I have data such as the following:
Field 1 (unique ID)     Date/time                       
1                               1/1/2016 01:05:09
2                               1/1/2016 01:05:09
3                               2/1/2016 12:00:00
4                               3/1/2016 05:06:06

/*I have two views in a company’s HR database I am pulling from. 

One gives a list of hiring actions. The first time an employee is hired, they’re given a sequence of 1. Each time after they’re given a sequence of 2, 3, 4 and so on.

The second view I’m querying has similar information for employees who were terminated. 

Hi all,
I was extending the partition of some tables to host the 2015 and I have notice this strange behavior.
I have this table called Table1

Posted this on latest Paulsinclair blog post (2012) but I guess I'll it a try here :
Hi, I have a question about partition elimination in 14.00(.06.09) with a table partitioned on a timestamp column.
I've prepared 3 cases studies but first of all we need a materialized calendar table :

All of our tables have effective start and end date columns (EFF_STRT_DT, EFF_END_DT).  We want to create date partitions on some of our larger tables, and our fist instinct is to create the partition on EFF_END_DT, because we set that date to '9999-21-31' to represent "active" rows in our data warehouse.

Hi All,
I am aware that order of columns in defined PI doesn't matter.
But if we have multilevel Partition defined on the table and we are including the partitioning columns as a part of PI, does it matter in anyway how the PPI columns be included in the PI of the table.
I hope I made myself clear.
 
Thanks!

Hi,
I'm new to Teradata and have a question on creating a table with a CPPI.
I have a table with a few hundered million rows. I want to create partitioning on a column called SAP_CODE whose structure is always four digits. The first digit is alphanumeric the last 3 digits are numeric, e.g. S001, M123, Z012.

Hi  --

I've just started playing around with column partitioning in R14. I created a table using a recent example that Paul Sinclair put into a recent Teradata Magazine article:

Hi,

I have a tableA with partition given below,

PARTITION BY ( CASE_N(

sys =  1,

sys=   2,

sys=   3

NO CASE OR UNKNOWN),CASE_N(

cyc =  1,

cyc=   2,

cyc=   3 

 NO CASE OR UNKNOWN) );

All All,

I have a PPI table paritioned by end_dt

Hi,

Is there any way to partition directly on a small (ie. CHAR(3)) character column, without defining a CASE_N or RANGE_N for the column (ie. partition on all unique values of the column)?

What is the best way to extract the output of the 'Help statistics DbName.TblName column Partition;", for the purpose of extracting the value of the column 'Number of Rows'. If you have attempted this through SQL, please share your thoughts. What I've already have in place is the following methodology...
Run 'Collect Statistics on DbName.TblName Column (Partition);
Run 'Help Statistics DbName.TblName Column (Partition);' through BTEQ with SideTitles on

Is there any other way of reliably tell the optimizer to use the PPI other than hard-coding the minimal/maximal value (date) in a following case?

Here is a DDL for a big table with customer transactions:

transaction_table
PRIMARY INDEX NUPI_txn ( transaction_id )
PARTITION BY RANGE_N(transaction_date BETWEEN DATE '2008-01-01' AND DATE '2012-12-31' EACH INTERVAL '1' DAY , NO RANGE OR UNKNOWN)
INDEX NUSI_customer_id( customer_id)