Skip to main content

RANK Function

Computes the rank of an ordered set of value within groups. Tie values are assigned the same rank, and the next ranking is incremented by the number of tie values.

  • Rank values start at 1 and increment.

  • Ranking order varies depending on the data type of the source data.

  • You must use the group and order parameters to define the groups of records and the ordering column to which this function is applied.

  • This function works with the following transforms:

  • This function assigns ranking values to match the total number of rows in a group. If needed, tie values can be assigned the same rank. For more information, see DENSERANK Function.

Wrangle vs. SQL: This function is part of Wrangle, a proprietary data transformation language. Wrangle is not SQL. For more information, see Wrangle Language.

Basic Usage

rank()

Output: Generates the new column, which contains the ranking of mySales, grouped by the Salesman column.

Syntax and Arguments

rank() order: order_col group: group_col

For more information on the orderand groupparameters, see Window Transform.

For more information on syntax standards, see Language Documentation Syntax Notes.

Examples

提示

For additional examples, see Common Tasks.

Example - Rank Functions

This example demonstrates you to generate a ranked order of values.

Functions:

Item

Description

RANK Function

Computes the rank of an ordered set of value within groups. Tie values are assigned the same rank, and the next ranking is incremented by the number of tie values.

DENSERANK Function

Computes the rank of an ordered set of value within groups. Tie values are assigned the same rank, and the next ranking is incremented by 1.

Source:

The following dataset contains lap times for three racers in a four-lap race. Note that for some racers, there are tie values for lap times.

Runner

Lap

Time

Dave

1

72.2

Dave

2

73.31

Dave

3

72.2

Dave

4

70.85

Mark

1

71.73

Mark

2

71.73

Mark

3

72.99

Mark

4

70.63

Tom

1

74.43

Tom

2

70.71

Tom

3

71.02

Tom

4

72.98

Transformation:

You can apply the RANK() function to the Time column, grouped by individual runner:

Transformation Name

Window

Parameter: Formulas

RANK()

Parameter: Group by

Runner

Parameter: Order by

Time

You can use the DENSERANK() function on the same column, grouping by runner:

Transformation Name

Window

Parameter: Formulas

DENSERANK()

Parameter: Group by

Runner

Parameter: Order by

Time

Results:

After renaming the columns, you have the following output:

Runner

Lap

Time

Rank

Rank-Dense

Mark

4

70.63

1

1

Mark

1

71.73

2

2

Mark

2

71.73

2

2

Mark

3

72.99

4

3

Tom

2

70.71

1

1

Tom

3

71.02

2

2

Tom

4

72.98

3

3

Tom

1

74.43

4

4

Dave

4

70.85

1

1

Dave

1

72.2

2

2

Dave

3

72.2

2

2

Dave

2

73.31

4

3