BookmarkSubscribeRSS Feed
dbcrow
Quartz | Level 8

Hey, folks-

 

I'm trying to build a simple pipeline for ensembling logistic regression using the SAS Viya Studio's "Build Model" feature and I get these errors: "This project could not be initialized"  and "The data advisor failed to run on the selected data source." 

 

Here's a sample of the log file: 

 

NOTE: Added action set 'cardinality'.
ERROR: Could not load the format YESNO_ALT of the variable ewe.
ERROR: Generic error.
ERROR: Action summarize failed.
ERROR: The action stopped due to errors.
NOTE: PROCEDURE CARDINALITY used (Total process time):
      real time           0.17 seconds
      cpu time            0.01 seconds
      

NOTE: The table DMDSLIB.STATISTICS does not exist in Cloud Analytic Services.
ERROR: File DMDSLIB.STATISTICS.DATA does not exist.

NOTE: The SAS System stopped processing this step because of errors.
NOTE: Due to ERROR(s) above, SAS set option OBS=0, enabling syntax check mode. 
      This prevents execution of subsequent data modification statements.
WARNING: The data set WORK.STATISTICS may be incomplete.  When this step was stopped there were 0 observations and 0 variables.
NOTE: DATA statement used (Total process time):
      real time           0.02 seconds
      cpu time            0.01 seconds
      


ERROR: Variable _VARNAME_ not found.

NOTE: The SAS System stopped processing this step because of errors.
NOTE: PROCEDURE SORT used (Total process time):
      real time           0.00 seconds
      cpu time            0.00 seconds
      


ERROR: Variable _VARNAME_ is not on file WORK.STATISTICS.
ERROR: Variable _ORDER_ is not on file WORK.STATISTICS.
ERROR: Variable _CARDINALITY_ is not on file WORK.STATISTICS.
ERROR: Variable _SUMFREQS_ is not on file WORK.STATISTICS.
ERROR: Variable _NMISS_ is not on file WORK.STATISTICS.
ERROR: Variable _MIN_ is not on file WORK.STATISTICS.
ERROR: Variable _MAX_ is not on file WORK.STATISTICS.
ERROR: Variable _STDDEV_ is not on file WORK.STATISTICS.
ERROR: Variable _MEAN_ is not on file WORK.STATISTICS.
ERROR: Variable _NOBS_ is not on file WORK.STATISTICS.
ERROR: Variable _SKEWNESS_ is not on file WORK.STATISTICS.
ERROR: Variable _KURTOSIS_ is not on file WORK.STATISTICS.

I'm attaching the full log file. My computer is a PC, 64-bit, running Windows Enterprise 10 and SAS Viya is "Data Explorer" release 2.5, SAS Viya release V.03.05. 

 

How can I get the SAS Viya to initialize the project? I'm at my wit's end with SAS Viya. It looks like a really powerful tool for ML, but I've found that even basic things like reading in data are hard to figure out and time-consuming. 

 

Thanks,

David

 

4 REPLIES 4
tom_grant
SAS Super FREQ
Moving post to SAS Viya community.
tom_grant
SAS Super FREQ

Hi David,

Sorry you are having difficulties with Model Studio (Build Models) - could you attach a screenshot of your pipeline and perhaps the Data tab (if not sensitive)?  Thanks.

dbcrow
Quartz | Level 8

Hi, Tom-

 

I'm attaching the following .pdf's: 

1) Pipeline (SAS5.pdf)

2) Data tab (SASData.pdf)

 

And for good measure: 

3) Project creation interactive dialog (SASProject)

4) Error message (SASError)

 

This is one of several errors I get when creating or trying to open the project. When I try to open the project from the "Projects" screen, I get: "The project could not be opened and returned the following status: creatingProviderError [Details] The advisor for project "MLClimate" could not be executed". 

 

I would greatly appreciate any help you could provide. 

 

Thanks,

David

dbcrow
Quartz | Level 8

Hi, all-

 

The solution turned out to be stripping the datasets of their formats, then importing them into SAS Studio. Two ways to remove the formats are: 

 

1) Exporting the data to a .csv file with PROC EXPORT:  

 

proc export data=data.climate23_nf
  outfile='H:\CHIS\Data\climate23_nf.csv'
  dbms=csv
  replace;
run;

2) Removing the formats manually from within a data step with this code: 

data data.climate23_nf;
	set data.climate23;
		format _numeric_   ;
		format _character_  ;
run;

I did both steps as part of preprocessing in SAS EGP and saved the files to a local directory. I then imported the files in SAS Studio using Build Models --> New Project --> browse for data --> Import --> Local Files.