HI @silvernet2<>, What version of SAS Customer Intelligence are you using ? is this Marketing Automation or Real Time Decision Manager ? Regards James ... View more

HI Muralidhar, Unfortunately that generic error message could be many different things. You will need to take a look at {SASHOME}/Config/Lev1/Web/Logs/SASCustomerIntelligenceCore.6.x.log (where x is your version number), around the time that you experienced the generic error. This sort of a problem is best tackled with SAS Technical Support - so you are advised to open a track with them. Regards James ... View more

Hi Faisal, This is a complex topic with many possible causes and solutions. Your best course of action is to open a track with SAS Technical Support. Regards James ... View more

HI Bhanu, Im not familiar with Python, however its important to note that for RTDM we are using Jython. To extend this with pure-Python packages you can append to the Jython module search paths with the following -D option (for SASServer7): -Dcom.sas.analytics.crm.rtdm.pythonPaths={Paths to your python modules and code} Regards James ... View more

Hi Bhanu, The Data Process does not support mapping List type variables as an input to a criteria variable. To achieve what you want you would need to create a SAS process using DS2 code (or using Groovy or Jython), to accept a List and then translate that into a IN clause in the SQL query. Your first query that retrieves the initial list is OK, then downstream from it add a Process Node. An example of the type of DS2 code you would need is below. Note the FedServer only supports 16 values in an IN Clause. This code essentially loops through the String List, and builds the IN Clause with the values from the list. It returns a flag to indicate if the customer has one or more of the input products, you can modify this for your needs. HTH, James package CheckCustomerProducts / overwrite=yes; /** This package will support up to 16 values for the IN CLAUSE. **/ /** FedServer throws a SQL prepare error if more than 16 parameters are passed in the IN clause **/ /** If the products list contains > 16 items, only the first 16 are used. The remainder will be ignored **/ dcl package tap_logger m_logger(); dcl package sqlstmt m_sqlstatement('select CustomerID, prod_nm from {Catalog}.{Schema}.{SomeTable} where CustomerID = ? AND prod_nm IN (?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)') ; method execute( double CustomerID, package tap_string_array products, in_out Integer HasProducts); dcl int i size rc ; dcl int param_num ; dcl int m_maxBind; dcl double startDTTM endDTTM elapsedTime; m_maxBind = 16 ; if (m_logger.isDebugEnabled()) then do; startDTTM=datetime() ; m_logger.debug(catt('CheckCustomerProducts: Entered execute method for CustomerID: ',CustomerID)); end ; size = products.size(); if size > 0 then do; if (m_logger.isDebugEnabled()) then do ; if (size > m_maxBind) then m_logger.debug('CheckCustomerProducts: Product list has '||size||' items. Only the first '||m_maxBind||' values will be used.'); else m_logger.debug('CheckCustomerProducts: Product list has '||size||' items.'); end ; m_sqlstatement.setdouble(1,CustomerID); param_num = 1 ; /** Set parameters from list up to the value of m_maxBind or size of list **/ do while (param_num <= size and param_num <= m_maxBind) ; m_sqlstatement.setchar(1+param_num, products.get(param_num)); param_num = param_num + 1 ; end ; /** If size is smaller than m_maxBind, fill remaining parameters with last value in list **/ do while (param_num <= m_maxBind) ; m_sqlstatement.setchar(1+param_num, products.get(size)); param_num = param_num + 1 ; end ; m_sqlstatement.execute(); rc = m_sqlstatement.fetch(); if rc = 0 then do; /* successful fetch therefore has products*/ HasProducts = 1; if (m_logger.isDebugEnabled()) then m_logger.debug('CheckCustomerProducts: FETCH returned records therefore HasProducts = '||HasProducts); end; else if rc = 1 then do; /* ERROR unsuccessful fetch therefore has NO products*/ HasProducts = 0; if (m_logger.isDebugEnabled()) then m_logger.debug('CheckCustomerProducts: FETCH returned ERROR therefore HasProducts = '||HasProducts); end; else if rc = 2 then do; /* NODATA therefore has NO products*/ HasProducts = 0; if (m_logger.isDebugEnabled()) then m_logger.debug('CheckCustomerProducts: FETCH returned NODATA therefore HasProducts = '||HasProducts); end; end; else do; HasProducts=0; if (m_logger.isDebugEnabled()) then do ; endDTTM = datetime() ; elapsedTime=endDTTM - startDTTM ; m_logger.debug('CheckCustomerProducts: No Products to search for provided. Exiting DS2 process'); end ; end; if (m_logger.isDebugEnabled()) then do ; endDTTM = datetime() ; elapsedTime=endDTTM - startDTTM ; m_logger.debug('CheckCustomerProducts: Execution time for CustomerID '||CustomerID|| ' is '||left(put(elapsedTime*1000, COMMA6.2))||'ms.'); end ; end; /** execute method **/ endpackage; run; ... View more

Hi, Sorry - but your new example confuses me. Can you try using a Character Custom detail inside a treatment (ensuring you have created a Custom Detail Tag and attached it to the new Custom Detail), assign its value using a calculated item with the TOSTRING function, and then assigning the packageTag in the reply. Regards James ... View more

HI vasuthevan, I think you are going to have a problem with your approach. The package.Tags is an array (or List), so when you pass this to the TOSTRING function, it converts the whole array to a string, and encloses it in curly braces. There are no functions available to convert a whole array to one of another type. My only suggestion it to change the Custom Detail on your treatment to be a Character, then when assigning the value use a Calculated item with a date literal or date function to create the date, and then use TOSTRING function to convert it to a Character (whilst still inside the Calculated Item). Then the packageTag will be a Character List and will map directly to your reply variable. Rgds James ... View more

Hi AlanisN, The ANY or ALL rule significantly changes the logic of the query generated. It is used when there is a one to many join between the main customer table and the table you are filtering (in your case the Contact History table). A customer might have numerous records in the contact history table, and you are looking for customers who have records with a contact date of last 3, 6 or 10 days. Some customers may have records that fall outside this range (for example 20 days, or 100 days). For example: Customer 1: Contact Date 5 days ago, and 30 days ago Customer 2: Contact Date 5 days ago If your query is "in the last 6 days", and you use the ANY rule, the query will find customers with at least one record within the last 6 days, but may also have records outside of the last 6 days. In this example both Customer 1 and 2 will be selected. If you query is "in the last 6 days", and you use the ALL rule, the query will find customers who only have records within the last 6 days (i.e. ALL of the contacts happened within the last 6 days). In this example only customer 2 will be selected. The results of these two rules is very different. The sql for these two rules is: ANY: PROC SQL; Create table WORK.MA3 as SELECT DISTINCT table0.CustomerID AS CustomerID LABEL='Customer ID' FROM CDM.CI_CONTACT_HISTORY table2 Inner join CDM.CI_CELL_PACKAGE table1 on table2.CELL_PACKAGE_SK = table1.CELL_PACKAGE_SK Inner join DataMart.CUSTOMER_SUMMARY table3 on table3.CustomerID = table2.CustomerID Inner join DataMart.CUSTOMER table0 on table0.CustomerID = table3.CustomerID WHERE table1.COMMUNICATION_CD = 'COMM265681' AND ( ( (( table2.CONTACT_DTTM ) >= '29Jul2018:00:00:00'dt AND ( table2.CONTACT_DTTM ) < '30Jul2018:00:00:00'dt) /* 3 days ago */ OR (( table2.CONTACT_DTTM ) >= '26Jul2018:00:00:00'dt AND ( table2.CONTACT_DTTM ) < '27Jul2018:00:00:00'dt) /* 6 days ago */ ) OR (( table2.CONTACT_DTTM ) >= '22Jul2018:00:00:00'dt AND ( table2.CONTACT_DTTM ) < '23Jul2018:00:00:00'dt) /* 10 days ago */ ) ; quit; ALL: PROC SQL; Create table WORK.MA4 as SELECT table4.CustomerID AS CustomerID LABEL='idFromSelection1__0 DIR' FROM ( SELECT DISTINCT table0.CustomerID AS CustomerID LABEL='Customer ID', ( count(*) ) AS DII_16 LABEL='' FROM CDM.CI_CONTACT_HISTORY table2 Inner join CDM.CI_CELL_PACKAGE table1 on table2.CELL_PACKAGE_SK = table1.CELL_PACKAGE_SK Inner join DataMart.CUSTOMER_SUMMARY table3 on table3.CustomerID = table2.CustomerID Inner join DataMart.CUSTOMER table0 on table0.CustomerID = table3.CustomerID WHERE table1.COMMUNICATION_CD = 'COMM265681' AND ( ( ( ( table2.CONTACT_DTTM ) >= '29Jul2018:00:00:00'dt AND ( table2.CONTACT_DTTM ) < '30Jul2018:00:00:00'dt) /* 3 days ago */ OR ( ( table2.CONTACT_DTTM ) >= '26Jul2018:00:00:00'dt AND ( table2.CONTACT_DTTM ) < '27Jul2018:00:00:00'dt) /* 6 days ago */ ) OR ( ( table2.CONTACT_DTTM ) >= '22Jul2018:00:00:00'dt AND ( table2.CONTACT_DTTM ) < '23Jul2018:00:00:00'dt) /* 10 days ago */ ) GROUP BY 1 ) table4 Inner join ( SELECT table8.CustomerID AS DIR_1 LABEL='Customer ID', ( count(*) ) AS DII_2 LABEL='' FROM DataMart.CUSTOMER table8 Inner join DataMart.CUSTOMER_SUMMARY table9 on table8.CustomerID = table9.CustomerID Inner join CDM.CI_CONTACT_HISTORY table7 on table9.CustomerID = table7.CustomerID Inner join CDM.CI_CELL_PACKAGE table6 on table7.CELL_PACKAGE_SK = table6.CELL_PACKAGE_SK GROUP BY 1 ) table5 on (table4.CustomerID = table5.DIR_1 AND table4.DII_16 = table5.DII_2) ; quit; Cheers James ... View more

Hi, Do have access to (or can you request a copy of) the SASCustomerIntelligenceCore log file from your environment ? This log file will show you the type of code that is generated by the Map Node. Rgds James ... View more