rolling#

class pylibcudf.rolling.BoundedClosed#

A bounded closed window.

This window contains rows with delta of the current row, endpoints included.

Parameters:

delta: Offset from current row, must be valid. If floating point must not be inf/nan.

Attributes

delta

class pylibcudf.rolling.BoundedOpen#

A bounded open window.

This window contains rows with delta of the current row, endpoints excluded.

Parameters:

delta: Offset from current row, must be valid. If floating point must not be inf/nan.

Attributes

delta

class pylibcudf.rolling.CurrentRow#

A current row rolling window.

This window contains all rows that are equal to the current row in the group.

class pylibcudf.rolling.RollingRequest(Column values, size_type min_periods, Aggregation aggregation)#

A request for a rolling aggregation.

Parameters:

values: The column of values to aggregate.
min_periods: The minimum number of observations required for a valid result in a given window.
aggregation: The aggregation to perform.

class pylibcudf.rolling.Unbounded#

An unbounded rolling window.

This window runs to the begin/end of the current row’s group.

pylibcudf.rolling.grouped_range_rolling_window(signatures, args, kwargs, defaults, _fused_sigindex_ref=[None])#

Perform grouping-aware range-based rolling window aggregations on some columns.

Parameters:

group_keys: Possibly empty table of sorted keys defining groups.
orderby: Column defining window ranges. Must be sorted, if group_keys is not empty, must be sorted groupwise.
order: Sort order of the orderby column.
null_order: Null sort order in the sorted orderby column
preceding: The type of the preceding window offset.
following: The type of the following window offset.
requests: List of RollingRequest objects.
streamStream | None: CUDA stream on which to perform the operation.
mrDeviceMemoryResource | None: Device memory resource used to allocate the returned table’s device memory.

Returns:

pylibcudf.rolling.rolling_window(signatures, args, kwargs, defaults, _fused_sigindex_ref=[None])#

Perform a rolling window operation on a column

For details, see cudf::rolling_window documentation.

Parameters:

sourceColumn: The column to perform the rolling window operation on.
preceding_windowUnion[Column, size_type]: The column containing the preceding window sizes or a scalar value indicating the sizes of all windows.
following_windowUnion[Column, size_type]: The column containing the following window sizes or a scalar value indicating the sizes of all windows.
min_periodsint: The minimum number of periods to include in the result.
aggAggregation: The aggregation to perform.
streamStream | None: CUDA stream on which to perform the operation.
mrDeviceMemoryResource | None: Device memory resource used to allocate the returned column’s device memory.

Returns: