Skip to main content Skip to complementary content
Announcements

Denormalizing on multiple columns

Procedure

  1. Drop the following components: tFileInputDelimited, tDenormalize, tLogRow from the Palette to the design workspace.
  2. Connect all components using a Row main connection.
  3. On the tFileInputDelimited Basic settings panel, set the filepath to the file to be denormalized.
  4. Define the Row and Field separators, the Header and other information if required.
  5. The file schema is made of four columns including: Name, FirstName, HomeCity, WorkCity. 
    Name;FirstName;HomeCity;WorkCity
Pitt;Brad;Berverly Hills;Los Angeles
Pitt;Brad;Paris;London
Joli;Angelina;Berlin;Berlin
Joli;Angelina;Berlin;Los Angeles
Joli;Angelina;Los Angeles;Los Angeles
Willis;Bruce;Paris;Los Angeles
Willis;Bruce;Paris;Madrid
Willis;Bruce;Madrid;Paris
Willis;Bruce;Roma;Dublin
Moore;Demi;New York;Paris
Moore;Demi;Rio de Janeiro;Los Angeles
  6. In the tDenormalize component Basic settings, select the columns that contain the repetition.
    These are the column which are meant to occur multiple times in the document. In this example, FirstName, HomeCity and WorkCity are the columns against which the denormalization is performed.
  7. Click the to add rows and define the columns to denormalize.
  8. In the Delimiter column, define the separator between double quotes, to split concatenated values.
  9. Save your Job and press F6 to execute it.
    The denormalized values are concatenated using a semicolon. 
    |=-----+--------------------------+-------------------------+------------------------------=|
|Name  |FirstName                 |HomeCity                 |WorkCity                       |
|=-----+--------------------------+-------------------------+------------------------------=|
|Joli  |Angelina;Angelina;Angelina|Berlin;Berlin;Los Angeles|Berlin;Los Angeles;Los Angeles |
|Moore |Demi;Demi                 |New York;Rio de Janeiro  |Paris;Los Angeles              |
|Willis|Bruce;Bruce;Bruce;Bruce   |Paris;Paris;Madrid;Roma  |Los Angeles;Madrid;Paris;Dublin|
|Pitt  |Brad;Brad                 |Berverly Hills;Paris     |Los Angeles;London             |
'------+--------------------------+-------------------------+-------------------------------'
  10. To delete the duplicate occurrences, open the Basic settings of tDenormalize.
  11. Select the Merge same value check box.
  12. Save your Job again and press F6 to execute it.

Results

The duplicate occurrences are merged.
|=-----+---------+-----------------------+------------------------------=|
|Name  |FirstName|HomeCity               |WorkCity                       |
|=-----+---------+-----------------------+------------------------------=|
|Joli  |Angelina |Berlin;Los Angeles     |Berlin;Los Angeles             |
|Moore |Demi     |New York;Rio de Janeiro|Paris;Los Angeles              |
|Willis|Bruce    |Paris;Madrid;Roma      |Los Angeles;Madrid;Paris;Dublin|
|Pitt  |Brad     |Berverly Hills;Paris   |Los Angeles;London             |
'------+---------+-----------------------+-------------------------------'

Did this page help you?

If you find any issues with this page or its content – a typo, a missing step, or a technical error – let us know how we can improve!

Leave your feedback here