An important analytical technique is to look at the rate of change over time. An easy graph to investigate this is a bar chart with the period over period change in percent or value. In this snippet, you can calculate the change for various time periods leveraging the built-in aggregation functions. If the user wants to use the aggregation filter, they can pass “aggregation” in the aggregation parameter. The “direction” field is for series to allow red/green or other types of color schemes to quickly indicate positives and negatives.

 Base: Bar Chart

• With period_over_period_direction as series
  • ‘positive’ color = #b5ca92
  • ‘negative’ color = #df5c5c

Name: period_over_period_change(table,field,date)

select
  date
  , [field]
  , case
    when lag([field]) over(order by date) is null
      then null
    else 1.0 *[field] / lag([field]) over(order by date) - 1
  end as period_over_period_change_perc
  , case
    when lag([field]) over(order by date) is null
      then null
    else [field] - lag([field]) over(order by date) - 1
  end as period_over_period_change_value
  , case
    when lag([field]) over(order by date) is null
      then null
    else (case when lag([field]) over(order by date) > [field] then 'negative' else 'positive' end)
  end as period_over_period_change_direction
from
  (
    select
      [field]
      , [date] as date
    from
      [table]
  )