Free tool runs in your browser

Tableau to SQL converter

Paste a calculated field. The xmetric engine runs it the way Tableau does, writes the SQL for it, then runs that SQL on the same rows and tells you whether the two agree. Nothing is uploaded.

Updated 2026-11-01 · Checked on the xmetric engine

Convert a formula

Paste a calculated field. Fields: [id], [Order Date], [Region], [Customer], [Category], [Sales], [Profit].

Tableau formula

In the view (for aggregates and level of detail) Order Date Region Customer Category

Use my CSV Sample: 10 orders

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

SELECT
  (CASE WHEN ("Profit" > 0.0) THEN 'Gain' ELSE 'Loss' END) AS result
FROM input

Profit	Tableau	SQL
18	Gain	Gain
−24	Loss	Loss
12	Gain	Gain
40	Gain	Gain
90	Gain	Gain
	Loss	Loss
−15	Loss	Loss
9	Gain	Gain
4	Gain	Gain
102	Gain	Gain

SQL notes and traps for each function: IF

The 10 sample rows

id	Order Date	Region	Customer	Category	Sales	Profit
1	2024-01-05	East	Ana Silva	Furniture	120	18
2	2024-01-19	East	Ben Ode	Technology	300	-24
3	2024-02-02	East	Ana Silva	Office	80	12
4	2024-02-14	West	Cai Lin	Furniture	200	40
5	2024-03-03	West	Dev Rao	Technology	450	90
6	2024-03-21	West	Cai Lin	Office	60	null
7	2024-04-08	South	Eve Moss	Technology	150	-15
8	2024-04-30	South	Eve Moss	Furniture	90	9
9	2024-05-12	North	Finn Kai	Office	40	4
10	2024-06-01	North	Gus Pell	Technology	510	102

What do common formulas look like in SQL?

Each one was converted by the engine and run both ways on the 10 sample rows.

How does an IF become a CASE?

Tableau

IF [Profit] > 0 THEN "Gain" ELSE "Loss" END

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

SELECT
  (CASE WHEN ("Profit" > 0.0) THEN 'Gain' ELSE 'Loss' END) AS result
FROM input

Profit	Tableau	SQL
18	Gain	Gain
−24	Loss	Loss
12	Gain	Gain
40	Gain	Gain
90	Gain	Gain
	Loss	Loss
−15	Loss	Loss
9	Gain	Gain
4	Gain	Gain
102	Gain	Gain

SQL notes and traps for each function: IF

How do I count weeks between dates like Tableau?

Tableau

DATEDIFF('week', [Order Date], #2024-06-01#)

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

SELECT
  CAST((floor((date_diff('day', DATE '1970-01-01', DATE '2024-06-01') + 4) / 7.0) - floor((date_diff('day', DATE '1970-01-01', CAST("Order Date" AS DATE)) + 4) / 7.0)) AS BIGINT) AS result
FROM input

Order Date	Tableau	SQL
2024-01-05	21	21
2024-01-19	19	19
2024-02-02	17	17
2024-02-14	15	15
2024-03-03	12	12
2024-03-21	10	10
2024-04-08	7	7
2024-04-30	4	4
2024-05-12	2	2
2024-06-01	0	0

SQL notes and traps for each function: DATEDIFF

How does ZN with a division translate?

Tableau

ZN([Profit]) / [Sales]

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

SELECT (coalesce("Profit", 0) / NULLIF("Sales", 0)) AS result
FROM input

Sales	Profit	Tableau	SQL
120	18	0.15	0.15
300	−24	−0.08	−0.08
80	12	0.15	0.15
200	40	0.2	0.2
450	90	0.2	0.2
60		0	0
150	−15	−0.1	−0.1
90	9	0.1	0.1
40	4	0.1	0.1
510	102	0.2	0.2

SQL notes and traps for each function: ZN

What is DATETRUNC by quarter in SQL?

Tableau

DATETRUNC('quarter', [Order Date])

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

SELECT
  CAST(date_trunc('quarter', "Order Date") AS DATE) AS result
FROM input

Order Date	Tableau	SQL
2024-01-05	2024-01-01	2024-01-01
2024-01-19	2024-01-01	2024-01-01
2024-02-02	2024-01-01	2024-01-01
2024-02-14	2024-01-01	2024-01-01
2024-03-03	2024-01-01	2024-01-01
2024-03-21	2024-01-01	2024-01-01
2024-04-08	2024-04-01	2024-04-01
2024-04-30	2024-04-01	2024-04-01
2024-05-12	2024-04-01	2024-04-01
2024-06-01	2024-04-01	2024-04-01

SQL notes and traps for each function: DATETRUNC

What does a profit ratio look like per region?

Tableau

SUM([Profit]) / SUM([Sales])

In the view: Region

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

WITH "marks" AS (
  SELECT "Region" COLLATE NOCASE AS "Region"
  FROM input
  GROUP BY "Region"
), "per_mark" AS (
  SELECT
    "Region" COLLATE NOCASE AS "Region",
    (sum("Profit") / NULLIF(sum("Sales"), 0)) AS "result"
  FROM input
  GROUP BY "Region"
)
SELECT m."Region" AS "Region", o1."result" AS "result"
FROM "marks" m LEFT JOIN "per_mark" o1 ON m."Region" IS NOT DISTINCT FROM o1."Region"
ORDER BY m."Region"

Text compares with COLLATE NOCASE because Tableau compares text without regard to case. Drop it if your data is all one case.

Region	Tableau	SQL
East	0.01	0.01
North	0.19	0.19
South	−0.03	−0.03
West	0.18	0.18

SQL notes and traps for each function: SUM

How does a FIXED level of detail become SQL?

Tableau

{ FIXED [Region] : SUM([Sales]) }

In the view: Region

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

WITH "lod_0" AS (
  SELECT
    "Region" COLLATE NOCASE AS "lod_key_0",
    sum("Sales") AS "value"
  FROM input
  GROUP BY "lod_key_0"
), "rows_1" AS (
  SELECT b.*, l."value" AS "lod_a81fe6"
  FROM input b LEFT JOIN "lod_0" l ON b."Region" COLLATE NOCASE IS NOT DISTINCT FROM l."lod_key_0"
), "marks" AS (
  SELECT "Region" COLLATE NOCASE AS "Region"
  FROM "rows_1"
  GROUP BY "Region"
), "per_mark" AS (
  SELECT
    "Region" COLLATE NOCASE AS "Region",
    MIN("lod_a81fe6") AS "result"
  FROM "rows_1"
  GROUP BY "Region"
)
SELECT m."Region" AS "Region", o1."result" AS "result"
FROM "marks" m LEFT JOIN "per_mark" o1 ON m."Region" IS NOT DISTINCT FROM o1."Region"
ORDER BY m."Region"

Text compares with COLLATE NOCASE because Tableau compares text without regard to case. Drop it if your data is all one case.

Region	Tableau	SQL
East	500	500
North	550	550
South	240	240
West	710	710

SQL notes and traps for each function: SUM FIXED

What is COUNTD in SQL?

Tableau

COUNTD([Customer])

In the view: Category

Run on 10 rows: this SQL returns the same values as the Tableau formula.

SQL (DuckDB)

WITH "marks" AS (
  SELECT "Category" COLLATE NOCASE AS "Category"
  FROM input
  GROUP BY "Category"
), "per_mark" AS (
  SELECT
    "Category" COLLATE NOCASE AS "Category",
    count(DISTINCT lower("Customer" COLLATE NOCASE)) AS "result"
  FROM input
  GROUP BY "Category"
)
SELECT m."Category" AS "Category", o1."result" AS "result"
FROM "marks" m LEFT JOIN "per_mark" o1 ON m."Category" IS NOT DISTINCT FROM o1."Category"
ORDER BY m."Category"

Text compares with COLLATE NOCASE because Tableau compares text without regard to case. Drop it if your data is all one case.

Category	Tableau	SQL
Furniture	3	3
Office	3	3
Technology	4	4

SQL notes and traps for each function: COUNTD

How do I link to this converter?

Copy this into your guide or answer:

<a href="https://getxmetric.com/reference/tableau-to-sql/">Tableau to SQL converter</a>

What do people ask about converting Tableau to SQL?

Which SQL does it write?

DuckDB SQL that reads a table named input. Most of it runs unchanged on Postgres. For Snowflake and BigQuery, each function's reference page gives the dialect notes.

Does it handle level of detail expressions?

Yes. FIXED, INCLUDE and EXCLUDE become common table expressions at the grain they ask for, joined back to the marks of the view you tick.

What about table calculations like RUNNING_SUM or RANK?

Those run over the marks of a view in a set order, so the converter shows the Tableau result and links each function's page, which gives the window-function SQL.

Why does the SQL say COLLATE NOCASE?

Tableau compares text without regard to case, so east and East are one region. The collation keeps that behaviour in SQL.

Is my file uploaded?

No. The engine and DuckDB run in your browser tab; your CSV never leaves your device.

Where is this documented?

Tableau functions (Tableau Desktop help) checked November 1, 2026
Level of detail expressions (Tableau Desktop help) checked November 1, 2026
DuckDB SQL introduction checked November 1, 2026