Translate Tableau table calculations to SQL windows.

What is the direct answer to this workbook question?

Many Tableau table calculations can be rewritten as SQL window functions when partitioning, addressing, sort order and frame bounds match. RUNNING_SUM commonly uses SUM() OVER, LOOKUP uses an offset window, and WINDOW_AVG uses AVG() OVER; missing dates and filter timing can change the result.

How do addressing and partitioning map to SQL?

Tableau's "Edit Table Calculation" dialog uses two terms that confuse SQL writers: Partitioning and Addressing (labeled "Compute Using" / "Specific Dimensions").

Here is the exact mechanical rule for translating any worksheet's table calculation into SQL:

1. First, aggregate the base measures at the grain of all Dimensions currently on the worksheet (Rows, Columns, Detail, Color) in a CTE.
2. Unchecked boxes in "Specific Dimensions" (Partitioning) become the columns inside PARTITION BY.
3. Checked boxes in "Specific Dimensions" (Addressing) become the columns inside ORDER BY.

Which SQL windows correspond to common table calculations?

Assuming your first CTE (view_grain) has already computed SUM(sales) AS total_sales grouped by region (partitioning) and order_month (addressing):

How can I calculate monthly growth and a rolling average?

In a Tableau workbook, you often see two table calculations layered on a monthly revenue sheet partitioned by Region:

// 1. MoM Growth %
(SUM([Sales]) - LOOKUP(SUM([Sales]), -1)) / ABS(LOOKUP(SUM([Sales]), -1))

// 2. 3-Month Moving Average
WINDOW_AVG(SUM([Sales]), -2, 0)

To get the exact same numbers in DuckDB, Postgres, or BigQuery SQL, aggregate to the worksheet grain first, then apply the window functions in the outer query:

WITH view_grain AS (
  SELECT
    region,
    DATE_TRUNC('month', order_date) AS order_month,
    SUM(sales) AS monthly_sales
  FROM orders
  GROUP BY region, DATE_TRUNC('month', order_date)
),
windowed AS (
  SELECT
    region,
    order_month,
    monthly_sales,
    LAG(monthly_sales, 1) OVER (
      PARTITION BY region
      ORDER BY order_month
    ) AS prev_month_sales,
    AVG(monthly_sales) OVER (
      PARTITION BY region
      ORDER BY order_month
      ROWS BETWEEN 2 PRECEDING AND CURRENT ROW
    ) AS rolling_3m_avg
  FROM view_grain
)
SELECT
  region,
  order_month,
  monthly_sales,
  (monthly_sales - prev_month_sales) / NULLIF(ABS(prev_month_sales), 0) AS mom_growth_pct,
  rolling_3m_avg
FROM windowed
ORDER BY region, order_month;

How do missing dates change a table calculation?

One subtle difference between Tableau and SQL: if a region has sales in January and March but zero rows in February, SQL LAG(monthly_sales, 1) on March will grab January's sales, and ROWS BETWEEN 2 PRECEDING AND CURRENT ROW will average 3 non-consecutive months.

If Tableau has "Show Missing Values" turned on for the date pill, your SQL query must CROSS JOIN a calendar spine of all months with the distinct regions before running LAG() or AVG() OVER().

What data does this worked example use?

The initial view uses authored synthetic data, not customer data or a captured agent session.

The example record lists its file, calculation or SQL and expected values. The live browser result is computed by xmetric’s existing engine or workbook inspector. Samples are excluded from own-data activation reporting.