Most recent date

Question

hi guys,
I have a few tables which I want to query, basically one table with customer information lets call it customer and one table with appointment information, lets call it appointment(this table has an appointment_date field). Now, one customer can have multiple appointments. What I want to do is select back all the customers which have NOT had an appointment in the last 100 days.
select *
from customer cust, appointment apt
where apt.appointment_date < trunc(sysdate-100)
The flaw with the above statment is it could be using the appointment_date from any of the appointments against the customer, not necessarily the latest. Is there a way I could do this using the latest appoinment this customer has had?
Thanks in advance for any help I may get.

Answer 1

786733 wrote:
hi guys,
I have a few tables which I want to query, basically one table with customer information lets call it customer and one table with appointment information, lets call it appointment(this table has an appointment_date field). Now, one customer can have multiple appointments. What I want to do is select back all the customers which have NOT had an appointment in the last 100 days.
select *
from customer cust, appointment apt
where apt.appointment_date < trunc(sysdate-100)
The flaw with the above statment is it could be using the appointment_date from any of the appointments against the customer, not necessarily the latest. Is there a way I could do this using the latest appoinment this customer has had?There are several possibilities. It would help us greatly to help you, when you post some create table and insert statements so that we can reproduce it.
Assuming you are only interested in the maximum appointment date value and nothing else from the apointment table. Then you can group.
select cust.col1, cust.col2, ....  /* all columns from table cust that you need */
       ,max(apt.appointment_date) as appointment_date
from customer cust
join appointment apt on add_the_missing_join_condition
where apt.appointment_date < trunc(sysdate-100)
group by cust.col1, cust.col2, ....  /* all columns from table cust that you need */If you want one more column from appointment table, then you can add it using the MAX KEEP syntax
select cust.col1, cust.col2, ....  /* all columns from table cust that you need */
       ,max(apt.appointment_date) as appointment_date
       ,max(apt.appointment_location) keep (dense_rank last order by apt.appointment_date) as location
from customer cust
join appointment apt on add_the_missing_join_condition
where apt.appointment_date < trunc(sysdate-100)
group by cust.col1, cust.col2, ....  /* all columns from table cust that you need */If you want all rows from the appointment table, then using the analytic row_number function is helpfull.
select * from (
   select cust.*, apt.*,
           row_number() over (partition by cust.id order by apt.appointment_date desc) rn
   from customer cust
   join appointment apt on add_the_missing_join_condition
   where apt.appointment_date < trunc(sysdate-100)
where rn = 1 /* only fetch first ordered apointment per customer */