I’m trying to create a rule for rule engine which searches a combined string for a particular value for a data set is quite messy in terms of string content .
for example: combined string row - “28000, male, employed at twitter”
rule: $combined string$ LIKE “employed”
to search for employed people
is it possible to create a rule with NOT LIKE i.e. NOT $combined string$ LIKE “employed”
to exclude rows where this word appears?
to exclude rows you could use a
Row Filter Node.
Thank you for your reply! Much appreciated.
I’m using a Rule Engine Dictionary on multiple healthcare data sets, which is why i’m using rules. As they will change depending on what set i’m working on.
In this case i want to add a rule to exclude a condition, additionally to a set of rules already being applied.
Hope this makes sense
Row Filter can also be provided with a dictionary. It is called
Rule-based Row Filter (Dictionary). If you want to use the filtered records further, a
Rule-based Row Splitter (Dictionary) could also be useful.
I hope it helps.
You need to include the wildcard characters in your string to get an inexact match e.g
"*employed*" to match “is employed at work”
. Usual wildcard characters apply ?*`.
To invert the match you can, as you suggest, use the
NOT clause. I tested
NOT both with and without brackets around the main clause and it appears that
LIKE is higher precedence than
NOT. In practice, I always put brackets around terms to ensure that processing is explicit e.g.
NOT ($column1$ LIKE "*employee*") => TRUE. This prevents any risk that the computer processes operators in a different sequence from the one I intended.
LinkedIn | Medium | GitHub
Thank you, will definitely look into this
Thank you! would of took me a long time to figure this out
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.