Skip to main content
Announcements
Qlik Connect 2024! Seize endless possibilities! LEARN MORE
cancel
Showing results for 
Search instead for 
Did you mean: 
Anonymous
Not applicable

Comparative search for multiple fields from 2 sources

I would like help in performing the following comparative analysis and count the results.

I have two data sources, from DB1 and DB2. 

Source is Database 1   
    
DB1_User_NameTicket_Create_DateTicket_noDB1_Issue_type
Dave10/11/2018123Outlook
Bell10/11/2018124PC
Scott12/10/2018125Application1
Don21/10/2018126Login
Ravi28/10/2018127Outlook
Tim1/11/2018128Application2
Dave5/11/2018129Outlook
Scott17/11/2018130Application3
Bell14/11/2018131Mouse
Don11/11/2018132Application4

 

Source is Database 2    
     
DB2_User_NameSearch_DateSearch_Session_IDAction TypeDB2_Issue_type
Mathew10/10/2018345CallPC
Dave10/10/2018456SearchOutlook
Scott11/10/2018567SearchApplication1
Don20/10/2018678SearchLogin
Paul28/10/2018789StartMouse
Bob1/11/2018900PingApplication3
Chris5/11/20181011CallApplication3
Don17/11/20181122SearchApplication4
Tim14/10/20181233SearchApplication2
Don11/11/20181344SearchApplication4

 

I would like to check if a user listed in DB2 has searched for a particular issue category and later has logged a ticket for the same category in DB1. But I would like to count only tickets logged in DB1 within 3 days from when the search was made in DB2.

 

So in the above example, Dave has made a search for an Outlook issue on 10/10/2018 and later logged a ticket in DB1 for the same category on 10/11/2018 (This falls within the 3 days bracket). I want to be able to count all such ticket (which are unique) for all such cases.

Labels (5)
2 Replies
Saravanan_Desingh

Check the date format. It seems to be DD/MM/YYYY. In that case your expected output is wrong.

Saravanan_Desingh

One solution is:

SET DateFormat='D/M/YYYY';
tab1:
LOAD * INLINE [
    DB1_User_Name, Ticket_Create_Date, Ticket_no, DB1_Issue_type
    Dave, 10/11/2018, 123, Outlook
    Bell, 10/11/2018, 124, PC
    Scott, 12/10/2018, 125, Application1
    Don, 21/10/2018, 126, Login
    Ravi, 28/10/2018, 127, Outlook
    Tim, 1/11/2018, 128, Application2
    Dave, 5/11/2018, 129, Outlook
    Scott, 17/11/2018, 130, Application3
    Bell, 14/11/2018, 131, Mouse
    Don, 11/11/2018, 132, Application4
];
Join 
tab2:
LOAD *, DB2_User_Name As DB1_User_Name, DB2_Issue_type As DB1_Issue_type
;
LOAD * INLINE [
    DB2_User_Name, Search_Date, Search_Session_ID, Action Type, DB2_Issue_type
    Mathew, 10/10/2018, 345, Call, PC
    Dave, 10/10/2018, 456, Search, Outlook
    Scott, 11/10/2018, 567, Search, Application1
    Don, 20/10/2018, 678, Search, Login
    Paul, 28/10/2018, 789, Start, Mouse
    Bob, 1/11/2018, 900, Ping, Application3
    Chris, 5/11/2018, 1011, Call, Application3
    Don, 17/11/2018, 1122, Search, Application4
    Tim, 14/10/2018, 1233, Search, Application2
    Don, 11/11/2018, 1344, Search, Application4
];

tab3:
LOAD *, If(Ticket_Create_Date >= Search_Date And Ticket_Create_Date <= Search_Date+3,'Y') As Flag
Resident tab1;

Drop Table tab1;

commQV24.PNG