Discussions
Categories
Groups
Community Home
Categories
INTERNAL ENABLEMENT
POPULAR
PUBLIC CLOUD
PRIVATE CLOUD
Quick Links
MY LINKS
HELPFUL TIPS
Back to website
Home
Intelligence (Analytics)
Possible to compose some data cubes in one data cube or table?
ManInBlack
Hello all,
I have some scripted data sources, where some queries will be sent to different databases.
Now I would like to have one cross tab or table embedding the data sets (data cubes). The header of the table shall contain
the dates and the following rows the data (aggregations). I don't know if it is possible.
My approach was to build for each data set a data cube. Next, I created a grid and in each line
of the grid I put a single cross tab and pushed the data cubes inside.
Each cross tab contains the date in the header, therefore I hid the date cells from row two.
(The first row have to contain the date cells)
I uploaded a example. The example doesn't contain the query retrieving to dbs, instead it gets the data over java classes which have the data hardcoded.
Please copy the class files to
eclipse_birt\plugins\org.eclipse.birt.report.viewer_2.6.0.v20100605\birt\WEB-INF\classes
Are there any smarter solutions?
Thanks a lot.
Bye
ManInBlack
Find more posts tagged with
Comments
ManInBlack
I will explain it in easier words:
I have 3 "Scripted Data Sets". Each data set contains this following data structure:
StartDate | StopDate | Count
2010-02-01 | 2010-02-07 | 5
2010-02-07 | 2010-02-14 | 3
2010-02-14 | 2010-02-21 | 4
...
With the data cube or cross-tab it looks like following:
time 2010-02-01 - 2010-02-07 | 2010-02-07 - 2010-02-14 | 2010-02-14 - 2010-02-21
count 5 | 3 | 4
And now all three data cubes or cross-tabs together:
data-set 1
time 2010-02-01 - 2010-02-07 | 2010-02-07 - 2010-02-14 | 2010-02-14 - 2010-02-21
count 5 | 3 | 4
data-set 2
time 2010-02-01 - 2010-02-07 | 2010-02-07 - 2010-02-14 | 2010-02-14 - 2010-02-21
count 2 | 4 | 6
data-set 3
time 2010-02-01 - 2010-02-07 | 2010-02-07 - 2010-02-14 | 2010-02-14 - 2010-02-21
count 5 | 6 | 8
But my aim is to put all data in only one table with one header
(data-set 1 - 3)
time 2010-02-01 - 2010-02-07 | 2010-02-07 - 2010-02-14 | 2010-02-14 - 2010-02-21
count 5 | 3 | 4
count 2 | 4 | 6
count 5 | 6 | 8
(I hid the header of cross-tab 2 and 3)
This way is not really nice.
I hope that you can understand my problem.
Please could you help me, I spent already many days for finding a better solution, but still I am stuck.
Thank you
Bye,
ManInBlack
mwilliams
ManInBlack,
You could add a field to each dataSet to be able to make a row dimension out of, something like, Set1/Set2/Set3. Then, you could join the dataSets on the new field in full outer joins in a joint dataSet in BIRT. Then, you could create computed columns in your joint dataSet to merge the data into the same columns. Finally, you can use these new columns to create a single dataCube out of to display in a single crosstab in your report. The Set1/Set2/Set3 value from each dataSet that you merged into a single column in your joint dataSet will now be used as a row dimension to separate the values into their own rows.
Hope this helps. Let me know if you need further details.