Qlik Community

Ask a Question

QlikView App Dev

Discussion Board for collaboration related to QlikView App Development.

Announcements
Join us at the Cloud Data and Analytics Tour! REGISTER TODAY
cancel
Showing results for 
Search instead for 
Did you mean: 
Not applicable

join and outerjoin in unique QVD file result


Hi,

Can you help me please,

I receive each month two files xlsx (fileA, FileB) and I want to store in one QVD file:

1) the intersection of the two files (using unique ID)

2) the records of fileB not present in fileA

like this :

fileA

ID

MR

1

1

2

4

3

8

fileB

ID

MR

1

2

2

5

4

15

file result in one QVD?

ID

MR

data source from

1

1

fileA

2

4

fileA

4

15

fileB

thanks for your help,

Thierry

1 Solution

Accepted Solutions
Nicole-Smith

Sure it is.  Change this:

Final:

NOCONCATENATE LOAD *

RESIDENT FileA

WHERE Exists(IDX,ID);

CONCATENATE (Final)

LOAD IDX as ID, MR, Amount, FilenameB, FiledateB

RESIDENT FileB

WHERE not Exists(ID,IDX);

DROP TABLES FileA, FileB;

store Final into Final.qvd (QVD)

to

Final:

NOCONCATENATE LOAD *

RESIDENT FileA

WHERE Exists(IDX,ID);

CONCATENATE (Final)

LOAD IDX as ID, MR, Amount, FilenameB as Filename, FiledateB as Filedate

RESIDENT FileB

WHERE not Exists(ID,IDX);

DROP TABLES FileA, FileB;

store Final into Final.qvd (QVD)

View solution in original post

6 Replies
datanibbler
Champion
Champion

Hi thiery,

the issue of JOINs is always the same, regardless of the exact scripting_language or whatever you use:

=> The best way to visualize it IMHO is two circles that have an intersection.

    => If you want just that intersection (present in circle_A AND in circle_B) -> inner_join (standard in QV)

   => If you want the part of circle_A outside the intersection (present in A, BUT NOT in B) -> left_outer/ left

    => If you want the part of circle_B outside the intersection (present in B, BUT NOT in A) -> right_outer/ right

    => If you want the parts of both circles, but not the intersection (present EITHER in A or in B) -> full_outer

HTH

Best regards,

DataNibbler

Nicole-Smith

This should do the trick:

FileA:

LOAD *, 'File A' as [Data Source] INLINE [

ID,MR

1,1

2,4

3,8

];

FileB:

LOAD ID as IDB, MR, 'File B' as [Data Source] INLINE [

ID,MR

1,2

2,5

4,15

];

Final:

NOCONCATENATE LOAD *

RESIDENT FileA

WHERE Exists(IDB,ID);

CONCATENATE (Final)

LOAD IDB as ID, MR, [Data Source]

RESIDENT FileB

WHERE not Exists(ID,IDB);

DROP TABLES FileA, FileB;

Example file is also attached.

simondachstr
Luminary Alumni
Luminary Alumni

DataNibbler said it all. Make sure to use different table names for each and use NoConcatenate to avoid the tables automatically merging again after the load because of the same field names.

Not applicable
Author

Thanks a lot,

1) If I good understand, in the line "LOAD IDB as ID, MR, [datasource]" the datasource can be replaced by any field from fileB ?

2) I tried to write the script. I take in account the name of the two files like date and filename. How can I put the filedateB and filenameB in the field filedate and filename ?

My script :

FileA :

LOAD

    filename() as Filename,

    DATE(DATE#(left(right(filename(),11),6),'YYYYMM'),'MM/YYYY') as Filedate,

    ID,

    MR,

    Amount

FROM

[201202.xlsx]

(ooxml, embedded labels, table is Sheet1);

store FileA into FileA.qvd (QVD);

FileB :

LOAD

    filename() as FilenameB,

    DATE(DATE#(left(right(filename(),12),6),'YYYYMM'),'MM/YYYY') as FiledateB,

    ID as IDX,

    MR,

    Amount

FROM

[201202X.xlsx]

(ooxml, embedded labels, table is Sheet1);

store FileB into FileB.qvd (QVD);

Final:

NOCONCATENATE LOAD *

RESIDENT FileA

WHERE Exists(IDX,ID);

CONCATENATE (Final)

LOAD IDX as ID, MR, Amount, FilenameB, FiledateB

RESIDENT FileB

WHERE not Exists(ID,IDX);

DROP TABLES FileA, FileB;

store Final into Final.qvd (QVD)

result :

IDMRfiledatefilenameAfiledateBfilenameB
1102/2012201202.xlsx
2402/2012201202.xlsx
415 02/2012201202X.xlsx

is it possible to have in the final :

IDMRfiledatefilename
1102/2012201202.xlsx
2402/2012201202.xlsx
41502/2012201202X.xlsx

Nicole-Smith

Sure it is.  Change this:

Final:

NOCONCATENATE LOAD *

RESIDENT FileA

WHERE Exists(IDX,ID);

CONCATENATE (Final)

LOAD IDX as ID, MR, Amount, FilenameB, FiledateB

RESIDENT FileB

WHERE not Exists(ID,IDX);

DROP TABLES FileA, FileB;

store Final into Final.qvd (QVD)

to

Final:

NOCONCATENATE LOAD *

RESIDENT FileA

WHERE Exists(IDX,ID);

CONCATENATE (Final)

LOAD IDX as ID, MR, Amount, FilenameB as Filename, FiledateB as Filedate

RESIDENT FileB

WHERE not Exists(ID,IDX);

DROP TABLES FileA, FileB;

store Final into Final.qvd (QVD)

View solution in original post

Not applicable
Author

Thanks a lot,

It is perfect.

Thierry