SELECTCOLUMNS

A table's rows with the columns named, each a column or an expression.

Group: Table manipulation. Compatible with DAX.

Syntax

SELECTCOLUMNS ( <Table>, [<Name>,] <Expression> [, [<Name>,] <Expression>] … )

Parameter	Type	Description
`Table`	Table	The rows to go through: a table of the model or any table expression.
`Name`	Text optional, repeatable	The name of a column of the result. It may be left out before a plain column reference, which then keeps its own name.
`Expression`	Expression repeatable, evaluated for each row	The column's value, worked out for each row.

Returns

Table. One row for each row of the table, with only the columns given, in that order.

Remarks

The rows are not grouped, so duplicates remain. Wrap it in DISTINCT, or use SUMMARIZE or VALUES, to get each value once.
A plain column reference stays tied to its column of the model under its new name, so the result still filters that column in CALCULATE. Any other expression, even Products[Product] & "", gives plain values.
Every name must be different; reusing one is an error, whatever the case of its letters.
Unlike ADDCOLUMNS, it leaves out every column of the table you do not name.

Examples

A renamed column and a computed one

EVALUATE
VAR Orders =
    DATATABLE (
        "Product", STRING, "Liters", INTEGER, "Price", DOUBLE,
        { { "Pils", 120, 38 }, { "IPA", 80, 64 }, { "Cider", BLANK (), 52 }, { "Stout", 45, 70 } }
    )
RETURN
    SELECTCOLUMNS ( Orders, "Beer", [Product], "Revenue", [Liters] * [Price] )
ORDER BY [Beer]

Beer	Revenue
Cider	(blank)
IPA	5120
Pils	4560
Stout	3150

A column keeps its tie to the model

Both tables hold the three beers' names. The first still filters Products[Product]; the second holds plain text and filters nothing, so the total of all products comes back.

EVALUATE
VAR Beer = FILTER ( Products, Products[Category] = "Beer" )
RETURN
    ROW (
        "By the column", CALCULATE ( [Liters sold], SELECTCOLUMNS ( Beer, "Name", Products[Product] ) ),
        "By an expression", CALCULATE ( [Liters sold], SELECTCOLUMNS ( Beer, "Name", Products[Product] & "" ) )
    )

By the column	By an expression
1505389	1806463

A name left out

Places[Place] keeps its own name.

EVALUATE
SELECTCOLUMNS (
    TOPN ( 3, Places, Places[Crates], DESC, Places[Place], ASC ),
    Places[Place],
    "Crates", Places[Crates]
)
ORDER BY [Crates] DESC, Places[Place]

Places[Place]	Crates
Shop Grünerløkka	18
Shop Bjørvika	16
Shop St. Hanshaugen	15

ADDCOLUMNS, SUMMARIZE, DISTINCT, VALUES, ROW.

Last updated 3 November 2026 · OpenBI Studio 0.3.74