Monday, April 6, 2009

Clementine is dead, long live PASW Modeller

For the new Clementine homepage see

SPSS have gone for new product names, including changing Clementine to PASW. I'm more interested in the new features and bug fixes than buzz words. I'll hopefully be getting the new version shortly and will let you know if Clementine 13 (aka Predictive Analytics Soft Ware Modeller) adds value.


Dominic F said...

Hi Tim. We will also be getting PASW in our company in the next couple of weeks. I think it is an evolutionary release, but still a few things that should make life easier: sticky notes for comments, automated data preparation, preview of imported files, auto cluster node.

Call me boring, but I am quite excited!

Tim Manns said...

For me any SQL pushback (better integration of features into SQL) improvements would probably have a big impact. I'm keen to see what's be added in this area.

I've logged plenty of enhancements over the years. For example, I'd like to see improvements to the merge node (joins) that allows between joins for example (where rows are matched if the value of one field is between the values two other fields). Stuff like that currently has to be done by direct sql. Not a problem, but would be nice if it were added.

Anonymous said...

Hi Tim!

Can you tell me if is possible in version 11 to call from visual basic (for example) a previous constructed Clementine stream and give to that streams some parameters like the file with data?
Thanks in advance.

Jorge from Portugal

Tim Manns said...

Hello Jorge,

Yes it is :)

Probably the easiest way is to use Clementine batch in a DOS command line.

Clementine can be run in the background, accepting values for predefined parameters such as a date, data file, field name etc

From memory to run a simple stream without the user interface would be something like this;

clemb -stream "your_stream.str" -execute

If you want to use a stream parameter that you have already created, then use -P to call the parameter and an equal sign to set the value of that parameter. So for example if in your stream you have a parameter named "your_parameter" which contains a text string value for your data file then;

clemb -stream "your_stream.str" -Pyour_parameter=c:\\datafile.txt -execute



Anonymous said...

Thank you very much Tim!

Best regards,

Jorge Duarte

Pedro said...

HI Tim,
Do you got the new version? How can i get one to evaluate?

Pedro said...

Do you fill mutch differents!?

Ashok Bottu said...

Hi Tim,
Iam using PASW Modeler 13 version and making use of it for predicting fraudulent nature. I have done building a model using C5.0 and applied new customer data to that model, it is going to give $C-Fraud and $CC-Fraud. Am, i doing in right way for predicting customers fraud nature?
How can we say, which modeling algorithm is best suitable to our application? As per my understanding C5.0 modeling algorithm which is giving much accuracy for my data. Please share your information about this topic.

Thanks in Advance,

Anonymous said...

Hi Tim, I new in Clementine and your blog and your forum are being very important for me. I think your forum is not working yet, is it? And I have a question very urgent for me and I would apreciate any help.

I have a stream that execute in batch mode and i need to schedule its execution. Could you give an example of a scheduled module? I can't find any help about this on the web.

Many thanks for your help.
Hisai- from Spain