I have a column defined as follows in a TeraData table: RunDay DATE FORMAT 'YYYY-MM-DD' I try to insert one row into this table as follows given a macro variable RunDay: %let RunDay = %sysfunc(datetime());
proc sql noerrorstop;
connect to teradata(User=&user. password=&passwd. mode=teradata);
Execute(
INSERT INTO SomeGreatTable
(
RunDay
)
VALUES
(
input("&RunDay.",YYMMDD.)
);
);
)By Teradata;
quit; Any ideas why this does not work? Thanks!
... View more
I am using this skeleton code: PROC GLMSELECT DATA = Training TESTDATA = Test;
...
MODEL
... /selection=lasso(stop=none choose=AIC) stb showpvalues;
...
"Parameter Estimates" = WORK.ParameterEstimates
RUN; When I look at WORK.ParameterEstimates I noticed that: StdErr Probt are missing. Can I somehow ensure that these values are included? Also can I force a selection method to use certain IVs/features?
... View more
Could anyone please be so kind and explain why the 'number of observations used' could ver be lower than the 'number of observations read' when I use proc glmselect? Thanks!
... View more
Thanks, sorry this does not make much sense. I am able to use proc plm. I could use this to potentially calculate the r-square against the test dataset myself. The final aim is to put it into a macro variable.
... View more
Using this code snippet: PROC GLMSELECT DATA = WORK.Training TESTDATA = WORK.Test;
MODEL
.... / selection=none stb showpvalues;
ods output
"Fit Statistics" = WORK.Model_Fit
"Parameter Estimates" = WORK.ParameterEstimates
Nobs = WORK.Nobs;
RUN; The dataset WORK.Model_Fit contains: Root MSE Dependent Mean R-Square Adj R-Sq AIC AICC SBC ASE (Train) ASE (Test) Is the R-Square measured on the training dataset? Is it possible to obatin the R-Square fro the training and test dataset?
... View more
Hope this is the right board. This publication: http://support.sas.com/documentation/cdl/en/hadoopov/68100/PDF/default/hadoopov.pdf mentions that SAS can only run (MapReduce?) some modeling PROCs on Hadoop. Does anyone know these PROCs or even better is there an exhaustive list? Thanks. Christian
... View more
I am using this code: title1 "&Title.";
proc sgplot data=WORK.TempGraphData;
density X / type=normal lineattrs=(color=red) legendlabel='Bla';
density Y / type=normal lineattrs=(color=blue) legendlabel='Bla';
histogram X / transparency=0.75 fillattrs=(color=red) legendlabel='Bla';
histogram Y / transparency=0.75 fillattrs=(color=blue) legendlabel='Bla';
keylegend / location=outside position=bottom;
xaxis label="Normal Curves";
run; but the 2 histograms' bin sizes appear to be very different. can I force the rendering to use the same bin sizes in both diagrams?
... View more
Let us say I have a massive tab separated file stored in a network drive visible to SAS. The file contains these columns: Can I access this file via, for axample PROC SQL, to only populate a SAS dataset based on a subset of this data (e.g. where Id = 1234)?
... View more
Thanks for all your help, This: %let StartDate = 2016/01/01;
proc sql;
connect to teradata
(BULKLOAD=YES MODE=TERADATA user=&user. Password=&passwd.);
CREATE TABLE WorkTest.test AS
select * from connection to teradata
(
SELECT
TOP 10 *
FROM SomeCalendarData
WHERE SomeDate = CAST(%bquote('&StartDate.') AS DATE FORMAT 'YYYY-MM-DD');
);
quit; Does the trick. Although I am still unsure why an explicit CAST like this should be necessary?! Bizare ...
... View more
As i said: SELECT TOP 10 * FROM SomeCalendarData WHERE SomeDate = '2016/01/01'; returns the expected results when I issue it from another teradata client (sql assistent) so I am a bit lost what's going on. Maybe SAS's explicit tries to be clever whilst causing issues.
... View more
