Creating custom dimensions

[#1]

I have a requirement where I have data as

ID	Name
1	Pa
1	as
3	asa
3	wer
3	asssc
2	aaw

I Want a dimension that would be like Pro1 for ID as 1, Pro2 for ID as 2, Pro3 for ID as 3

[sunny_talwar]

May be this as your calculated dimension:

Pick(ID, 'Pro1', 'Pro2', 'Pro3')

or

Pick(Match(ID, 1, 2, 3), 'Pro1', 'Pro2', 'Pro3')

or

'Pro' & ID

[#3]

I have around lakhs of ID's which would need me to add all the id and their respective names is there not a way where in the Names are give on the basis by grouping all the similar ID's as "Pro"+"id"

[reddy-s]

Hi Shashank,

As you say that there are loads of records to modify, I would suggest you to do it in the load using an ApplyMap or simply make use of the last option which Sunny T mentioned in your load statement.

data:

Load 'Pro' & ID,

         name

       from ........

[Anonymous]

may be like this?

load distinct *,

concat('Pro',ID) as ID

from

TableName;

[#6]

Is there no way to create a dimension itself and not performing data load for that dimension

[sunny_talwar]

Yes there is and that way is to use a calculated dimension. If you don't want to use either of them, then I am afraid there is no 3rd way here.

[reddy-s]

Hi Shashank,

I would suggest you to create the dimension in the load itself as that would improve your performance.

[sunny_talwar]

I agree with Sangram