This is very interesting,I will test it soon. Thanks! ... View more

rsubmit remote=U_server WAIT=YES; ... View more

@GeorgeSAS wrote: Other than two lines, is there a better visual effect graph that can compare them in percentage? Thanks! What does that mean?   Other plot suggestions are: 1. Plot the difference rather than the values 2. Plot it as a band, with the area between coloured to highlight the differences 3. Plot bars instead of line   All of these can be accomplished with SGPLOT.  ... View more

As far as I know sas dataset can be read concurently. Please checkout  interesting link below   http://support.sas.com/resources/papers/proceedings15/3380-2015.pdf ... View more

@GeorgeSAS wrote: Thank you Kurt,   Does it must be a .sh file to use 2>&1 ?   if it is a sas file, I can't use this : "sas sortdata.sas 2>&1"  thanks! Can be anything runnable from the commandline. I wrote a quick program: data sasuser.test; x1 = datetime(); format x1 e8601dt22.; put x1=; run; and stored it as test.sas in my home directory. Then I ran this from my EG session: filename oscmd pipe "/sas/SASFoundation/9.4/sas $HOME/test.sas -altlog $HOME/test.log 2>&1"; data _null_; infile oscmd; input; put _infile_; run; Note that I specified the whole path to the SAS executable, as we do not have that in the PATH environment variable. The log in EG shows this: 24 filename oscmd pipe "/sas/SASFoundation/9.4/sas $HOME/test.sas -altlog $HOME/test.log 2>&1"; 25 26 data _null_; 27 infile oscmd; 28 input; 29 put _infile_; 30 run; NOTE: The infile OSCMD is: Pipe-Kommando="/sas/SASFoundation/9.4/sas $HOME/test.sas -altlog $HOME/test.log 2>&1" NOTE: 0 records were read from the infile OSCMD. NOTE: DATA statement used (Total process time): real time 0.41 seconds cpu time 0.00 seconds which indicates no message came back from the system. $HOME/test.log looks like this (excerpt): 1 data sasuser.test; 2 x1 = datetime(); 3 format x1 e8601dt22.; 4 put x1=; 5 run; x1=2017-08-22T07:39:14 NOTE: The data set SASUSER.TEST has 1 observations and 1 variables. NOTE: DATA statement used (Total process time): real time 0.03 seconds cpu time 0.00 seconds and dataset SASUSER.TEST shows this: x1 2017-08-22T07:39:14 ... View more

This is how I name my graphic files:   ods listing gpath="&path.\Graphiques" style=journal image_dpi=600; ods graphics / imagename="Biomarker VG" noborder reset=index height=600 width=600; proc sgplot .... This will create a graph in a square picture in a file named &path.\Graphiques\Biomarker VG.png with the journal style. The reset=index option forces SAS to start the creation of graphic files with Biomarker VG even if the name already exists in the directory. Otherwise SAS would add a number suffix to the name to make it unique.   ... View more

Thanks   whatever size I gave ,the maxium is 4 gb ... View more

Yes. will run SAS program at UNIX ... View more

What do you mean by "large SAS format(more than 1 G)" ?   Assuming you have a format of too many lines you can do: options obs=100; proc format lib=library cntlout=cntl; select format_name; run; options obs=max; Otherwise you may mean a dataset with large number of variables and observations ? still use options obs= to limit number of observations. ... View more

Thank you Shmuel, but I want use upstream excluding instead of down stream.because I want save memory size.the where clause will not save any SAS memory size. Thanks! ... View more

Get help from your local systems team. If you really are running MVS then change the memory settings in the JCL. Are you sure the file is really a format catalog? Can you see the list of formats in it using PROC CATALOG? Or using the DICTIONARY views? Are you sure the file is a format catalog for the version of SAS that you are running?  Format catalogs a normally not compatible across releases or operating systems. ... View more

You need to have /et= in the SELECT statement.   proc format lib=work.formats ; value myfmt 1='Made on Local Session'; run; signon remote=r1 sascmd='!SASCMD' ; rsubmit r1; proc format lib=work.formats; value myfmt 1='Made on Remote Session'; run; proc download incat=work.formats outcat=work.formats; select myfmt / et=format; run; endrsubmit ; signoff r1; proc format fmtlib lib=work.formats; select myfmt ; run; ... View more

Those key definitions are stored in the user profile catalog, sasuser.profile (see Files Used by SAS). Normally, this happens automatically when you close the KEYS window. A subsequent SAS session would then (automatically) access the modified profile catalog including the new key definition.   However, if the SAS system option RSASUSER is in effect, the user profile catalog is read-only and on closing the KEYS window you will get the error messages: ERROR: File I/O Error. Unable to permanently save KEYS settings. ERROR: Cannot open entry DMKEYS.KEYS for write access - catalog SASUSER.PROFILE is opened for read access. This option can only be changed (to NORSASUSER) at SAS invocation or in your config file. The latter typically also contains the definition of library SASUSER where the mentioned profile catalog is located.   System option RSASUSER enables multiple SAS sessions to share a common profile catalog (so that in particular the two notes mentioned below do not occur when a second, third, ... session is started). With the shipped default, NORSASUSER, a second session cannot access the same user profile catalog, hence the modified key definitions are not available.   So, to make the key definition permanent, open SAS with option -NORSASUSER while no other SAS session (with option RSASUSER) is open. Make sure that the log of this new SAS session does not contain notes like these: NOTE: Unable to open SASUSER.PROFILE. WORK.PROFILE will be opened instead. NOTE: All profile changes will be lost at the end of the session. Then modify the key definition and close the session.   Personally, I have created two start icons for SAS in my Windows Start Menu: "SAS" and "SAS (NORSASUSER)". The latter sets option NORSASUSER and I use it only if I want to change key definitions (or make other changes to my user profile), the former sets option RSASUSER and is for everyday use. ... View more

See the TRANSREG doc for specifics.  For regular data analysis, it uses ordinary least squares (OLS). For optimal variable transformations, it iterates between OLS estimates of the parameters and OLS estimates of the transformation parameters (a method that is called the method of alternating least squares). ... View more