I needed to write a filter that used an OR statement to pull the department values from one explore join (the IN part of the referral) OR the another explore join (the OUT part of the referral). Unfortunately, this functionality just doesn't exist in Looker. I had to find another way. Here's my explore with 5 joins. The last join contains the OR statement: the final solution to this problem. The first 4 joins set up the in and out assignment of departments for both the referral and the revenue from the referral. explore: flatfile { join: referred_to_dept { view_label: "referred to department" from: v_dim_dept type: left_outer relationship: one_to_one sql_on: ${flatfile.referred_to_dept_id} = ${referred_to_dept.branch};; } join: referrers_dept { view_label: "referrer department" from : v_dim_dept type: left_outer relationship: one_to_one sql_on: ${flatfile.referrers_dept_id} = ${referrers_dept.branch};; } jo
Each bar is created with a separate dimension, because the periods we are tracking overlap. The problem is that in the data table, the periods repeat, which creates a weird Axis Value Label, Figure 1. Workaround with not-great results. To fix, add a Table Calculation to the data in Looker and hide the dimensions that were making the weird axis value labels in Figure 1. Axis Label if(length(${dim_week.4_week})>1, "4 week", if(length(${dim_week.13_week})>1, "13 week", "YTD"))