge claims100mm(keep=claim_number) claims100mm(in=in_loss where=(loss_code = '27')); by claim_number; if not in_loss; if first.claim_number; keep claim_number; run; 542 data want; 543 merge claims100mm(keep=claim_number) 544 claims100mm(in=in_loss where=(loss_code = '27')); 545 by claim_number; 546 if not in_loss; 547 if first.claim_number; 548 keep claim_number; 549 run; NOTE: MERGE statement has more than one data set with repeats of BY values. NOTE: There were 100000000 observations read from the data set WORK.CLAIMS100MM. NOTE: There were 20000000 observations read from the data set WORK.CLAIMS100MM. WHERE loss_code='27'; NOTE: The data set WORK.WANT has 5 observations and 1 variables. NOTE: DATA statement used (Total process time): real time 27.63 seconds cpu time 27.62 seconds data want(keep=claim_number); retain gotone 0; do until (last.claim_number); set claims100mm; by claim_number; if loss_code='27' then gotone=1; end; do until (last.claim_number); set claims100mm; by claim_number; if last.claim_number and gotone =0 then output; end; gotone=0; run; NOTE: There were 100000000 observations read from the data set WORK.CLAIMS100MM. NOTE: There were 100000000 observations read from the data set WORK.CLAIMS100MM. NOTE: The data set WORK.WANT has 5 observations and 1 variables. NOTE: DATA statement used (Total process time): real time 27.08 seconds cpu time 27.00 seconds

rogerjdeangelis · ‎02-27-2017

May not exactly solve your problem, but you may be able to enhance it. Ian Whitlock - Create flatfile with layout HAVE (sashelp.class) ==================== Up to 40 obs from SASHELP.CLASS total obs=19 Obs NAME SEX AGE HEIGHT WEIGHT 1 Alfred M 14 69.0 112.5 2 Alice F 13 56.5 84.0 3 Barbara F 13 65.3 98.0 4 Carol F 14 62.8 102.5 5 Henry M 14 63.5 102.5 6 James M 12 57.3 83.0 7 Jane F 12 59.8 84.5 8 Janet F 15 62.5 112.5 9 Jeffrey M 13 62.5 84.0 10 John M 12 59.0 99.5 11 Joyce F 11 51.3 50.5 12 Judy F 14 64.3 90.0 13 Louise F 12 56.3 77.0 14 Mary F 15 66.5 112.0 15 Philip M 16 72.0 150.0 16 Robert M 12 64.8 128.0 17 Ronald M 15 67.0 133.0 18 Thomas M 11 57.5 85.0 19 William M 15 66.5 112.0 WANT flatfile with layout) =========================== Layout for SASHELP.CLASS to d:/txt/class.txt Column Column Name Type fmtlen --------------------------------- NAME char $8. +1 SEX char $1. +1 AGE num best10. +1 HEIGHT num best10. +1 WEIGHT num best10. +1 d:/txt/class.txt ----|----10---|----20---|----30---|----40---| Alfred M 14 69 112.5 Alice F 13 56.5 84 Barbara F 13 65.3 98 Carol F 14 62.8 102.5 Henry M 14 63.5 102.5 James M 12 57.3 83 Jane F 12 59.8 84.5 Janet F 15 62.5 112.5 Jeffrey M 13 62.5 84 John M 12 59 99.5 Joyce F 11 51.3 50.5 Judy F 14 64.3 90 Louise F 12 56.3 77 Mary F 15 66.5 112 Philip M 16 72 150 Robert M 12 64.8 128 Ronald M 15 67 133 Thomas M 11 57.5 85 William M 15 66.5 112 FULL SOLUTION ============= * Ian Whitlock; %macro utl_flatfile(lib=, /* libref of input sas dataset to convert */ sasdsn=, /* second-level sas dataset name */ file=print, /* name of file to write to */ tl=3, /* default is to generate a title3 stmt */ spaces=1 /* number of spaces separating output fields */ ); %local putlist ; %let lib=%upcase(&lib); %let sasdsn=%upcase(&sasdsn); proc sql noprint ; select case when format ^= " " then name || " " || format when upcase(type) = "CHAR" then name || " $char"||trim(left(put(length,3.)))||"." else name || " best10." end into :putlist separated by " +&spaces " from dictionary.columns where libname = "&lib" and memname = "&sasdsn" ; reset print ; title&tl "Layout for &lib..&sasdsn to &file" ; select name , type , case when format ^= " " then trim(format) when upcase(type) = "CHAR" then "$"||trim(left(put(length,3.)))||"." else "best10." end || " +&spaces " as fmtlen from dictionary.columns where libname = "&lib" and memname = "&sasdsn" ; title&tl ; quit; data _null_; set &lib..&sasdsn; file "&file" ; put &putlist ; run; %mend utl_flatfile; %utl_flatfile(lib=sashelp,sasdsn=class,file=d:/txt/class.txt); * check; data cls; infile "d:/txt/class.txt"; input name $8. +1 sex $1. +1 age best10. +1 height best10. +1 weight best10. +1 ; run;quit;

rogerjdeangelis · ‎02-27-2017

If you have a very large table this may be the fastest method. I think it reuses the cache created by the first 'set'? data want; retain gotone 0; do until (last.claim_number); set claims; by claim_number; if loss_code='27' then gotone=1; end; put gotone; do until (last.claim_number); set claims; by claim_number; if last.claim_number and gotone =0 then output; end; gotone=0; run;quit;

rogerjdeangelis · ‎02-27-2017

Remove empty rows and empty columns from excel inspired by https://goo.gl/ZlgFGn https://communities.sas.com/t5/Base-SAS-Programming/how-to-check-if-a-macro-variable-exists/td-p/206121 Slightly more complex with a mixture of numeric and character columns in excel. libname xel "d:/xls/colrow.xlsx"; * create some data; data xel.have; array cols[*] $8 Mary Joe Mike Jane Ted Tim; do rep=1 to 5; do var= 1 to dim(cols); jane=' '; if rep=3 then cols[var]=''; else cols[var]='A'!!put(int(100*uniform(-1)),z3.); end; drop rep var; output; end; run;quit; libname xel clear; HAVE (Excel sheet with missing column and missing row) Up to 40 obs WORK.HAVE total obs=5 Obs MARY JOE MIKE JANE TED TIM 1 A092 A041 A069 A069 A074 2 A010 A022 A069 A057 A068 3 4 A092 A006 A010 A009 A095 5 A009 A002 A068 A086 A094 WANT Up to 40 obs from want total obs=4 Obs MARY JOE MIKE TED TIM 1 A005 A082 A037 A067 A088 2 A019 A071 A026 A060 A043 3 A044 A077 A070 A041 A062 4 A086 A062 A008 A051 A081 WORKING CODE proc freq data=xel.have nlevels; set xel.have(drop=&colmis); if coalescec(of _character_)="" then delete; FULL SOLUTION ============= libname xel "d:/xls/colrow.xlsx"; * create excel workbook with sheet have and table have; data xel.have; array cols[*] $8 Mary Joe Mike Jane Ted Tim; do rep=1 to 5; do var= 1 to dim(cols); jane=' '; if rep=3 then cols[var]=''; else cols[var]='A'!!put(int(100*uniform(-1)),z3.); end; drop rep var; output; end; run;quit; ods output nlevels=nlevels; proc freq data=xel.have nlevels; run;quit; ods select all; /* Up to 40 obs from WORK.NLEVELS total obs=6 Obs TABLEVAR TABLEVARLABEL NLEVELS NMISSLEVELS NNONMISSLEVELS 1 MARY MARY 5 1 4 2 JOE JOE 5 1 4 3 MIKE MIKE 5 1 4 4 JANE JANE 1 1 0 5 TED TED 5 1 4 6 TIM TIM 5 1 4 */ %symdel colmis; * just incase it exists; data _null_; length colmis $1024; do until (dne); set nlevels end=dne; if nnonmisslevels=0 then colmis=catx(' ',colmis,tablevar); end; call symputx('colmis',colmis); rc=dosubl(' data want; set xel.have(drop=&colmis); if coalescec(of _character_)="" then delete; run;quit; '); run;quit; %put &=colmis; /* COLMIS=JANE Up to 40 obs from want total obs=4 Obs MARY JOE MIKE TED TIM 1 A005 A082 A037 A067 A088 2 A019 A071 A026 A060 A043 3 A044 A077 A070 A041 A062 4 A086 A062 A008 A051 A081 */

rogerjdeangelis · ‎02-26-2017

Looping and Concat col in QB This problem does not lend itself to mouse surfing? inspired by https://goo.gl/C5qmVX https://communities.sas.com/t5/SAS-Enterprise-Guide/Looping-and-Concat-col-in-QB/m-p/335958 This solution is a bit advanced and probably should not be used by inexperienced programmers. You do not need to know how many columns are in the data. HAVE ==== Up to 40 obs WORK.HAVE total obs=5 O COL1 COL2 COL3 COL4 COL5 COL6 COL7 COL8 COL9 COL10 FRUIT 1 A086 A090 A098 A002 A083 A042 A031 A090 A091 A088 APPLE 2 A055 A008 A014 A031 A074 A066 A084 A000 A024 A062 APPLE 3 A071 A004 A010 A038 A006 A044 A054 A033 A039 A085 APPLE 4 A057 A047 A098 A010 A024 A002 A041 A073 A025 A046 APPLE 5 A081 A098 A052 A099 A036 A081 A069 A048 A024 A040 APPLE WANT ==== COLALL FRUIT ------------------------------------------------------------------- A086,A090,A098,A002,A083,A042,A031,A090,A091,A088 APPLE A055,A008,A014,A031,A074,A066,A084,A000,A024,A062 APPLE A071,A004,A010,A038,A006,A044,A054,A033,A039,A085 APPLE A057,A047,A098,A010,A024,A002,A041,A073,A025,A046 APPLE A081,A098,A052,A099,A036,A081,A069,A048,A024,A040 APPLE WORKING CODE rc=dosubl(' proc sql; select catx(",",&colcat.) as colall length=44 ,fruit from have ;quit; '); FULL SOLUTION ============= * create some data; data have; array cols[10] $8 col1-col10; do rep=1 to 5; do var= 1 to 10; cols[var]='A'!!put(int(100*uniform(-1)),z3.); end; fruit='APPLE'; keep c: fruit; output; end; run;quit; * select concatenated cols and fruit; data null_; length getcol $1024; set have(obs=1); array cols[*] c:; do i=1 to dim(cols); getcol=catx(',',getcol,vname(cols[i])); end; call symputx('colcat',strip(getcol)); put getcol=; rc=dosubl(' proc sql; select catx(",",&colcat.) as colall length=60 ,fruit from have ;quit; '); stop; run;quit;

rogerjdeangelis · ‎02-26-2017

Pass a parameter in macro When this is resolved Internet/Catalog Customers your program tries to do a division, I almost always double quote my macro comparison arguments. HAVE ==== Up to 40 obs WORK.CUSTOMER total obs=5 Obs NAME SEX CUSTOMER_GROUP 1 Alfred M Store Customers 2 Alice F Store Customers 3 Barbara F Internet/Catalog Customers 4 Carol F Store Customers 5 Henry M Store Customers WANT ==== the least abudant group is INTERNET/CATALOG CUSTOMERS Obs NAME SEX AGE CUSTOMER_GROUP 3 Barbara F 13 Internet/Catalog Customers BUT I GET THE ERRO MPRINT(TEST): where customer_group="Internet/Catalog Customers"; ERROR: A character operand was found in the %EVAL function or %IF condition where a numeric operand is required. The condition was: &param=A members WITH THIS CODE ============== %macro test(param); proc print data=customer; where customer_group="&param"; %if &param=A members %then %do; title "the most abudant group is %upcase(&param)"; %end; %else %do; title "the least abudant group is %upcase(&param)"; %end; var name gender age customer_group; run; %mend test; options mprint mlogic; %let least=Internet/Catalog Customers; %test(&least); WORKING CODE ============ Change %if &param =A members %then %do; to %if "&param" ="A members" %then %do; Also she need a run;quit; FULL SOLUTION ============= data customer( keep=name sex age customer_group); set sashelp.class(obs=5); if name='Barbara' then customer_group ='Internet/Catalog Customers'; else customer_group='Store Customers'; run;quit; %macro test(param); proc print data=customer; where customer_group="&param"; %if "&param" ="A members" %then %do; title "the most abudant group is %upcase(&param)"; %end; %else %do; title "the least abudant group is %upcase(&param)"; %end; var name sex age customer_group; run;quit; run; %mend test; options mprint mlogic; %let least=Internet/Catalog Customers; %test(&least);

rogerjdeangelis · ‎02-26-2017

Dynamic Matrix Multiplication If you have IML you can cut and paste the code into IML/R interface or just use IML directly I simplified the problem so I could check the algorithm. Should be fairly easy to adapt this algorithm to IML a datastep ,or even an FCMP function. The hard part is breaking the problem down. I hope I did it correctly? inspired by https://goo.gl/XMGUKM https://communities.sas.com/t5/SAS-Enterprise-Guide/Dynamic-Matrix-Multiplication/m-p/335968 Problem Multiply Y12 ( only 9 X8) with Y2 (8X1 i.e only last column.) and then from the result select all columns except last which means (9 X8 matrix ) multiplied by only last column from Y2 except the last value (8 X1) and this gives result a vector (9X1) HAVE two SAS sdatasets Y1 and Y2 Up to 40 obs from sd1.y1 total obs=3 Obs C1 C2 C3 1 1 0 0 2 1 1 0 3 1 1 1 Up to 40 obs from sd1.y2 total obs=3 Obs C1 C2 C3 1 2 2 2 2 0 2 2 3 0 0 2 WANT Up to 40 obs from wantwps total obs=3 Obs V1 1 8 2 12 3 12 DETAILS 1. Matrix mutiply Y1 * Y2 =Sum(elementwise mutiplication of row of y1 times column of y2) Sometimes called an inner product Y1 Y2 Y12 1 0 0 2 2 2 2 2 2 1 1 0 X 0 2 2 = 2 4 4 1 1 1 0 0 2 2 4 6* Example 2 *6= [1 1 1] *2 = 1*2 + 1*2 + 1*2 =6 2. Reduce the size of Y12 by removing the last column and reduce the size of Y2 by selecting the rows 1 and 2 of the last column Reduced Reduced Y12 Y2 = 2 2 2 = 1*2 + 2*2 = 8 2 4 2 = 2*2 + 4*2 = 12 2 4 2*2 + 4*2 = 12 WORKING CODE ============ R 1. Compute y12 y12<-y1 %*% y2 Multiply reduced matrices 2. y12[,c(1,2)] %*% y2[-1,3] FULL SOLUTION ============= Multiply Y12 ( only 9 X8) with Y2 (8X1 i.e only last column.) and then from the result select all columns except last which means (9 X8 matrix ) multiplied by only last column from Y2 except the last value (8 X1) and this gives result a vector (9X1) * create some data options validvarname=upcase; libname sd1 "d:/sd1"; data sd1.y1 (keep=c:) sd1.y2 (keep=c:); input Year $ c1 c2 c3; if year='Y1' then output sd1.y1; else output sd1.y2; cards4; Y1 1 0 0 Y1 1 1 0 Y1 1 1 1 Y2 2 2 2 Y2 0 2 2 Y2 0 0 2 ;;;; run;quit; %utl_submit_wps64(' libname sd1 "d:/sd1"; options set=R_HOME "C:/Program Files/R/R-3.3.1"; libname wrk "%sysfunc(pathname(work))"; proc r; submit; library(haven); library(dplyr); y1<-as.matrix(read_sas("d:/sd1/y1.sas7bdat")); y2<-as.matrix(read_sas("d:/sd1/y2.sas7bdat")); y1; y2; y12<-y1 %nrstr(%%)*%y2; y12; y3x2<-y12[,c(1,2)]; y3x2; y2[-1,3]; want<-y3x2%nrstr(%%)*%y2[-1,3]; want; endsubmit; import r=want data=wrk.wantwps; run;quit; '); proc print data=wantwps; run;quit; The WPS System C1 C2 C3 [1,] 1 0 0 [2,] 1 1 0 Y1 [3,] 1 1 1 C1 C2 C3 [1,] 2 2 2 [2,] 0 2 2 Y2 [3,] 0 0 2 C1 C2 C3 [1,] 2 2 2 Y12 [2,] 2 4 4 [3,] 2 4 6 C1 C2 [1,] 2 2 Reduced Y12 [2,] 2 4 [3,] 2 4 [1] 2 2 Reduced Y2 [,1] [1,] 8 [2,] 12 [3,] 12

rogerjdeangelis · ‎02-24-2017

Number of sheets and sheet names in the order they appear in an excel workbook If you have IML interface to R you can cut and paste the code into IML. Note proc contents gives the sheetnames in alphabetic order. HAVE Excel workbook with two sheets (Tab1 is ZZZZ and Tab2 is AAAA) ==================================================================== SHEET(ZZZZ ist sheet) CLASS IN WORKBOOK D:/XLS/TWO_SEETS.XLSX +----------------------------------------------------------------+ | A | B | C | D | E | +----------------------------------------------------------------+ 1 | NAME | SEX | AGE | HEIGHT | WEIGHT | +------------+------------+------------+------------+------------+ 2 | ALFRED | M | 14 | 69 | 112.5 | +------------+------------+------------+------------+------------+ ... +------------+------------+------------+------------+------------+ N | WILLIAM | M | 15 | 66.5 | 112 | +------------+------------+------------+------------+------------+ [ZZZZ] SHEET(AAAA 2nd sheet) +----------------------------------------------------------------+ | A | B | C | D | E | +----------------------------------------------------------------+ 1 | NAME | SEX | AGE | HEIGHT | WEIGHT | +------------+------------+------------+------------+------------+ 2 | ALFRED | M | 14 | 69 | 112.5 | +------------+------------+------------+------------+------------+ ... +------------+------------+------------+------------+------------+ N | WILLIAM | M | 15 | 66.5 | 112 | +------------+------------+------------+------------+------------+ [AAAA] WANT Number of sheets in the order they appear in workbook =========================================================== Up to 40 obs WORK.WANT total obs=1 SHEETS_BY_ NUMBER_ Obs POSITION SHEETS 1 ZZZZ AAAA 2 WORKING CODE R strset<-paste0(getSheets(wb),collapse=" "); FULL SOLUTION ============= %utl_submit_r64(' source("c:/Program Files/R/R-3.3.2/etc/Rprofile.site",echo=T); library(XLConnect); wb <- loadWorkbook("d:/xls/two_sheets.xlsx",create = FALSE); strset<-paste0(getSheets(wb),collapse=" "); strset; write(strset,"d:/txt/sheets.txt"); '); > source("c:/Program Files/R/R-3.3.2/etc/Rprofile.site",echo=T); library(XLConnect); wb <- loadWorkbook("d:/xls/two_sheets.xlsx",create = FALSE); strset<-paste0(getShee ts(wb),collapse=" "); strset; write(strset,"d:/txt/sheets.txt"); [1] "zzzz aaaa" d:/txt/sheets.txt zzzz aaaa data want; infile "d:/txt/sheets.txt"; input; sheets_by_position=_infile_; number_sheets=countw(_infile_); run;quit; Up to 40 obs WORK.WANT total obs=1 SHEETS_BY_ NUMBER_ Obs POSITION SHEETS 1 zzzz aaaa 2

rogerjdeangelis · ‎02-24-2017

Difference is sales for store1-store2 and store3-store2 Nicely layed out question Other solutions could use lag or 'by not sorted' No need to transpose. My solution will stop processing if you have more then three 'stores' . inspired by https://goo.gl/OgQQNz https://communities.sas.com/t5/Base-SAS-Programming/Is-there-any-way-to-transpose-the-observation-into-variable/m-p/335664 HAVE ==== Up to 40 obs WORK.HAVE total obs=12 Obs DATE DEPARTMENT SALE 1 30NOV2016 WEST 175 2 31DEC2016 WEST 134 3 31JAN2017 WEST 124 4 30NOV2016 EAST 198 5 31DEC2016 EAST 182 6 31JAN2017 EAST 117 7 30NOV2016 SOUTH 200 8 31DEC2016 SOUTH 203 9 31JAN2017 SOUTH 205 10 30NOV2016 NORTH 199 11 31DEC2016 NORTH 194 12 31JAN2017 NORTH 106 WANT ==== Up to 40 obs from want total obs=4 SALE_ SALE_ DEPARTMENT DIFF1 DIFF2 WEST -41 -10 EAST -16 -65 SOUTH 3 2 NORTH -5 -88 DETAILS ======= ???? Obs DEPARTMENT DIF1 Sale1-Sale2 DIF2 Sale3-Sale2 1 WEST 41 175-134 -10 124-134 2 EAST 16 198-182 -65 117-182 3 SOUTH -3 200-203 2 205-203 4 NORTH 5 199-194 -88 106-194 WORKING CODE =========== select; when (mod(_n_,3)=2) vals[1]=sale; when (mod(_n_,3)=1) vals[2]=sale; when (mod(_n_,3)=0) do; dif1=vals[2]-vals[1]; dif2=sale-vals[1]; output; end; end; * leave off otherwise to force error; FULL SOLUTION ============= DATA have; INPUT DATE $ 1-9 DEPARTMENT $ SALE ; CARDS; 30NOV2016 WEST 175 31DEC2016 WEST 134 31JAN2017 WEST 124 30NOV2016 EAST 198 31DEC2016 EAST 182 31JAN2017 EAST 117 30NOV2016 SOUTH 200 31DEC2016 SOUTH 203 31JAN2017 SOUTH 205 30NOV2016 NORTH 199 31DEC2016 NORTH 194 31JAN2017 NORTH 106 ; RUN; data want; array vals[2] _temporary_; set have; select; when (mod(_n_,3)=2) vals[1]=sale; when (mod(_n_,3)=1) vals[2]=sale; when (mod(_n_,3)=0) do; dif1=vals[2]-vals[1]; dif2=sale-vals[1]; output; end; end; * leave off otherwise to force error; keep department dif:; run;quit; Up to 40 obs from want total obs=4 Obs DEPARTMENT DIF1 DIF2 1 WEST 41 -10 2 EAST 16 -65 3 SOUTH -3 2 4 NORTH 5 -88

rogerjdeangelis · ‎02-24-2017

Difference is sales for store1-store2 and store3-store2 Nicely layed out question Other solutions could use lag or 'by not sorted' No need to transpose. My solution will stop processing if you have more then three 'stores' . inspired by https://goo.gl/OgQQNz https://communities.sas.com/t5/Base-SAS-Programming/Is-there-any-way-to-transpose-the-observation-into-variable/m-p/335664 HAVE ==== Up to 40 obs WORK.HAVE total obs=12 Obs DATE DEPARTMENT SALE 1 30NOV2016 WEST 175 2 31DEC2016 WEST 134 3 31JAN2017 WEST 124 4 30NOV2016 EAST 198 5 31DEC2016 EAST 182 6 31JAN2017 EAST 117 7 30NOV2016 SOUTH 200 8 31DEC2016 SOUTH 203 9 31JAN2017 SOUTH 205 10 30NOV2016 NORTH 199 11 31DEC2016 NORTH 194 12 31JAN2017 NORTH 106 WANT ==== Up to 40 obs from want total obs=4 SALE_ SALE_ DEPARTMENT DIFF1 DIFF2 WEST -41 -10 EAST -16 -65 SOUTH 3 2 NORTH -5 -88 DETAILS ======= ???? Obs DEPARTMENT DIF1 Sale1-Sale2 DIF2 Sale3-Sale2 1 WEST 41 175-134 -10 124-134 2 EAST 16 198-182 -65 117-182 3 SOUTH -3 200-203 2 205-203 4 NORTH 5 199-194 -88 106-194 WORKING CODE =========== select; when (mod(_n_,3)=2) vals[1]=sale; when (mod(_n_,3)=1) vals[2]=sale; when (mod(_n_,3)=0) do; dif1=vals[2]-vals[1]; dif2=sale-vals[1]; output; end; end; * leave off otherwise to force error; FULL SOLUTION ============= DATA have; INPUT DATE $ 1-9 DEPARTMENT $ SALE ; CARDS; 30NOV2016 WEST 175 31DEC2016 WEST 134 31JAN2017 WEST 124 30NOV2016 EAST 198 31DEC2016 EAST 182 31JAN2017 EAST 117 30NOV2016 SOUTH 200 31DEC2016 SOUTH 203 31JAN2017 SOUTH 205 30NOV2016 NORTH 199 31DEC2016 NORTH 194 31JAN2017 NORTH 106 ; RUN; data want; array vals[2] _temporary_; set have; select; when (mod(_n_,3)=2) vals[1]=sale; when (mod(_n_,3)=1) vals[2]=sale; when (mod(_n_,3)=0) do; dif1=vals[2]-vals[1]; dif2=sale-vals[1]; output; end; end; * leave off otherwise to force error; keep department dif:; run;quit; Up to 40 obs from want total obs=4 Obs DEPARTMENT DIF1 DIF2 1 WEST 41 -10 2 EAST 16 -65 3 SOUTH -3 2 4 NORTH 5 -88

rogerjdeangelis · ‎02-24-2017

Strong Encryption sha256 is probably enough but if you are paranoid then do sha512 /* T1002220 SAS/WPS/R: De-identify PII(name) using irrevesible strong hashing SAS/WPS/R: De-identify PII(name) using irrevesible strong hashing Joining tables with de-identified names sha256 and sha512 are 'theoretically' secure You need SAS Secure for the SAS solution. For WPS/R you don't; HAVE Two sas datasets sashelp.class and sashelp.classfit need to de-identify name ========================================================= Up to 40 obs SD1.WPSCLASS total obs=4 Obs NAME SEX AGE HEIGHT WEIGHT 1 Roger F 14 62.8 102.5 2 Janet F 15 62.5 112.5 3 Judy F 14 64.3 90 4 Robert M 12 64.8 128 And Up to 40 obs SD1.WPSCLASSFIT total obs=9 Obs NAME SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 Louise F 12 56.3 77 76.48 67.96 2 James M 12 57.3 83 80.38 72.66 3 John M 12 59 99.5 87.01 80.47 4 Janet F 15 62.5 112.5 100.66 95.22 5 Carol F 14 62.8 102.5 101.83 96.37 6 Judy F 14 64.3 90 107.68 101.84 7 Barbara F 13 65.3 98 111.57 105.26 8 William M 15 66.5 112 116.25 109.18 9 Alfred M 14 69 112.5 126.00 116.94 WANT ( Join the two datasets but with de-identified name ) ========================================================== Up to 40 obs SD1.WPSCLASSFIT total obs=9 Obs NAME SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 Janet F 15 62.5 112.5 100.66 95.22 2 Judy F 14 64.3 90 107.68 101.84 WITH DE-IDENTIFIED 'NAME' using 256bit hash =========================================== Up to 40 obs from classwps total obs=4 Obs NAME SEX AGE HEIGHT WEIGHT 1 689ffebb9149498fb8189d6e3ad1625c8bc1e1c0f0849ca40f2df83eb2930c7b F 14 62.8 102.5 2 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc F 15 62.5 112.5 3 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 F 14 64.3 90 4 7d359106f754613d9354353f2a403b88b76fb1569ea81524f54d479f4c0d610f M 12 64.8 128 Obs NAME SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 b68d4a840d7e218045d98247e00d79ed7bb1eff3b1fff117dc93db32f4699bfe F 12 56.3 77 76.48 67.96 2 7007b18f0d9cb87601ba50c83e761006dc185270a4cf2d55812e0863be8ff566 M 12 57.3 83 80.38 72.66 3 632dc5c5235dc0b5b411f8f5e4a6f619ce4155ef99d3d8adfcd026288c795187 M 12 59 99.5 87.01 80.47 *4 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc F 15 62.5 112.5 100.66 95.22 5 f90982571d7ace45426530062c7008fc370be6a7faeb9ecb05733758960b0c39 F 14 62.8 102.5 101.83 96.37 *6 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 F 14 64.3 90 107.68 101.84 7 6615925a48704e803ee5e32ad0209bb348698ce575effe73da268d9e606c3208 F 13 65.3 98 111.57 105.26 8 2b008dfe3f79c53c59ab5392929c06c970d71b62c58527ef51d07a50616dbbd0 M 15 66.5 112 116.25 109.18 9 177d26a3ce597f633b7194aee7fa886111ab1f2edfd37854e0a179ea533154c6 M 14 69 112.5 126.00 116.94 OR Up to 40 obs WORK.NTRSEC total obs=2 Obs NAME SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc F 15 62.5 112.5 100.66 95.22 2 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 F 14 64.3 90 107.68 101.84 MAPPING THE HASH TO INTEGERS ============================ Up to 40 obs from CLASSWPSFIX total obs=4 NAME CRYPTO SEX AGE HEIGHT WEIGHT 1 F 14 62.8 102.5 4 M 12 64.8 128 *2 F 15 62.5 112.5 ** Match 1 *3 F 14 64.3 90 ** Match 2 Up to 40 obs from CLASSWPS total obs=9 NAME CRYPTO SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 5 F 12 56.3 77 76.48 67.96 6 M 12 57.3 83 80.38 72.66 7 M 12 59 99.5 87.01 80.47 *2 F 15 62.5 112.5 100.6 95.22 ** Match 1 9 F 14 62.8 102.5 101.8 96.37 *3 F 14 64.3 90 107.6 101.8 ** Match 2 11 F 13 65.3 98 111.5 105.2 12 M 15 66.5 112 116.2 109.1 13 M 14 69 112.5 126.0 116.9 AFTER MATCH ON NUM_CRYPTO NAME Obs CRYPTO SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 2 F 15 62.5 112.5 100.66 95.225 2 3 F 14 64.3 90 107.68 101.84 *____ _ ____ / ___| / \ / ___| \___ \ / _ \ \___ \ ___) / ___ \ ___) | |____/_/ \_\____/ ; libname sd1 "d:/sd1"; data sd1.sasclass; length name $32; format name $hex64.; set sashelp..class; name=sha256(name); if mod(_n_,4) = 0 then do; if _n_=4 then name=sha256('Roger'); output; end; run;quit; data sd1.sasclassfit; length name $32; format name $hex64.; set sashelp.classfit; name=sha256(name); if mod(_n_,2) = 0; run;quit; proc sql; create table ntrsec as select l.* ,r.predict ,r.lowermean from sd1.sasclass as l, sd1.sasclassfit as r where l.name = r.name ;quit; Predicted Bound of Value of 95% C.I. NAME SEX AGE HEIGHT WEIGHT Weight for Mean -------------------------------------------------------------------------------------------------------------------- D4B7A1817B10284BBC7135509F57659210CAEB70C06F91644CDB0D6B19054C47 F 15 62.5 112.5 100.6625 95.22579 E8EE6BCC3464E592F1DD07E0EE6D3FB93EC132AA4D867AA52818A641304475F5 F 14 64.3 90 107.6807 101.8416 #! WPS-R ; *_ ______ ____ ____ \ \ / / _ \/ ___| | _ \ \ \ /\ / /| |_) \___ \ _____| |_) | \ V V / | __/ ___) |_____| _ < \_/\_/ |_| |____/ |_| \_\ ; proc datasets lib=work kill; run;quit; * need at aleast one dataset in work for WPS to switch to sas7bdats instead of wpd datasets; data makesas; set sashelp.class; run;quit; options validvarname=upcase; libname sd1 "d:/sd1"; data sd1.wpsclass; set sashelp.class; if mod(_n_,4) = 0 then do; if _n_=4 then name='Roger'; output; end; run;quit; data sd1.wpsclassfit; set sashelp.classfit; if mod(_n_,2) = 0; run;quit; * I think there is an issue with binary data and WPS import; * so I switched to hex; %utl_submit_wps64(' libname sd1 "d:/sd1"; options set=R_HOME "C:/Program Files/R/R-3.3.2"; libname wrk "%sysfunc(pathname(work))"; libname hlp "C:\Program Files\SASHome\SASFoundation\9.4\core\sashelp"; proc r; submit; source("C:/Program Files/R/R-3.3.2/etc/Rprofile.site", echo=T); library(digest); library(haven); class <- read_sas("d:/sd1/wpsclass.sas7bdat"); shaclass <- as.character(lapply(class$NAME,function(x) paste0(digest(x, algo=c("sha256"),raw=TRUE),collapse=""))); classwps <- cbind(name_crypto=c(1:nrow(class)),SHACLASS=shaclass,class[,-1]); classfit <- read_sas("d:/sd1/wpsclassfit.sas7bdat"); shaclassfit <- as.character(lapply(classfit$NAME,function(x) paste0(digest(x, algo=c("sha256"),raw=TRUE),collapse=""))); classfitwps <- cbind(name_crypto=c((nrow(class)+1):(nrow(class)+nrow(classfit))),SHACLASS=shaclassfit,classfit[,-1]); endsubmit; import r=classwps data=wrk.classwps; import r=classfitwps data=wrk.classfitwps; run;quit; '); /* The WPS System name_crypto SHACLASS SEX AGE HEIGHT WEIGHT 1 1 689ffebb9149498fb8189d6e3ad1625c8bc1e1c0f0849ca40f2df83eb2930c7b F 14 62.8 102.5 2 2 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc F 15 62.5 112.5 3 3 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 F 14 64.3 90.0 4 4 7d359106f754613d9354353f2a403b88b76fb1569ea81524f54d479f4c0d610f M 12 64.8 128.0 [1] 5 6 7 8 9 10 11 12 13 name_crypto SHACLASS SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 5 b68d4a840d7e218045d98247e00d79ed7bb1eff3b1fff117dc93db32f4699bfe F 12 56.3 77.0 76.48849 67.96005 2 6 7007b18f0d9cb87601ba50c83e761006dc185270a4cf2d55812e0863be8ff566 M 12 57.3 83.0 80.38752 72.66709 3 7 632dc5c5235dc0b5b411f8f5e4a6f619ce4155ef99d3d8adfcd026288c795187 M 12 59.0 99.5 87.01587 80.47925 4 8 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc F 15 62.5 112.5 100.66247 95.22579 5 9 f90982571d7ace45426530062c7008fc370be6a7faeb9ecb05733758960b0c39 F 14 62.8 102.5 101.83218 96.37505 6 10 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 F 14 64.3 90.0 107.68073 101.84160 7 11 6615925a48704e803ee5e32ad0209bb348698ce575effe73da268d9e606c3208 F 13 65.3 98.0 111.57976 105.26025 8 12 2b008dfe3f79c53c59ab5392929c06c970d71b62c58527ef51d07a50616dbbd0 M 15 66.5 112.0 116.25859 109.18222 9 13 177d26a3ce597f633b7194aee7fa886111ab1f2edfd37854e0a179ea533154c6 M 14 69.0 112.5 126.00617 116.94168 */ proc sql; select l.* from classfitwps as l, classwps as r where l.shaclass = r.shaclass ;quit; SHACLASS SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN --------------------------------------------------------------------------------------------------------------------- 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc F 15 62.5 112.5 100.6625 95.22579 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 F 14 64.3 90 107.6807 101.8416 *__ __ _ ____ ___ _ _ _____ | \/ | / \ | _ \ |_ _| \ | |_ _| | |\/| | / _ \ | |_) |____| || \| | | | | | | |/ ___ \| __/_____| || |\ | | | |_| |_/_/ \_\_| |___|_| \_| |_| ; data class_classfit; set classwps classfitwps; run;quit; /* Up to 40 obs WORK.CLASS_CLASSFIT total obs=13 Obs NAME_CRYPTO SHACLASS 1 1 689ffebb9149498fb8189d6e3ad1625c8bc1e1c0f0849ca40f2df83eb2930c7b 2 2 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc ***** DUP 3 3 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 ** DUP 4 4 7d359106f754613d9354353f2a403b88b76fb1569ea81524f54d479f4c0d610f 5 5 b68d4a840d7e218045d98247e00d79ed7bb1eff3b1fff117dc93db32f4699bfe 6 6 7007b18f0d9cb87601ba50c83e761006dc185270a4cf2d55812e0863be8ff566 7 7 632dc5c5235dc0b5b411f8f5e4a6f619ce4155ef99d3d8adfcd026288c795187 8 8 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc ***** DUP 9 9 f90982571d7ace45426530062c7008fc370be6a7faeb9ecb05733758960b0c39 10 10 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 ** DUP 11 11 6615925a48704e803ee5e32ad0209bb348698ce575effe73da268d9e606c3208 12 12 2b008dfe3f79c53c59ab5392929c06c970d71b62c58527ef51d07a50616dbbd0 13 13 177d26a3ce597f633b7194aee7fa886111ab1f2edfd37854e0a179ea533154c6 */ * remove dups we need 1 to 1 proc sort data=class_classfit out=lookup(keep=name_crypto shaclass) nodupkey; by shaclass; run;quit; /* Up to 40 obs WORK.LOOKUP total obs=11 NAME Obs CRYPTO SHACLASS 1 13 177d26a3ce597f633b7194aee7fa886111ab1f2edfd37854e0a179ea533154c6 2 12 2b008dfe3f79c53c59ab5392929c06c970d71b62c58527ef51d07a50616dbbd0 3 7 632dc5c5235dc0b5b411f8f5e4a6f619ce4155ef99d3d8adfcd026288c795187 4 11 6615925a48704e803ee5e32ad0209bb348698ce575effe73da268d9e606c3208 5 1 689ffebb9149498fb8189d6e3ad1625c8bc1e1c0f0849ca40f2df83eb2930c7b 6 6 7007b18f0d9cb87601ba50c83e761006dc185270a4cf2d55812e0863be8ff566 7 4 7d359106f754613d9354353f2a403b88b76fb1569ea81524f54d479f4c0d610f 8 2 909428873964ece00ba2d2fbca40035e7685a04687821c9fc7ea4cbb047b70fc 9 5 b68d4a840d7e218045d98247e00d79ed7bb1eff3b1fff117dc93db32f4699bfe 10 3 f43da163b5f48c69f909facc9a5880c5666c5ccffa4837890bb4519542f03e45 11 9 f90982571d7ace45426530062c7008fc370be6a7faeb9ecb05733758960b0c39 */ * add mapped integers to each dataset; data _null_; do dsn="classfitwps","classwps"; call symputx('dsn',dsn); rc=dosubl(' proc sql; create table &dsn.fix(drop=shaclass) as select l.name_crypto ,r.* from lookup as l, &dsn as r where l.shaclass = r.shaclass ;quit; '); end; run;quit; * change sha256 to an unique integer for each datset ; Up to 40 obs from CLASSWPSFIX total obs=4 NAME Obs CRYPTO SEX AGE HEIGHT WEIGHT 1 1 F 14 62.8 102.5 2 4 M 12 64.8 128 3 2 F 15 62.5 112.5 4 3 F 14 64.3 90 NAME Obs CRYPTO SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 5 F 12 56.3 77 76.48 67.960 2 6 M 12 57.3 83 80.38 72.667 3 7 M 12 59 99.5 87.01 80.4 4 2 F 15 62.5 112.5 100.6 95.225 5 9 F 14 62.8 102.5 101.8 96.375 6 3 F 14 64.3 90 107.6 101.84 7 11 F 13 65.3 98 111.5 105.26 8 12 M 15 66.5 112 116.2 109.18 9 13 M 14 69 112.5 126.0 116.94 proc sql; create table ntrsec as select l.* ,r.predict ,r.lowermean from classwpsfix as l, classfitwpsfix as r where l.name_crypto = r.name_crypto ;quit; Up to 40 obs WORK.NTRSEC total obs=2 NAME Obs CRYPTO SEX AGE HEIGHT WEIGHT PREDICT LOWERMEAN 1 2 F 15 62.5 112.5 100.66 95.225 2 3 F 14 64.3 90 107.68 101.84

rogerjdeangelis · ‎02-23-2017

This seems to work * create some data; data have; infile cards; input id1; cards; 567123456 104111111 311555555 442819299 224262144 591048576 661048111 999999999 008111111 008110000 010101010 ; run;quit; * encrypt and decrypt; data want; set have; encrypt=encrypt(id1); decrypt=decrypt(encrypt); put id1 encrypt decrypt; run;quit; options cmplib = (work.functions /* add other libs here */); proc fcmp outlib=work.functions.hashssn; function encrypt(ssn); length rev $17; key=8192; ssn_remainder = mod(ssn, key); ssn_int = round((ssn - ssn_remainder)/key,1); * 8192 ssn_remainder = 1 and ssn_int = 1 ie 1*key + 1 = original value; ssn_big = ssn_int*100000 + ssn_remainder; rev=reverse(put(ssn_big,17.)); if substr(rev,1,1)=0 then rev=cats('-1',substr(rev,2)); ssn_encrypt=input(rev,17.); return(ssn_encrypt); endsub; run;quit; proc fcmp outlib=work.functions.hashssn; function decrypt(ssn); length rev $17; key=8192; rev=strip(reverse(put(ssn,17.))); if index(rev,'-')>0 then substr(rev,index(rev,'-')-1)='0 '; ssn_big=input(rev,17.); ssn_decrypt = round(ssn_big/100000,1)*key + mod(ssn_big,10000); return(ssn_decrypt); endsub; ;run;quit; Up to 40 obs WORK.WANT total obs=11 Obs ID1 ENCRYPT DECRYPT 1 567123456 -1867082296 567123456 2 104111111 5717080721 104111111 3 311555555 3065013083 311555555 4 442819299 9370055045 442819299 5 224262144 4416057372 224262144 6 591048576 8693094127 591048576 7 661048111 3682049608 661048111 8 999999999 95520070221 999999999 9 8111111 13010099 8111111 10 8110000 21180989 8110000 11 10101010 472003321 10101010

rogerjdeangelis · ‎02-23-2017

There is an issue when encrypting ssns that end in 0

rogerjdeangelis · ‎02-23-2017

Weak reversible encryption (set up for SSNs) Using rounded random numbers does not insure 1 to 1 or easy reversibility and can be problematic. If your ids are considerable larger than your key and your ids are less than 9999999999 then this should work. key has to be a power of 2 (shift bits) HAVE ==== Up to 40 obs WORK.HAVE total obs=8 Obs ID1 1 104111111 2 567123456 3 311555555 4 442819299 5 224262144 6 591048576 7 661048111 8 999999999 WANT ==== Up to 40 obs WORK.WANT total obs=8 Obs ID1 ENCRYPT DECRYPT 1 104111111 5717080721 104111111 2 567123456 867082296 56713984 3 311555555 3065013083 311555555 4 442819299 9370055045 442819299 5 224262144 4416057372 224262144 6 591048576 8693094127 591048576 7 661048111 3682049608 661048111 8 999999999 95520070221 999999999 SOLUTION ======== * create some data; data have; infile cards; input id1; cards; 104111111 567123456 311555555 442819299 224262144 591048576 661048111 999999999 ; run;quit; * encrypt and decrypt; data want; set have; encrypt=encrypt(id1); decrypt=decrypt(encrypt); put id1 encrypt decrypt; run;quit; options cmplib = (work.functions /* add other libs here */); proc fcmp outlib=work.functions.hashssn; function encrypt(ssn); length rev $16; key=8192; ssn_remainder = mod(ssn, key); ssn_int = round((ssn - ssn_remainder)/key,1); * 8192 ssn_remainder = 1 and ssn_int = 1 ie 1*key + 1 = original value; ssn_big = ssn_int*100000 + ssn_remainder; rev=reverse(putn(ssn_big,16.)); ssn_encrypt=input(rev,16.); return(ssn_encrypt); endsub; run;quit; proc fcmp outlib=work.functions.hashssn; function decrypt(ssn); length rev $16; key=8192; rev=reverse(putn(ssn,16.)); ssn_big=input(rev,16.); ssn_decrypt = round(ssn_big/100000,1)*key + mod(ssn_big,10000); return(ssn_decrypt); endsub; ;run;quit;

rogerjdeangelis · ‎02-23-2017

data have; input x ID Year rate ; datalines; 1 1 2010 1 2 1 2011 2 3 1 2012 1 4 2 2010 4 5 2 2011 3 6 2 2012 2 7 3 2012 2 8 1 2013 1 ; run; Ods Output Observed=havsum(Rename=(Label=Product sum=Total)); Proc Corresp Data=have Observed dim=1; Table id, year; weight rate; run;quit; data want; set havsum; avg=mean(of _:); run;quit; Up to 40 obs WORK.WANT total obs=4 Obs PRODUCT _2010 _2011 _2012 _2013 TOTAL AVG 1 1 1 2 1 1 5 1.25 2 2 4 3 2 0 9 2.25 3 3 0 0 2 0 2 0.50 4 Sum 5 5 5 1 16 4.00

Online Status	Offline
Date Last Visited	‎12-04-2021 02:20 PM

Youtube Point and Shoot Programming with the Original Display Manger E...

Re: Fun With SAS ODS Graphics: YAAEB (Yet another animated Easter Bunn...

Re: SAS Macros

Re: Deleting datasets older than 24 months

Re: Macro for using data step to several dataset

Re: Large integer field format in SAS

Re: SAS Grid Pros and Cons

Re: Paradox Table Import into SAS EG Problems

Re: Export to .csv value comes in a separate line

Re: Splitting columns

Fun With SAS ODS Graphics: YAAEB (Yet another animated Easter Bunny)

Re: Splitting up dataset based on unique variable

Youtube Point and Shoot Programming with the Original Display Manger E...

Re: Convert a column to a vector

Re: change rows order

Re: Proc Import multiple CSV files

Re: Urgent Help

Re: Keep blanks when exporting to .txt

Re: Urgent Help

Re: Remove empty rows and empty columns

Re: Looping and Concat col in QB

Re: Pass a parameter in macro

Re: Dynamic Matrix Multiplication

Re: How to select Excel sheets to import based on position, and import...

Re: Is there any way to transpose the observation into variable

Re: Is there any way to transpose the observation into variable

Re: Randomly generated IDs and sequential IDs

Re: Randomly generated IDs and sequential IDs

Re: Randomly generated IDs and sequential IDs

Re: Randomly generated IDs and sequential IDs

Re: Transponse and add average - a solution, but maybe not the most ef...

SAS Analytics Explorers