pysatl_tsp.core.scrubber.abstract

Module Contents

Classes

Scrubber	Abstract base class for handlers that produce window views of time series data.
ScrubberWindow	A sliding window container for time series data processing.

Data

__all__

API

class pysatl_tsp.core.scrubber.abstract.Scrubber(source: pysatl_tsp.core.Handler[Any, pysatl_tsp.core.T] | None = None)

Bases: pysatl_tsp.core.Handler[pysatl_tsp.core.T, pysatl_tsp.core.scrubber.abstract.ScrubberWindow[pysatl_tsp.core.T]]

Abstract base class for handlers that produce window views of time series data.

Scrubbers consume individual data points and produce windows (sections) of the data stream. They are essential components for algorithms that need to analyze multiple data points together, such as moving averages, pattern detection, or feature extraction.

Concrete implementations of this class define specific windowing strategies such as fixed-size sliding windows, tumbling windows, or context-based windows.

Parameters:

source – The handler providing input data, defaults to None

Example:

# Example with a fixed-size sliding window scrubber (implementation not shown)

# Create a data source
data_source = SimpleDataProvider([10, 20, 30, 40, 50, 60, 70, 80])

# Create a sliding window scrubber with window size 3
window_scrubber = SlidingWindowScrubber(window_size=3, source=data_source)

# Process windows
for window in window_scrubber:
    # Each window is a ScrubberWindow instance
    print(f"Window values: {list(window.values)}")
    print(f"Window indices: {list(window.indices)}")

    # Calculate window statistics
    avg = sum(window.values) / len(window)
    print(f"Window average: {avg}")

# Output:
# Window values: [10, 20, 30]
# Window indices: [0, 1, 2]
# Window average: 20.0
#
# Window values: [20, 30, 40]
# Window indices: [1, 2, 3]
# Window average: 30.0
#
# ... and so on

Initialization

Initialize a scrubber.

Parameters:

source – The handler providing input data, defaults to None

abstractmethod __iter__() → collections.abc.Iterator[pysatl_tsp.core.scrubber.abstract.ScrubberWindow[pysatl_tsp.core.T]]

Create an iterator that yields window views of the input data.

Concrete implementations define specific windowing strategies.

Returns:

Iterator yielding ScrubberWindow instances

class pysatl_tsp.core.scrubber.abstract.ScrubberWindow(values: collections.deque[pysatl_tsp.core.T] | None = None, indices: collections.deque[int] | None = None)

Bases: typing.Generic[pysatl_tsp.core.T]

A sliding window container for time series data processing.

ScrubberWindow provides a specialized container for holding a window of time series data values along with their corresponding indices. It’s optimized for efficient append and remove operations at the ends of the window, making it suitable for sliding window algorithms in time series processing.

This class manages two parallel deques: one for the actual data values and another for their corresponding indices or positions in the original data stream.

Parameters:

• values – Deque containing the data values, defaults to None (empty deque)

• indices – Deque containing the indices corresponding to values, defaults to None (if not provided, sequential indices starting from 0 are used)

Raises:

ValueError – If the lengths of values and indices don’t match

Example:

# Create an empty window
window = ScrubberWindow()

# Add values with automatic indices
window.append(10.5)
window.append(11.2)
window.append(9.8)

# Add value with explicit index
window.append(12.1, index=100)

# Get value by position in window
first_value = window[0]  # 10.5

# Get a slice of the window
sub_window = window[1:3]  # Contains 11.2 and 9.8

# Iterate through values
for value in window:
    print(value)

# Get original position of a value
third_value_index = window.indices[2]  # 2
fourth_value_index = window.indices[3]  # 100

Initialization

Initialize a scrubber window.

Parameters:

• values – Deque containing the data values, defaults to None (empty deque)

• indices – Deque containing the indices corresponding to values, defaults to None (if not provided, sequential indices starting from 0 are used)

Raises:

ValueError – If the lengths of values and indices don’t match

__eq__(other: object) → bool

Check if this window equals another window.

Parameters:

other – Another object to compare with

Returns:

True if other is a ScrubberWindow with equal values and indices

__getitem__(key: int | slice) → pysatl_tsp.core.T | pysatl_tsp.core.scrubber.abstract.ScrubberWindow[pysatl_tsp.core.T]

Get a value or sub-window by index or slice.

Parameters:

key – Integer index or slice to retrieve

Returns:

Single value (if key is int) or sub-window (if key is slice)

Raises:

TypeError – If key is not an int or slice

__hash__() → int

Get window’s hash code.

Returns:

The hash value of the object

__iter__() → collections.abc.Iterator[pysatl_tsp.core.T]

Create an iterator over the values in the window.

Returns:

Iterator yielding window values

__len__() → int

Get the number of values in the window.

Returns:

The window size

__repr__() → str

Get a string representation of the window.

Returns:

String representation showing values and indices

append(value: pysatl_tsp.core.T, index: int | None = None) → None

Add a new value to the end of the window.

Parameters:

• value – The data value to append

• index – The index/position of the value in the original data stream, defaults to None (auto-assigned as len(self))

clear() → None

Remove all values from the window.

copy() → pysatl_tsp.core.scrubber.abstract.ScrubberWindow[pysatl_tsp.core.T]

Create a deep copy of the window.

Returns:

A new ScrubberWindow with copies of the values and indices

popleft() → None

Remove the oldest (leftmost) value from the window.

pysatl_tsp.core.scrubber.abstract.__all__ = ['Scrubber', 'ScrubberWindow', 'T']