Rank

Create a ranking of values in a column.

Parameters

NameTypeDescriptionIs Optional

rank_columns

column_list

The column or expression to order (rank) by

partition_by

column_list

The column or expression to partition the window by. Accepts .

True

order

sort_direction

Instruction on how to order the rank_column. Accepts , but defaults to ASC. ASC = Ascending DESC = Descending

True

rank_type

value

Type of ranking [ | dense | percent | unique ]. Accepts . = ANSI SQL rank function dense = ANSI SQL dense_rank function percent = ANSI SQL percent_rank function unique = ANSI SQL row_number function

True

qualify_filter

value

A filter to apply to the ranked column. Accepts .

True

alias

string

Name of the output column containing rankings

True

overwrite_columns

boolean

Optional: if true, the columns passed in 'rank_columns' will be dropped from the output

True

Example

#Return the 5 highest temperatures per date:
ds = rasgo.get.dataset(dataset_id)

ds2 = ds.rank(rank_columns=['DS_DAILY_HIGH_TEMP'],
      partition_by=['DATE'],
      ORDER='DESC',
      RANK_TYPE='unique',
      QUALIFY_FILTER='>=5'
})

ds2.preview()

Source Code

Last updated