combining cubes into one cube
Dear community,
I have a problem and need suggestion to solve it. I am working on a project that will show projects status that the team members are working for. I've developed an input screen for project information. This screen includes data entry areas like status, person in charge, priority. These areas are defined as cubes with project id dimension.
Now I need to build a new project cube with status & person in charge & priority dimensions. This cube will include project name. But I don't know how can I fill this cube with the entered information.
Do you have any idea? How can I do that? I tried to use data loader but I couldn't succeeded.
I will be grateful with your answers. Thank you.
Answers
-
Hi Selvet Yönsel,
do I understand you correctly that in stage one you have a screen with
* status (defined as text cube)
* person in charge (defined as text cube)
* priority (defined as text cube)
with project-id on the rows axis.
and want to use these cubes as input for entities?
If so, you could for example use rolap-cubes for status, person and priority. The dataentry will then write the user input into a SQL table.Then you could create a datareader for the entities, with as input the SQL tables.
BR, Ray
0 -
If what you want is to have suggested values for a field such as person, priority etc... (in the example below it is responsibility). Then you can define suggested values which can be written or existing entities.
If what you want is to have all the information in one cube. You can create a text cube with a dimension (project info)
Project info
-person responsible
-status
-etc
And basically take the information in to each one of the elements (slices) of that cube with a dataflow. This is well explained in the E-Learning and if you need further help you can ask, but basically you would create 3 dataflows one for each and define the target entity element (person responsible for example).
0 -
Hi Ray Martens,
Thank you for your reply. I've created RDB cube and defined a csv file as a source. I defined column maps and tried to enter value. But I failed. I checked the db log but cannot see anything.
Do I need to define something to be able to enter data to csv file?
I will be grateful with your answer. Thank you.
0 -
I don't know about RDBs with CSV files and the problems that may arise with it. Are you sure the file permissions are set correctly? Have you tried writing to that file through the OLE/ODBC driver you're using with a different program such as Excel?
I remember that for the SQL connection it is necessary for the SQL user used in the connection to have the rights to write into the table.
BR, Ray
0 -
Hi Miguel Angel Sisternas Seaton,
Thank you for your answer.
I have developed screen to enter project data and also I've defined configuration for suggested values. What I want to is:
-
to make a single data reader that reads data from different excel/csv/txt files and put data into one analyzer cube?
For example, I have data1.csv, data2.csv, data3.csv files. Both have id dimension. Each of them has fact columns like data1 has fact1 column, data2 has fact2 column and data3 has fact3 column.
I want to join all 3 files and write data into cube (with the structure id, fact1, fact2, fact3).
Is it possible?
I will be really grateful with your answer.
1 -
Try putting them all in the same folder creating the data reader with the path to that folder and then executing it as a procedure set for all.
0 -
Dear community,
I have solved this problem like this:
I have defined a schema.ini file on the server with the same path with my source text files. Then, I have defined my source text files in that ini doc..
Here is an example definition of source file in ini :
-----------------------------------------
[SourceFileName.txt]
ColNameHeader=True
Format=TabDelimited
MaxScanRows=25
CharacterSet=OEM
TextDelimiter=^
Col1=colname1 Char Width 255
Col2=colname2Char Width 255-------------------------------------------------
(Of course, I also need to define connection on the server for that path.)
Thank you for your replies and supports.
0
