maxframe.liteframe.window.rolling.LiteFrameRolling#

class maxframe.liteframe.window.rolling.LiteFrameRolling(liteframe, window, min_periods=None, center=False, columns=None, groupby_params=None, order_by=None, order_by_ascending=True, window_shift=0)[source]#

Intermediate object returned by LiteFrame.rolling().

Provides rolling-window aggregations over a fixed-size integer window. Execution uses the sliced-data (concatenate-compute-slice) cross-chunk strategy: for multi-chunk inputs, up to window - 1 boundary rows are fetched from neighbor chunks, concatenated, the aggregation is computed over the concatenation, and the result is sliced back to the chunk’s range.

Parameters:

  • liteframe (LiteFrame) – The frame to compute over.

  • window (int) – Size of the moving window (number of rows). Must be a positive integer. Offset/time-based windows are not yet supported.

  • min_periods (int or None, default None) – Minimum number of observations in a window required to produce a value (otherwise null). Defaults to window (matching pandas).

  • center (bool, default False) – If True, the window is centered on the current row.

  • columns (str, list of str, or None, default None) – Column(s) to aggregate. If None, all (non-hidden, non-group-key) columns are aggregated. Non-selected columns are passed through.

  • groupby_params (dict or None, default None) – Set when created from a groupby.rolling() call so the tiler partitions data by group key before computing per-group windows.

__init__(liteframe, window, min_periods=None, center=False, columns=None, groupby_params=None, order_by=None, order_by_ascending=True, window_shift=0)[source]#

Methods

__init__(liteframe, window[, min_periods, ...])

agg(func)

Aggregate using one of the supported function names.

count()

Rolling count of non-null observations.

max()

Rolling maximum.

mean()

Rolling mean.

median()

Rolling median.

min()

Rolling minimum.

std()

Rolling standard deviation (ddof=1).

sum()

Rolling sum.

var()

Rolling variance (ddof=1).