every time i create a dataset batch_id and _batch_last_run columns are generating and while joining

Hi,

every time i create a dataset batch_id and _batch_last_run columns are generating and while joining datasets an error is displaying bcoz of these duplicate columns present in both datasets

is there a way to create a dataset without generating these columns or how to delete it

and one more question 

is it possible to change column names after creating a dataset

thank you

Best Answers

  • guitarhero23
    Accepted Answer

    What type of join are you doing? How are you performing it, ETL, SQL dataflow, fusion? I ask because you should be able to join with both datasets having them, they are standard in Domo and I've yet to run into it. MySQL dataflow join example below

     

    Table 1: "Favorites"

    • ID
    • Favorite_Color
    • Favorite_Food
    • UserID
    • Batch_ID
    • Batch_Last_Run

     

    Table 2: "Users"

    • UserID
    • Name
    • Email
    • Batch_ID
    • Batch_Last_Run

     

    Despite those columns being duplicate if you wanted to join on them you could do

     

    SELECT 

    f.Favorite_Color

    ,f.Favorite_Food

    ,u.Name

    ,u.Email

    FROM Users as u

    LEFT JOIN Favorites as 'f' on f.UserID = u.UserID

     

     

    This shouldn't cause any errors. If you did want to include both batchID and batch last run from both datasets you could just rename them in the select statement

     

    SELECT 

    f.Favorite_Color

    ,f.Favorite_Food

    ,f.BatchID as 'Favorites BatchID'

    ,f.Batch_Last_Run as 'Favorites Batch Last Run'

    ,u.Name

    ,u.Email

    ,u.BatchID as 'UsersBatchID'

    ,u.Batch_Last_Run as 'USers Batch Last Run'

    FROM Users as u

    LEFT JOIN Favorites as 'f' on f.UserID = u.UserID

     

     

    This should also be possible in ETL by perhaps putting a select columns type of ETL block so you don't even select them prior to joining

  • ST_-Superman-_
    Accepted Answer

    You could also use the SQL transform option to drop the columns.  (you can only drop one column at a time though)

     

    Create a transform, select the SQL option:

    ALTER TABLE `table_name` DROP COLUMN `_BATCH_ID_`;

     

    then drop the other batch field:

    ALTER TABLE `table_name` DROP COLUMN `_BATCH_LAST_RUN_`

     

    You can then use 

    SELECT a.*, b.* from `table_name` a left join `table_b_name` b on a.`id`=b.`id`

     

    without a duplicat column.

     

    Alternatively, you can also remove the columns by clicking on the input data set and clicking on the 'x' to the right (this only works if your field names don't have a . in them):

     

    1.png

     

    2.png