I am trying to perform null validation on a table

Pooja_Agrawal

null validation has to performed on the table consisting of more than one column. Then storing it in two different tables in SQL Server such that records which pass the validation that is the records that are not null must get stored in one table and the records which fail the validation should get stored in separate table with proper error message.
Please help me to do this. I have used "Filter Examples Operator" and "Filter Examples with missing values Operator" but none of them are generating output as required. The null values are also coming.

BalazsBaranyRM

Hi!

Filter Examples is a good operator for separating data with and without missing values (that's how the statistics community calls null values).

The easiest way to separate with Filter Examples is by selecting "no missing attributes" as the condition class parameter. You get the matches (rows without missing attribute values) on the "exa" output and the rows with missing values from the "unm" (unmatched) output. 

So you could already store the two data sets into two different tables using Write Database. However, you need a "proper" error message. What makes an error message proper in your use case? Is "Has missing values" enough or do you need "Attribute X has a missing value"? Do you need it for just one attribute or all missing attribute values for each row?

Loop Attributes and Generate Attributes would be helpful for the more detailed error message. For example, in Generate Attributes you would use the missing() function in an if() construct to generate the error message or append to the existing one.

Regards,
Balázs