If I understand your customer count correctly, something like this should work. When you then apply a filter on date VA will only include the customer counts that obey the filter.   proc sql; create table want as select date ,custid ,count(*) as Count from have group by date ,custid ; quit; ... View more

As mentioned in https://communities.sas.com/t5/SAS-Visual-Analytics/Visual-Analytics-First-var-Last-var/m-p/484219/highlight/true#M10816 - VA doesn't allow row based calculations such as sequence or lag (as these require special sorting to work). Did you try using path analysis to see whether this is suitable for your needs? Things like customer count on given paths and measuring overall customer journey / contact points is part of sankey diagrams.   Otherwise you could create a SAS job to create and update the table using data step. Yes, you can use SAS Enterprise Guide (or even SAS Studio if you are on SAS Viya already). Probably easier just to have the job replacing the in-memory table on regular basis so any report consumer gets updated values when viewing the report.   Hope this helps. Falko ... View more

Thank you for your response ... View more

It seems to me that SAS VA doesn't support a dinamic report except for SAS Viya.   But if i use a data query in the Data Preparation Panel is it possible to treat the query builder as a stored process? Let's say i create the query. is it possible to associate a from_data and to_data prompts to trigger the execution of the quey in order to create the report? ... View more

Unfortunately there are some limitations in terms how row-based calculations can be done in VA. You have a related discussion here: Re: SAS Visual Analytics create a Sequence variable   You can certainly use your original table straight as input to path analysis (sankey diagram) - it would render similar to this:     Note, that a sankey diagram is really designed to be used on high volume data so you rarely look at an individual but rather at the overall customer journey here. For example, what percentage of my customers first contact by phone and then visit a store? How many customers do not come back (contact) after visiting a store? How many purchase a product? etc.   The sankey diagram shows you the total count (small numbers to the start and the end of a path) taking this particular step.   No, the labeling in a sankey diagram is done given the event identifier. in your example this is probably 'wayOfContact'. You can however use special path filters here to identify a particular path(s) of interest. For more detailed investigation - you can create other visualizations given a path or event selection (e.g. bar chart showing the customers behind this path). Such visualization can have additional labels, measures etc to describe this path subset.   Hope this helps. Falko ... View more

Unfortunately the VA query engine doesn't allow such sequence or row based calculation and I'm not aware of a work-around other than calculating such as part of the data preparation step. ... View more

Hi @LinusH,   i keep on not understanding so well how the LASR Schema works. I have one fact table 10M records and one dimension table 1M records. The cardinality it's 1 to n. Viewing the output i understand that: - VA performs an inner join - VA performs a sort by key: using the output table in an Exploration if i filter for a specific value, which should have two rows (i tried the query on Guide), i can see only the first row.   Could you help me?   Thank you ... View more

There are several techniques for data reduction and interpretation but the choice among which ones to try depends greatly on the nature of your data (number and type of fields, number of observations, presence of and/or type of target/outcome variable of interest) as well as the business problem you are trying to solve.   You must also consider the available computing power you have to solve the problem since some approaches might be more computationally intensive than others.   It would be helpful if you could describe the nature of your data as well as the objective you hope to accomplish after analyzing your data.     Cordially, Doug ... View more

Hi, do you know how can i run MCA in Enterprise MINER? ... View more

It is true that SAS Web Report Studio does not support column-level security. See   http://support.sas.com/kb/45/094.html   If that is what you were asking, then that cannot be done.   The option for cubes is to exclude (or include) members with only missing values. By default, these members are excluded. In this case, if an entire row or column contains nothing but missing values, then they are suppressed in the crosstab. As far as I am aware, you cannot use this feature as column-level security. ... View more

Use a varaible with shorter values for the ID variable and use the existing one for the IDLABEL variable instead.  If you do not have another variable you can use then make one.  Here is a method to just number the distinct values found. data have ; length NUMERO_CASO $15 DESCRIZIONE_SERVIZIO $50 COUNT_SERV 8; infile cards dsd dlm='|' truncover ; input NUMERO_CASO DESCRIZIONE_SERVIZIO COUNT_SERV ; cards; 1-1107594808|ESTRATTO PAGAMENTI SENZA PIN|1 1-1107594808|PENSIONI - DETTAGLIO RATA PAGAMENTO CON PIN|7 1-1107594808|PENSIONI - DETTAGLIO RATA PAGAMENTO SENZA PIN|4 1-1107594808|PIN INPS - GESTIONE RICHIESTE|3 1-1107594808|STATO DOMANDA|2 1-1107594808|VISUALIZZAZIONE CORRISPONDENZA INPS|5 ; proc sql ; create table fmt as select distinct DESCRIZIONE_SERVIZIO as START from have ; quit; data fmt; fmtname='id'; type='i'; label+1; set fmt; run; proc format cntlin=fmt fmtlib; run; data step1; set have; id = input(DESCRIZIONE_SERVIZIO,id.); run; proc transpose data=step1 prefix=SERV out=want(drop=_name_) ; by NUMERO_CASO; id id; idlabel DESCRIZIONE_SERVIZIO; var COUNT_SERV; run; proc print data=want ; run; proc print data=want label; run; ... View more

I don't know your data so it is hard for me to interpret your result.  I did experiment on a problem and I see that the results seem reasonable in my plot. More similar clusters "tend" to be plotted closer to one another.   That plot is using multidimensional scaling, based on the cluster similarities derived from the cluster means.  The result is a view of high dimensional data in a two-dimensional space. You do lose information when you do this. The optimization is over ALL the cluster pairs so the distance between any particular pair of clusters will not necessarily correspond to their true distance in the higher dimensional space.    Also perhaps if you explore terms that are in common below the initial reported terms you would discover more similarities then you first realized?   Russ  ... View more