How did you check the CSV file?   PROC EXPORT should not change the values, but it will have to convert numbers into text to be able to make a text file, so the formats that are attached to the value will determine how the number is converted to text.   So for example if you run this program: data have; input id value; cards; 1 10.895 2 45.42 ; filename csv temp; proc export data=have file=csv dbms=csv; run; The CSV file you get looks like this: id,value 1,10.895 2,45.42 But if you attach the format 5.1 to the VALUE variable then it will look like this: id,value 1,10.9 2,45.4 Or if you attach the format Z8.4 you will instead get: id,value 1,010.8950 2,045.4200     ... View more

The answer to your question lies in the post you have made: Warning: Apparent symbolic reference TEMP not resolved.  This means the macro variable TEMP is not being created.  Thus something is amis with the code (perhaps you are running on another OS or in a system where the environment variable does not exist): data _null_; temp = sysget('temp'); * Put the environment variable %temp% in SAS \ variable temp call symput ('temp', right(trim(temp))); * put the content of temp in the SAS macrovariable &temp; run; However, more importantly, this thread is 3 years old.  Start a new thread with your question, inputs, outputs, test data etc. I would advise against the use of VBS script as it may not work with your system. Are you trying to read in an Excel file?  No information to work with, hence why posting a full question would be betteras we have no idea what you are doing or how. ... View more

Thank you very much jiltao for your reply. So nice of you ... View more

Hi Steve,    I am a little confused about one thing in your example (and in the example in the SAS documentation) -- why there are y1 and y2, and why do they switch places in the two equations?    As the topic starter, I am also trying to run the Hausman Test, and I am not sure how to write the syntax... I thought that with y, endogenous x, and say two instruments (z1 and z2), I should write the first stage first and then the second stage... Is it not right? Where do y1 and y2 come from?    proc model data = data; endo x; instruments z1 z2; y = b0 + b1*x; x = b00 + b2*z1 + b3*z2; fit y / ols 2sls hausman; run; ... View more

Got it! Thank you. The problem is solved -- I just needed to move the SET statement inside the first loop.  ... View more

Hi, I have same problem as you. The computation takes very long time, especially for confirmatory  multidimensional models.    To make it faster, I relax the convergence criteria to gconv=1E-4 (I know this gives is less accurate estimates). I also found that the conjugate gradient optimization works fastest in complex models (technique=congra). And, for exploratory models I also relax the convergence on the criteria for finding the rotation matrix (rconverge=1E-3). ... View more

Thank you! That's very helpful. I ran into one more small problem. In the sgplot syntax above, I have datalabel option for scatter and I also have series on the same plot. The problem is my data labels overlap with my lines, and it doesn't look good (also difficult to see labels). Is there a way to get the software to place data labels in a way that they don't overlap with lines? ... View more