Ordering

class order.DataFrameOrdering

Ordering specification of a DataFrame.

__getitem__(rows)

Return a slice of DataFrame in the specified order.

Parameters: rows (slice) – slice: number of first rows
Returns: DataFrame with the rows in order
Return type: DataFrame

Example

>>> rows = [(4,), (3,), (2,), (1,)]
>>> df = db.create_dataframe(rows=rows, column_names=["id"])
>>> df.order_by("id")[:]
----
id
----
1
2
3
4
----
(4 rows)

order_by(column_name, ascending=None, nulls_first=None, operator=None)

Refine the order by adding another order definition to break the tie.

Parameters

column_name (str) – name of the column to order the GreenplumPython DataFrame by
ascending (Optional[bool]) – Optional[Bool]: Define ascending of order, True = ASC / False = DESC
nulls_first (Optional[bool]) – Optional[bool]: Define if nulls will be ordered first or last, True = First / False = Last
operator (Optional[str]) – Optional[str]: Define the order with the operator. Can’t be combined with ascending.

Returns

DataFrame ordered by the given arguments

Return type

DataFrameOrdering

Example

>>> rows = [(1, 2), (1, 3), (2, 2), (3, 1), (3, 4)]
>>> t = db.create_dataframe(rows=rows, column_names=["id", "num"])
>>> ret = t.order_by("id").order_by("num", ascending=False)[:5]
>>> t.order_by("id").order_by("num", ascending=False)[:]
----------
 id | num
----+-----
  1 |   3
  1 |   2
  2 |   2
  3 |   4
  3 |   1
----------
(5 rows)