ERROR: (execution) Matrices do not conform to the operation.

[AM2323]

Hi everyone. I have a question regarding multiplying matrices. I already read Interpreting PROC IML error messages: Matrices do not conform to the operation suggested in other boards and I still do not understand why I get this message. I understand that if I want to multiply xbeta = x # beta; the number of columns in X have to be equal to the rows in beta, right? I am running SAS 9.4

This is part of my code:

xbeta = x#beta;

se= sqrt (vecdiag(x#cov#x`));

lower =xbeta   - (1.96 # se);

upper =xbeta   + (1.96 # se);

When I run the operation I get this on the log:

ERROR: (execution) Matrices do not conform to the operation.

operation : # at line 15164 column 14

operands  : X, BETA

X    502 rows      6 cols    (numeric)

BETA      6 rows      1 col     (numeric)

And also when I try to perform this operation x#cov

operation : # at line 15164 column 80

operands  : X, COV

X    502 rows      6 cols    (numeric)

COV      6 rows      6 cols    (numeric)

What am I doing wrong? Thanks!

[gergely_batho]

# is elementwise multiplication.

* is matrix multiplication

[AM2323]

When I change # to * I get the following error.

ERROR: (execution) Invalid argument or operand; contains missing values.

operation : * at line 15164 column 14

operands  : X, BETA

X    502 rows      6 cols    (numeric)

BETA      6 rows      1 col     (numeric)

[Rick_SAS · Accepted solution]

Matrix multiplication does not support missing values.  If you are trying to mimic the usual OLS regression, you should remove the rows with missing values. This article might be useful:http://blogs.sas.com/content/iml/2015/02/23/complete-cases.html

[Rick_SAS]

See "Ways to multiply in the SAS/IML Language"