Monitor

This module provides a monitoring tool for tracking the performance of threaded and process-based generators.

It visualizes queue sizes, production rates, and consumption rates in real-time, either in the terminal or using matplotlib.

Examples

>>> from threaded_generator import ThreadedGenerator
>>> import time
>>>
>>> # Create a monitor
>>> monitor = Monitor()
>>>
>>> # Pass the monitor to a generator
>>> gen = ThreadedGenerator(range(100), monitor=monitor, name="MyGen")
>>>
>>> # Start the monitor (usually in a context manager or separate thread)
>>> with monitor:
...     for item in gen:
...         time.sleep(0.1)  # Simulate work
...
# (Terminal output will show a progress bar and stats for "MyGen")

class Monitor(name='Monitor')

Bases: Thread

A monitoring thread that aggregates and displays statistics for parallel generators.

It runs in the background, consuming statistics pushed by generators via a queue, and updates a live display in the terminal.

Parameters:

name (str) – The name of the monitor thread. Defaults to “Monitor”.

drain()

Process all currently available items in the statistics queue.

Calculates current throughputs and average queue sizes since the last drain.

Returns:

A tuple of (throughput_map, average_queue_sizes).

Return type:

tuple[dict[tuple[str, str], float], dict[str, int]]

plot()

Display live statistics using matplotlib on the main thread.

This method blocks until the plot window is closed. It requires matplotlib to be installed.

Return type:

None

put(x)

Push a statistic record to the monitor.

This is typically called internally by the ParallelGenerator classes.

Parameters:

x (MonitorFormat) – A tuple containing (name, type, duration, thread_id, qsize, maxsize).

Return type:

None

run()

Execute the main monitoring loop.

Repeatedly drains the queue, calculates stats, and updates the terminal display until the stop event is set.

Return type:

None