I created measure with a filter and the field I use in that expression is a boolean type with true/false values, however the field contains null values as well.
here’s how I define the measure:
measure: aged_25_34 {
type: count
filters: {
field: age
value: "TRUE"
}
}
The query returns:
No matching signature for function UPPER for argument types: BOOL. Supported signatures: UPPER(STRING); UPPER(BYTES)
Is there any hack to get around this?
+1
Age is a boolean here? By boolean, do you mean it’s a lookml dimension of type: yesno? If so, you actually say
measure: aged_25_34 {
type: count
filters: {
field: age
value: "Yes"
}
}
instead of TRUE. That might help!
Hi string and the values are shown as:
Looker translates my filter into this expression that expects the dimension to be of type boolean.
SELECT
COUNT(CASE WHEN (UPPER(age) = UPPER('Yes')) THEN 1 ELSE NULL END) AS customer_profile_age
FROM customer AS customer_profile
LIMIT 500
still returning the same error
+1
Ooookay my wager here is that somehow, case_sensitive: is being applied to the age field in the lookml. That’s the only thing in Looker that’d automatically insert those UPPER clauses.
Can you verify that there’s a case_sensitive: no parameter being applied somewhere in the reference chain there, and try removing it? If you’d like to maintain the boolean status of that field, you might look into using a yesno type:
🙂
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.