cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
xtc283x
‎03-24-2024
  • 74 Posts
  • 17 Likes Given
  • 0 Solutions
  • 0 Likes Received

Replace missing values with 0

by Quartz | Level 8 in SAS/IML Software and Matrix Computations
‎10-20-2015 10:25 AM
‎10-20-2015 10:25 AM
I have a matrix G that contains 2 missing values:   G 0.0909091 0.2 1 0.6666667 0.2857143 0.1428571 0.2857143 0.2 . 0.25 -2 0.1538462 -2 -0.181818 0.6666667 -0.666667 -1 0.1538462 -1 0.5 0.2222222 0.25 2 -1 0.5 0.2222222 2 -0.285714 0.4 -2 1 -0.333333 0.3333333 0.4 2 0.5 0.4 . -1 1   The CALL CHANGE function only works on text fields. And I don't understand the explanations for the REPLACE function.   What is the logic in SAS/IML to replace those two missing values with 0? ... View more

Re: Predictions from Proc QLIM

by Quartz | Level 8 in Statistical Procedures
‎10-12-2015 03:58 PM
‎10-12-2015 03:58 PM
I have not tried those but they are good suggestions which I will pursue. Regardless, I'm still curious about why QLIM is giving me negative predictions. Any insights? ... View more

Predictions from Proc QLIM

by Quartz | Level 8 in Statistical Procedures
‎10-12-2015 03:18 PM
‎10-12-2015 03:18 PM
My dependent variable contains positive integers truncated at 1. It is also quite skewed. The n is small with 48 observations. My first try was a multiplicative (log-log) model. Thinking I could improve on the Mean Squared Error, I tried quantile regression at the 25th percentile. This had the positive effect of reducing the MSE by 33% versus the OLS multiplicate model, but had the undesireable result of producing some negative predictions. So, I tried Proc QLIM specifying a truncated lower bound equal to 1, thinking that any predictions from this procedure would be constrained at the lower bound.    I was wrong. Proc QLIM produces negative predictions in spite of a lower bound at 1. This result is unexpected and surprising. The documentation doesn't address the issue.    What am I missing here?  ... View more

Re: Matrix reshaping

by Quartz | Level 8 in SAS/IML Software and Matrix Computations
‎10-12-2015 08:08 AM
‎10-12-2015 08:08 AM
That specification of the Shape function worked. When I print the resulting matrix within IML, it is square. However, when I create output containing that matrix, the output resolves to a vector:               Vec2 = the resulting square matrix              create Outputx var {Vec2};            append;            close Outputx;            run;*end Proc IML;     Outputx contains a single column with nxn rows.   Why does Outputx not contain the square matrix Vec2?   Thank you. ... View more

Matrix reshaping

by Quartz | Level 8 in SAS/IML Software and Matrix Computations
‎10-09-2015 06:53 AM
‎10-09-2015 06:53 AM
I have a matrix, x, that is numeric and 49X47. The second row contains all missing values. I would like to reshape it to a square matrix that is 47x47 by eliminating the first and second rows.    The SHAPE function isn't working:            x2=shape(x,3:49,47);   I'm getting execution errors telling me that the "argument should be a scalar."   How do I resolve this?   Thank you. ... View more

Re: IML SVD sensitivities

by Quartz | Level 8 in SAS/IML Software and Matrix Computations
‎09-29-2015 05:38 AM
‎09-29-2015 05:38 AM
Thank you. ... View more

Re: IML SVD sensitivities

by Quartz | Level 8 in SAS/IML Software and Matrix Computations
‎09-14-2015 02:04 PM
‎09-14-2015 02:04 PM
Rick-    Thanks again for your help with developing a robust PCA. I'm a novice user of IML. I'm working with the SAS document you wrote -- Paper 329-2010  Rediscovering SAS/IML® Software: Modern Data Analysis for the Practicing Statistician. In addition, I have your book on Statistical Programming with SAS/IML. The matrix I'm analyzing for the robust PCA is 40 columns by 900 rows. I have a question for you about some code in your paper:   On page 9 you write:       /* compute PCA of robust covariance matrix */    p = ncol(x); /* 20 variables */    RobustCov = est[3:2+p, ]; /* robust estimate of shape parameters */    call eigen(eVal, eVec, RobustCov); /* PCA = eigenvectors of RobustCov */   I can't find any documentation on the "est" function in your book. So, when you write                      RobustCov = est[3:2+p, ];   What exactly is that line of IML code doing? Is the "3:2+p" based on the sample data you're using? If I want to rewrite that line of code for the 40x900 matrix I'm working with, what would it look like? Currently, I'm getting errors with a simple replication of your code:   Cut and paste of IML error messages in my SAS log:   p=ncol(x); RobustCov=ext[3:2+p, ]; ERROR: (execution) Matrix has not been set to a value. operation : [ at line 1231 column 14 operands : ext, *LIT1001, _TEM1001, ext 0 row 0 col (type ?, size 0) *LIT1001 1 row 1 col (numeric) 3 _TEM1001 1 row 1 col (numeric) 39 statement : ASSIGN at line 1231 column 1 1232 call eigen(eVal, Evec, RobustCov); ERROR: (execution) Character argument should be numeric. operation : EIGEN at line 1232 column 1 operands : RobustCov RobustCov 0 row 0 col (type ?, size 0) statement : CALL at line 1232 column 1 1233 VarExplained=cusum(eval)/sum(eval); ERROR: (execution) Matrix has not been set to a value. operation : CUSUM at line 1233 column 19 operands : eval eVal 0 row 0 col (type ?, size 0) statement : ASSIGN at line 1233 column 1 1234 NumPC=1; 1235 RobustLoc=est[1,]; ERROR: (execution) Matrix has not been set to a value. operation : [ at line 1235 column 14 operands : est, *LIT1004, est 0 row 0 col (type ?, size 0) *LIT1004 1 row 1 col (numeric) 1 statement : ASSIGN at line 1235 column 1 1236 c=(x-RobustLoc); ERROR: (execution) Matrix has not been set to a value. operation : - at line 1236 column 5 operands : x, RobustLoc x 897 rows 37 cols (numeric) RobustLoc 0 row 0 col (type ?, size 0) statement : ASSIGN at line 1236 column 1 1237 Scores=c*eVec[,1:NumPC]; ERROR: (execution) Matrix has not been set to a value. operation : [ at line 1237 column 14 operands : eVec, , *LIT1005, NumPC Evec 0 row 0 col (type ?, size 0) *LIT1005 1 row 1 col (numeric) 1 NumPC 1 row 1 col (numeric) 1 statement : ASSIGN at line 1237 column 1 1238 print scores; ERROR: Matrix Scores has not been set to a value. statement : PRINT at line 1238 column 1 1239 1240 create RobustPCA var {pgm scores}; 1241 append; 1242 close robustpca; NOTE: The data set WORK.ROBUSTPCA has 897 observations and 2 variables.     What am I doing wrong?   THank you, Thomas     ... View more

Redirecting SAS log and output files using PROC PRINTTO

by Quartz | Level 8 in SAS Procedures
‎09-14-2015 01:27 PM
‎09-14-2015 01:27 PM
Hello-       I am running v9.3 TSIM1 and would like to redirect my SAS log and output to an external file using Proc Printto WITHOUT the output being automatically written to the "Results" log in PC SAS. The output being automatically directed to the "Results" log is large and slows down the performance of the SAS program dramatically. The reason for this is that I am running a SAS/STAT procedure using a "by" statement and this is useless output.       Here is the syntax I am using for the Proc Printto:                 proc printto new log='c:\data\logfilex.txt'    file='c:\data\logfilex.txt';               run;      The SAS log output is being successfully written out to the external file. In addition, SAS documentation recommends putting the log= and file= (or print=) options in the same Proc Printto statement. Note two things about this:      - Using two Proc Printtos -- one for the log= and a second for the file= -- has no effect on the "Results" log      - Using an "ODS Listing Close;" options statement also has no effect on output written to the "Results" log.    Any suggestions? ... View more

Re: Array subscript out of range

by Quartz | Level 8 in SAS Programming
‎08-05-2015 02:08 PM
‎08-05-2015 02:08 PM
Ok, thanks to all for their suggestions. The sly devils that created the %macro names and the format used in the put statements had them out of sync in terms of their element counts....they were just trying to confuse us mere mortals and protect the IP of their program. ... View more

Re: Array subscript out of range

by Quartz | Level 8 in SAS Programming
‎08-05-2015 11:06 AM
‎08-05-2015 11:06 AM
I am getting conversion comments to the log. Length 3 for the two numeric fields is fine as these are 0,1 dummies. Does this mean that applying character formatting to numeric fields in the first two statements below is the problem?          Nexus21=put(Nexus11,$group1z.);          Nexus22=put(Nexus12,$group2z.);          IndicatorMatrix{Nexus21}=1;          IndicatorMatrix{Nexus22}=1; Thanks! ... View more

Re: Graphing networks

by Quartz | Level 8 in Graphics Programming
‎08-05-2015 11:02 AM
‎08-05-2015 11:02 AM
Thank you! That worked despite a Java issue. ... View more

Re: Graphing networks

by Quartz | Level 8 in Graphics Programming
‎08-05-2015 10:22 AM
‎08-05-2015 10:22 AM
That would be so helpful! Especially if I can get it to work on my data. Btw, I want to create a network for over 200 medical diagnostic codes to graphically illustrate their co-occurrence in patients over a year's period. I'm running v9.3 with the TSIM1 maintenance install. Thank you. ... View more

Re: Array subscript out of range

by Quartz | Level 8 in SAS Programming
‎08-05-2015 10:19 AM
‎08-05-2015 10:19 AM
Interesting! The values of 61 for Nexus21 and Nexus22 are counts? I thought they were the resolved values of the format and put statements for Nexus11 and Nexus12... This may have diagnosed the problem. What's the solution? How do I fix it? Thank you! ... View more

Array subscript out of range

by Quartz | Level 8 in SAS Programming
‎08-05-2015 09:40 AM
‎08-05-2015 09:40 AM
I've never been a sophisticated array statement processor. Regardless, I've inherited some SAS code that contains the following statements:       data test(keep=keyno %names);set junk;        length nexus11 nexus12 $8 nexus21 nexus22 3;        retain %names None Other 0;        by keyno;         array IndicatorMatrix{*} %names None Other;          Nexus11=put(code_1,$charxz.);          Nexus12=put(code_2,$charyz.);          Nexus21=put(Nexus11,$group1z.);          Nexus22=put(Nexus12,$group2z.);          IndicatorMatrix{Nexus21}=1;          IndicatorMatrix{Nexus22}=1;         if last.keyno; So far, so good, right? The IndicatorMatrix array statement calls a little macro named %names -- which has 56 elements -- and adds two additional variables "None" and "Other" for a total of 58 elements, correct? Then, when I run this code, I get the following error: ERROR: Array subscript out of range at line 3 column 50. keyno=001220832MAR-11430 age=86.849315068 code_1=07819 code_2=7295 Sex=Female FIRST.keyno=1 LAST.keyno=0 nexus11=99900000 nexus12=11113456 nexus21=61 nexus22=61 AIDSH=0 CANH=0 CANL=0 CANM=0 CANVH=0 CAREL=0 CARL=0 CARM=0 CARVH=0 CERL=0 CNSH=0 CNSL=0 CNSM=0 DDL=0 DDM=0 DIA2L=0 DIA2M=0 EYEVL=0 GENEL=0 GIH=0 GIL=0 GIM=0 HEMEH=0 HEML=0 HEMM=0 HEMVH=0 HIVM=0 HLTRNS=0 INFH=0 INFL=0 INFM=0 METH=0 METM=0 METVL=0 PRGCMP=0 PRGINC=0 PSYH=0 PSYL=0 PSYM=0 PSYML=0 PULH=0 PULL=0 PULM=0 PULVH=0 RENEH=0 RENL=0 RENM=0 RENVH=0 SKCL=0 SKCM=0 SKCVL=0 SKNH=0 SKNL=0 SKNVL=0 SUBL=0 SUBVL=0 None=0 Other=0 _ERROR_=1 _N_=1 NOTE: The SAS System stopped processing this step because of errors. NOTE: There were 2 observations read from the data set WORK.JUNK. WARNING: The data set WORK.TEST may be incomplete.  When this step was stopped there were 0          observations and 57 variables. When I count the number of elements in the %names macro call, I get 56. If you count the elements dumped out in the error statement, you get 58 elements applied to the IndicatorMatrix array. The {*} is the problem, it would seem. Somehow it's returning the wrong number of elements. I've tried everything I can think of to get this code to work. Nothing has. I don't see what I'm doing wrong. Any suggestions as to how to fix this? Thanks! ... View more

Re: Graphing networks

by Quartz | Level 8 in Graphics Programming
‎08-04-2015 11:44 AM
‎08-04-2015 11:44 AM
Thank you. I do not have this Workshop in my SAS installation. Any other suggestions? ... View more
Contact Me
Online Status
Offline
Date Last Visited
‎03-24-2024 11:28 PM
Likes given to
Quick links