Log for node: The partition for the project "survival" could not be completed.

[mrahim]

am trying to build a model for "titanic" dataset like logistic regression. but am getting the following error; The partition for the project "survival" could not be completed. note that; i change 4 devices and send queries for many time. my account is student account granted by the university of Wollongong in Dubai. Appreciate your support so i complete my assignments within the deadlines.

[tom_grant]

Usually this is caused by incorrect target variable assignment- make sure only one variable is assigned to role of Target - if that is not it - please send the deployment number (in url) & the table you are using

[mrahim]

Thank you for your response. i have confirmed that i choose the correct variable as target (survived) and no other variable role as target and still i have same partition issue.

project url is : https://vfl-025.engage.sas.com/SASModelStudio/?projectUrl=/analyticsGateway/projects/dc74d03d-0846-4...

i dont how i can share the table.

one more question SAS should add partition column. i that is correct am not seeing SAS creating and new column.

[antonbcristina]

Hi @mrahim, I recommend using the ACADEMIC.TITANIC dataset rather than others that might be available in the PUBLIC library. I've had issues trying to user others and the TITANIC dataset in the ACADEMIC library is the source table. If you're following along with the traditional story line/activity that uses that dataset, you'll want to select Survived as your target variable. I've been able to recreate your error by incorrectly selecting Sex as the target and matching your partition and event-based sampling options. That indicates to me that the other variables in the data are not good target candidates based on the project settings you've made. 

[mrahim]

i tried academic titanic. still same problem 

[tom_grant]

The partition is created for you automatically - do you have another variable with role of Partition? - that will cause an issue if you use the default settings.

[mrahim]

Now the variables same as the attachment, no variable assigned as partition. and i run the pipeline and again same problem

[tom_grant]

If you are still having trouble - turn off the Event Based sampling - no need for that with this small data table.

[tom_grant]

I would say, start over with a brand new project, keep all the default settings, & just make Survival the Target

[mrahim]

Many thanks for your support. it works when i turned of the event based. however. what is the solution if i have bigger table