Friday, March 23, 2012

excel data mining add-in for categorical data

Thank you very much for your previous answers conserning clustering algorithm.

I change a bit my data and now I can get two and three clusters as I wish.

But now I have a different question. I wish to keep my Structure and Model for further trials with different algorithm values, like cardinality. I tried the following

I selected in excel Data Mining and then the option cluster button.

Then in the cluster wizard I chose analysis service data source and I reform the query as I need adding only at the end

where "STAFF_YES" = 1. Then I choose the needed columns and target value. Finally I select browse model, enable drill through only and I finish the wizard and get some results, rather good. But i need to try again with different parameters.

I tried Business intelligent Studio to check my Analysis service. The structure were here but when I changed any parameter and I tried to reprocess I got the following messages

Processing Mining Structure 'try Structure_1' completed successfully.
Start time: 5/8/2007 4:55:08 μμ; End time: 5/8/2007 4:55:08 μμ; Duration: 0:00:00
Processing Dimension 'try Structure_1 ~MC-__RowIndex' completed successfully.
Start time: 5/8/2007 4:55:08 μμ; End time: 5/8/2007 4:55:08 μμ; Duration: 0:00:00
Processing Dimension Attribute '(All)' completed successfully.
Start time: 5/8/2007 4:55:08 μμ; End time: 5/8/2007 4:55:08 μμ; Duration: 0:00:00
Processing Dimension Attribute 'EPI_CELL_SINO_GOOD' completed successfully.
Start time: 5/8/2007 4:55:08 μμ; End time: 5/8/2007 4:55:08 μμ; Duration: 0:00:00
Processing Dimension Attribute 'EPI_CELL_SINO_LOOSE' completed successfully.
Start time: 5/8/2007 4:55:08 μμ; End time: 5/8/2007 4:55:08 μμ; Duration: 0:00:00
Errors and Warnings from Response
Internal error: The operation terminated unsuccessfully.
Internal error: The operation terminated unsuccessfully.
Internal error: An unexpected error occurred (file 'pcprocbinding.cpp', line 6645, function 'PCDBProcBinding:Tongue TiedelectOrChangeCartridge').
Errors in the OLAP storage engine: An error occurred while the dimension, with the ID of 'try Structure_1 ~MC-__RowIndex', Name of 'try Structure_1 ~MC-__RowIndex' was being processed.
Errors in the OLAP storage engine: An error occurred while the 'EPI_CELL_SINO_LOOSE' attribute of the 'try Structure_1 ~MC-__RowIndex' dimension from the 'DMAddinsDB2' database was being processed.
Internal error: The operation terminated unsuccessfully.
Internal error: The operation terminated unsuccessfully.
Internal error: An unexpected error occurred (file 'pcprocbinding.cpp', line 6645, function 'PCDBProcBinding:Tongue TiedelectOrChangeCartridge').
Errors in the OLAP storage engine: An error occurred while the dimension, with the ID of 'try Structure_1 ~MC-__RowIndex', Name of 'try Structure_1 ~MC-__RowIndex' was being processed.
Errors in the OLAP storage engine: An error occurred while the 'EPI_CELL_SINO_GOOD' attribute of the 'try Structure_1 ~MC-__RowIndex' dimension from the 'DMAddinsDB2' database was being processed.

And no change to the structure take place.

Please help me if you can

Thank you in advnce

Best regards.

Manolis

We discovered this issue and it will be addressed in a future release but currently there is no workaround for this problem using BI Dev Studio.

Here is an alternate way of doing the same thing:

- in Excel, under the Data mining tab, select the "Advanced\Add model to mining structure" option

- create a new model using whatever algorithm (Clustering, I assume) and setting any parameters you might want.

The wizard will create a new model inside the same structure and process it.

If having multiple models does not bother your, you can leave them all inside the structure. Otherwise, you can delete the old models using the Manage Models option in the Data Mining tab.

Also, after executing the steps you mentioned in the BI Dev Studio, chances are that your mining structure is in an incosistent state. You can restore its original (processed) state by invoking, under Manage Models, Process structure with new data then point to the data source you used in creating the structure the first time

Hope this helps

sql

No comments:

Post a Comment