You cannot create a custom category from an aggregated measure even in the  most current release.  ... View more

@gabras wrote: The JSON is the format in which a value is stored in a dataset column, as you can see in the xlsx file i attached. Think the excel as the export of a sas dataset. But why did you go to the trouble of copying the dataset into an XLSX file and uploaded the XLSX as an attachment to your post? WHy not just dump the text into the SAS log and copy and paste the text into your posting so that everyone can actually read it.   So does each value of the variable constitute a complete JSON file?  Or are these instead individual lines from a JSONL file?  The JSON engine in SAS cannot read JSONL files.  So if you have the latter you will need to add either square or curly brackets around the output and commas between the lines to convert the JSONL lines into an actual JSON file.   If each value is an independent fully formed JSON object then that might not work. In which case you would need to write each line as its own file and then read each one back in.   How complex is the JSON (remember I can't see it since you hid it in a XLSX file so that when I viewed it in the browser it looks like four black horizontal lines)?  If it is a simple format you can probably just parse it instead by looking for the keywords and taking the next token as the value. ... View more

Just as a thought if you really need dbcommit=1; You could use a MERGE instead of an INSERT - and just do nothing in case of an Update. This would at least make it much simpler to re-run stuff after a partial insert. ... View more

Hi, Unfortunately Sankey and Network diagrams are only available in SAS Visual Analytics Explorer which does not support filter interaction between objects. I'm not aware of a work-around in 7.4 though.   We do support this kind of interaction in either VA 7.5 or in any of the newer SAS Viya based versions (VA 8.x). And yes, a node selection in a network would trigger the auto filtering of other visualizations including Sankey.   Regards, Falko ... View more

Sorry for the delay in responding. As far as I remember the graph shouldn't change (in terms of number of paths displayed) if you change the aggregation. Are you able to share your data so I can see more details (the screens you shared indicate you have used a different data set than the one used in the first example). Or alternatively - can you try using the original data and see whether you see similar results?   Thanks, Falko ... View more

Hi @gabras ,   It sounds like you are trying to create a recommendation engine. After a quick Google search of Recommendation Engine I came to this page. Even though it is implemented in Python, it looks like it gives a good beginning understanding of how they work. https://www.analyticsvidhya.com/blog/2018/06/comprehensive-guide-recommendation-engine-python/ It even looks like the Recommendation Engine might work very similarly to a market basket analysis.   Hope you discover the right model for you, Tom ... View more

As shown above, just use a retained variable, and change the value of the retained value as needed.  You could update it in a the first if, then use that in the second. ... View more

Just to add another possibility in case this issue is encountered - this can occur if the underlying HDFS cluster is in a compromised state.  Check and possibly restart the HDFS cluster.  It's not very clear from the error message but I experienced this.  2 VMs had failed and restarted.  I had started the LASR server again but didn't think to check HDFS, and it turned out 2 datanodes were also down.  Restarting HDFS fixed it. ... View more

@gabras wrote: Could you help me please Earlier, I stated the explanation   It is not necessary (and wrong) to include all three variables. You want the ORD variable to reset when SUBPROD changed, not when all three change.   so apply this explanation to your new situation involving variable ITEM. ... View more

Okay -- for that, we have the SAS IOM Provider for OLE DB (use this to connect to a SAS workspace, similar to the way SAS Enterprise Guide works) and the SAS ODBC Driver (connects to SAS/SHARE server or SPDS server).  Either way, I think to render a SAS view you'll need the driver to connect to a running instance of SAS that can assemble the data rows, with any SAS formats you need, etc. ... View more

OK   I was looking for the SAS AppDev Studio which you can use to develop web application that could permit the end-user to modify table.   Thanks  ... View more

Please provide screenshots of your parameter/prompt definition within the Stored Process. We need both information, the General with type etc., as well as the Prompt Type and Values definition   Are you using SAS Enterprise Guide to develop your Stored Process? ... View more