AddColumns, DropColumns, RenameColumns, and ShowColumns functions

Applies to: Canvas apps Desktop flows Model-driven apps Power Platform CLI

Shapes a table by adding, dropping, renaming, and selecting its columns.

Overview

These functions shape a table by adjusting its columns:

  • Reduce a table that contains multiple columns down to a single column for use with single-column functions, such as Lower or Abs.
  • Add a calculated column to a table (for example, a Total Price column that shows the results of multiplying Quantity by Unit Price).
  • Rename a column to something more meaningful, for display to users or for use in formulas.

A table is a value in Power Apps, just like a string or a number. You can specify a table as an argument in a formula, and functions can return a table as a result.

Note

The functions that this topic describes don't modify the original table. Instead, they take that table as an argument and return a new table with a transform applied. See working with tables for more details.

You can't modify the columns of a data source by using these functions. You must modify the data at its source. You can add columns to a collection with the Collect function. See working with data sources for more details.

Description

The AddColumns function adds a column to a table, and a formula defines the values in that column. Existing columns remain unmodified.

The formula is evaluated for each record of the table.

Fields of the record currently being processed are available within the formula. Use the ThisRecord operator or simply reference fields by name as you would any other value. The As operator can also be used to name the record being processed which can help make your formula easier to understand and make nested records accessible. For more information, see the examples below and working with record scope.

The DropColumns function excludes columns from a table. All other columns remain unmodified. DropColumns excludes columns, and ShowColumns includes columns.

Use the RenameColumns function to rename one or more columns of a table by providing at least one argument pair that specifies the name of a column that the table contains (the old name, which you want to replace) and the name of a column that the table doesn't contain (the new name, which you want to use). The old name must already exist in the table, and the new name must not exist. Each column name may appear only once in the argument list as either an old column name or a new column name. To rename a column to an existing column name, first drop the existing column with DropColumns, or rename the existing column out of the way by nesting one RenameColumns function within another.

The ShowColumns function includes columns of a table and drops all other columns. You can use ShowColumns to create a single-column table from a multi-column table. ShowColumns includes columns, and DropColumns excludes columns.

For all these functions, the result is a new table with the transform applied. The original table isn't modified. You can't modify an existing table with a formula. SharePoint, Microsoft Dataverse, SQL Server, and other data sources provide tools for modifying the columns of lists, tables, and tables, which are often referred to as the schema. The functions in this topic only transform an input table, without modifying the original, into an output table for further use.

The arguments to these functions support delegation. For example, a Filter function used as an argument to pull in related records searches through all listings, even if the '[dbo].[AllListings]' data source contains a million rows:

AddColumns( RealEstateAgents,
   "Listings",
   Filter(  '[dbo].[AllListings]', ListingAgentName = AgentName )
)

However, the output of these functions is subject to the non-delegation record limit. In this example, only 500 records are returned even if the RealEstateAgents data source has 501 or more records.

If you use AddColumns in this manner, Filter must make separate calls to the data source for each of those first records in RealEstateAgents, which causes a lot of network chatter. If [dbo](.[AllListings] is small enough and doesn't change often, you could call the Collect function in OnStart to cache the data source in your app when it starts. As an alternative, you could restructure your app so that you pull in the related records only when the user asks for them.

Syntax

AddColumns( Table, ColumnName1, Formula1 [, ColumnName2, Formula2, ... ] )

  • Table - Required. Table to operate on.
  • ColumnName(s) - Required. Name(s) of the column(s) to add. You must specify a string (for example, "Name" with double quotes included) for this argument. You must specify the column name(s) using the its logical name. Intellisense will only suggest logical names. The logical name is case sensitive.
  • Formula(s) - Required. Formula(s) to evaluate for each record. The result is added as the value of the corresponding new column. You can reference other columns of the table in this formula.

DropColumns( Table, ColumnName1 [, ColumnName2, ... ] )

  • Table - Required. Table to operate on.
  • ColumnName(s) - Required. Name(s) of the column(s) to drop. You must specify a string (for example, "Name" with double quotes included) for this argument. You must specify the column name(s) using the its logical name. Intellisense will only suggest logical names. The logical name is case sensitive.

RenameColumns( Table, OldColumnName1, NewColumnName1 [, OldColumnName2, NewColumnName2, ... ] )

  • Table - Required. Table to operate on.
  • OldColumnName - Required. Name of a column to rename from the original table. This element appears first in the argument pair (or first in each argument pair if the formula includes more than one pair). This name must be a string (for example "Name" with double quotation marks included). You must specify the column name(s) using the its logical name. Intellisense will only suggest logical names. The logical name is case sensitive.
  • NewColumnName - Required. Replacement name. This element appears last in the argument pair (or last in each argument pair if the formula includes more than one pair). You must specify a string (for example, "Customer Name" with double quotation marks included) for this argument. You must specify the column name(s) using the its logical name. Intellisense will only suggest logical names. The logical name is case sensitive.

ShowColumns( Table, ColumnName1 [, ColumnName2, ... ] )

  • Table - Required. Table to operate on.
  • ColumnName(s) - Required. Name(s) of the column(s) to include. You must specify a string (for example, "Name" with double quotes included) for this argument. You must specify the column name(s) using the its logical name. Intellisense will only suggest logical names. The logical name is case sensitive.

Examples

The examples in this section use the IceCreamSales data source, which contains the data in this table:

IceCream example.

None of these examples modify the IceCreamSales data source. Each function transforms the value of the data source as a table and returns that value as the result.

Formula Description Result
AddColumns( IceCreamSales, "Revenue", UnitPrice * QuantitySold ) Adds a Revenue column to the result. For each record, UnitPrice * QuantitySold is evaluated, and the result is placed in the new column. Result with Srawberry, Chocolate and Vanilla.
DropColumns( IceCreamSales, "UnitPrice" ) Excludes the UnitPrice column from the result. Use this function to exclude columns, and use ShowColumns to include them. Result with Strawberry, Chocolate, Vanilla having only QuantitySold column.
ShowColumns( IceCreamSales, "Flavor" ) Includes only the Flavor column in the result. Use this function include columns, and use DropColumns to exclude them. Only Flavor column.
RenameColumns( IceCreamSales, "UnitPrice", "Price") Renames the UnitPrice column in the result. Result with Flavor, Price and Revenue.
RenameColumns( IceCreamSales, "UnitPrice", "Price", "QuantitySold", "Number") Renames the UnitPrice and QuantitySold columns in the result. Result with 3 IceCreams and columns as Flavor, Price, Revenue.
DropColumns(
RenameColumns(
AddColumns( IceCreamSales, "Revenue",
UnitPrice * QuantitySold ),
"UnitPrice", "Price" ),
"Quantity" )
Performs the following table transforms in order, starting from the inside of the formula:
  1. Adds a Revenue column based on the per-record calculation of UnitPrice * Quantity.
  2. Renames UnitPrice to Price.
  3. Excludes the Quantity column.
Note that order is important. For example, we can't calculate with UnitPrice after it has been renamed.
IceCream example for unit price.

Step by step

Let's try some of the examples from earlier in this topic.

  1. Create a collection by adding a Button control and setting its OnSelect property to this formula:

    ClearCollect( IceCreamSales,
       Table(
           { Flavor: "Strawberry", UnitPrice: 1.99, QuantitySold: 20 },
           { Flavor: "Chocolate", UnitPrice: 2.99, QuantitySold: 45 },
           { Flavor: "Vanilla", UnitPrice: 1.50, QuantitySold: 35 }
       )
    )
    
  2. Run the formula by selecting the button while holding down the Alt key.

  3. Add a second Button control, set its OnSelect property to this formula, and then run it:

    ClearCollect( FirstExample,
       AddColumns( IceCreamSales, "Revenue", UnitPrice * QuantitySold )
    )
    
  4. On the File menu, select Collections, and then select IceCreamSales to show that collection.

    As this graphic shows, the second formula didn't modify this collection. The AddColumns function used IceCreamSales as a read-only argument; the function didn't modify the table to which that argument refers.

    Collection viewer showing three records of the Ice Cream Sales collection that doesn't include a Revenue column.

  5. Select FirstExample.

    As this graphic shows, the second formula returned a new table with the added column. The ClearCollect function captured the new table in the FirstExample collection, adding something to the original table as it flowed through the function without modifying the source:

    Collection viewer showing three records of the First Example collection that includes a new Revenue column.

Map columns in a component

See Map columns.