Calculate a percentage value based on time and value criteria in LookML

Question

Im trying to calculate a percentage value. Where the dividend is the amount of rows where a timestamp is within a certain range, as well as a column 'customer' is equal to 2. The divisor is the total amount of rows where 'customer' is equal to 2. No timeframe requirement. I tried SAFE_DIVIDE as well as making several measures and calculating the rows in Looker report instead but usually get an error.

Any idea what a solution could be? Is it possible to do in one measure or easier to split into several measures with type:sum, type: percent_1 and so on?

I believe Im using wrong datatypes in below example. Example code:

measure: customer_divided {
    type: number
    value_format_name: percent_1
    view_label: "view_label text"
    label: "label text"
    description: "descr text"
    sql:SAFE_DIVIDE((${Filtered_timeframe_value}),(${TABLE}.customer="2")+;;

}

Answer 1

One possible solution could be to create 2 measures and then use them to create the "customer_divided" measure on the explore view itself.

measure: customer_divided_numerator {
    type: number
    value_format_name: percent_1
    filter : [<timestamp filter>, customer:2]
}
measure: customer_divided_denominator {
    type: number
    value_format_name: percent_1
    filter : [customer:2]
}

Once you have created these measures, then you can use them to create the customer_divided measure.