Interested to Buy Any Domain ? << Click Here >> for more details...
Hi Friends,
I have a input data like,
class_id Marks
101 50
101 60
101 40
102 90
102 35
And i want my output data like
class_id Marks Rank
101 50 2
101 60 1
101 40 3
102 90 1
102 35 2
how to do this in datastage?
- 8 Answers
- 28719 Views
- Cognizant, HCL, TIAA CREF, I also Faced
- E-Mail Answers
Answers were Sorted based on User's Feedback
Answer / ankit gosain
Hi friends,
This question is a trick and as i found in every answer
given previously, there is a slight miss-queue.
Check out my solution for this question:
I/p File(Primary)
!
I/p
file---->Sort---->Sort---->transformer---->join/lookup---->O/p
File
(1). In the first sort stage, make class_id and marks fields
as keys, sort by descending and then create a Key change column.
(2). In second sort stage, make class_id only as a key
(don't sort previously sorted) and create a cluster key
change column.
(3). You will get the o/p as given below:
Class_Id, Marks, KeyChange, ClusterKeyChange
102,90,1,1
102,35,1,0
101,60,1,1
101,50,1,0
101,40,1,0
(4). In transformer stage take a integer type variable say
var1 and create one column say Rank.
Now, derive Var1 as :
if ClusterKeyChange=1 then 1 else Var1
And, derive Rank as :
if ClusterKeyChange=1 then 1 elseif (ClusterKeyChange=0 and
KeyChange=0) then Var1 else Var1+1
(5). Take only class_id, marks & rank as an o/p of the
transformer, so that you will have the o/p as given below:
Class_id, Marks, Rank
102,90,1
102,35,2
101,60,1
101,50,2
101,40,3
(6). Now finally, apply join/lookup and take this o/p as a
Reference data and Primary data will be the Source I/p File.
Take Class_id and Marks as keys and derive Rank to the o/p,
so that you will get the finally o/p as given below:
class_id Marks Rank
101 50 2
101 60 1
101 40 3
102 90 1
102 35 2
If you have any other doubt, you can mail me at
ankitgosain@gmail.com
Cheers,
Ankit :)
| Is This Answer Correct ? | 7 Yes | 1 No |
Answer / shar
If you think ur solution is correct just try out with below
input and recheck ur solution..
class_id,Marks
101,50
101,60
101,60
101,50
101,40
102,90
102,90
102,35
102,35
| Is This Answer Correct ? | 4 Yes | 0 No |
Answer / indu
Input->Sort1->Sort2->Trnsformer->Output
Sort1-->Declare class_id and Marks as key column and sort
in descding order.
Sort2-->Declare class_id and Marks as key column(Sort Mode
to Dont sort previously sorted for both) and set clustered
key change column to true.
Otuput for Sort 2 will be
class_id Marks Rank
102 90 1
102 35 0
101 60 1
101 50 0
101 40 0
In the Transformer declare stage variable temp and
initialize to o
Derive temp--> If Rnak=1 then Rank else temp+1
Derive output columns as --->
class_id ---> class_id
Marks ----> Marks
Rank----> temp
| Is This Answer Correct ? | 13 Yes | 10 No |
Answer / ankit gosain
Hi friends,
This question is a trick and as i found in every answer
given previously, there is a slight miss-queue.
Check out my solution for this question:
I/p File(Primary)
!
I/p
file---->Sort---->Sort---->transformer---->join/lookup---->O/p
File
(1). In the first sort stage, make class_id and marks fields
as keys, sort by descending and then create a Key change column.
(2). In second sort stage, make class_id only as a key
(don't sort previously sorted) and create a cluster key
change column.
(3). You will get the o/p as given below:
Class_Id, Marks, KeyChange, ClusterKeyChange
102,90,1,1
102,35,1,0
101,60,1,1
101,50,1,0
101,40,1,0
(4). In transformer stage take a integer type variable say
var1 and create one column say Rank.
Now, derive Var1 as :
if ClusterKeyChange=1 then 1 else Var1
And, derive Rank as :
if ClusterKeyChange=1 then 1 elseif (ClusterKeyChange=0 and
KeyChange=0) then Var1 else Var1+1
(5). Take only class_id, marks & rank as an o/p of the
transformer, so that you will have the o/p as given below:
Class_id, Marks, Rank
102,90,1
102,35,2
101,60,1
101,50,2
101,40,3
(6). Now finally, apply join/lookup and take this o/p as a
Reference data and Primary data will be the Source I/p File.
Take Class_id and Marks as keys and derive Rank to the o/p,
so that you will get the finally o/p as given below:
class_id Marks Rank
101 50 2
101 60 1
101 40 3
102 90 1
102 35 2
If you have any other doubt, you can mail me at
ankitgosain@gmail.com
Cheers,
Ankit :)
| Is This Answer Correct ? | 1 Yes | 0 No |
Answer / ankit gosain
Hi friends,
This question is a trick and as i found in every answer
given previously, there is a slight miss-queue.
Check out my solution for this question:
I/p File(Primary)
!
I/p
file---->Sort---->Sort---->transformer---->join/lookup---->O/p
File
(1). In the first sort stage, make class_id and marks fields
as keys, sort by descending and then create a Key change column.
(2). In second sort stage, make class_id only as a key
(don't sort previously sorted) and create a cluster key
change column.
(3). You will get the o/p as given below:
Class_Id, Marks, KeyChange, ClusterKeyChange
102,90,1,1
102,35,1,0
101,60,1,1
101,50,1,0
101,40,1,0
(4). In transformer stage take a integer type variable say
var1 and create one column say Rank.
Now, derive Var1 as :
if ClusterKeyChange=1 then 1 else Var1
And, derive Rank as :
if ClusterKeyChange=1 then 1 elseif (ClusterKeyChange=0 and
KeyChange=0) then Var1 else Var1+1
(5). Take only class_id, marks & rank as an o/p of the
transformer, so that you will have the o/p as given below:
Class_id, Marks, Rank
102,90,1
102,35,2
101,60,1
101,50,2
101,40,3
(6). Now finally, apply join/lookup and take this o/p as a
Reference data and Primary data will be the Source I/p File.
Take Class_id and Marks as keys and derive Rank to the o/p,
so that you will get the finally o/p as given below:
class_id Marks Rank
101 50 2
101 60 1
101 40 3
102 90 1
102 35 2
If you have any other doubt, you can mail me at
ankitgosain@gmail.com
Cheers,
Ankit :)
| Is This Answer Correct ? | 1 Yes | 0 No |
Answer / abhay agrawal
Source--Tx--Target
In Tx have 5 Stage Variables:
PreSID
CurrSID
PreMarks
CurrMarks
Rank
CurrSID=>PreSID
Input Sid=>CurrSID
CurrMarks=>PreMarks
Input Mark=>CurrMarks
If CurrSID=PreSID then If CurrMarks=PreMarks then Rank Else Rank+1 Else 1=>Rank
Take Rank as output column, let me know if any clarificationneeded.
| Is This Answer Correct ? | 0 Yes | 0 No |
Answer / anki_sri
Src>> Sort1 >> Sort2 >> Transfrmr >> Trgt
Sort1 --> sort with class_id and marks as well.
Sort2 --> declare key as id and select Dont Sort(Previously Sorted) and set create cluster key change column as True.
Transfrmr --> set two stage variables.
StageVar1=If Clusterkeychange=1 then Clusterkeychange else StageVar+1
StageVar=StageVar1
Create a new row as rank in transfrmr output and map StageVar1 to rank
| Is This Answer Correct ? | 3 Yes | 8 No |
Answer / amulya kumar panda
you are use
transformer if (Marks > 50 or 40 <Marks ) then rank=2 else
if Marks > 60 then rank=1 else rank=4
| Is This Answer Correct ? | 1 Yes | 7 No |
1.which index is follows the dimensions tables?why? 2.what is the use of trigger in job sequence? 3.what is the mean of optimization? 4.what is the job control?when we use it? what is difference bet batch and sequencer? 6.seq--->seq,seq--->copy--->seq which one is best and efficient?
explain about citrix scheduling tool in datastage
how to sort two columns in single job in datastage.
What is the diff between sort performed at sort stage and the stream sort performed at the input of few stages in DS Enterprise edition?
What is meta stage?
what is the differeces between hash and modulus partition methods
What is the difference between the local container and shared container?
Name the third party tools that can be used in datastage?
If we take 2 tables(like emp and dept),we use join stage and how to improve the performance?
how do u reduce warnings
how to find no.of records in sequntial file itself?
i have source like deptno,sal 1,2000 2,3000 3,4000 1,2300 4,5000 5,1100 i want target like target1 deptno,sal 1,2000 3,4000 4,5000 target2 2,3000 1,2300 5,1100 with out using transformerstage
Teradata 
