From 1e55a6ba7f065efeae02b2c7f091cfd7a2a421e8 Mon Sep 17 00:00:00 2001 From: codycbakerphd Date: Sun, 3 Mar 2024 12:43:46 -0500 Subject: [PATCH] demo refactor --- .gitignore | 2 + demo/parallel/server.py | 81 ++++++++++++++++++++++++++--------------- 2 files changed, 54 insertions(+), 29 deletions(-) diff --git a/.gitignore b/.gitignore index 34fbcc3..6f8f449 100644 --- a/.gitignore +++ b/.gitignore @@ -5,3 +5,5 @@ dist .coverage .coverage.* codecov.xml + +.spyproject/ \ No newline at end of file diff --git a/demo/parallel/server.py b/demo/parallel/server.py index 7a13a68..248e40a 100644 --- a/demo/parallel/server.py +++ b/demo/parallel/server.py @@ -1,54 +1,73 @@ """Demo of parallel tqdm.""" -# HTTP server addition -import http.server import json -import socket -import socketserver import sys -import threading import time import uuid from concurrent.futures import ProcessPoolExecutor -from typing import List, Tuple +from typing import List import requests from tqdm_publisher import TQDMPublisher -def _run_sleep_in_subprocess(args: Tuple[int, int]): - """The operation to run on each subprocess.""" - repeat = args[0] - iteration_index = args[1] - id = args[2] - url = args[3] +def to_main_process(id: int, url: str, n: int, total: int, **kwargs): + """ + This is the parallel callback definition. - iterable = range(repeat) + Its parameters are attributes of a tqdm instance and their values are what a typical default tqdm printout + to console would contain (update step `n` out of `total` iterations). + """ + json_data = json.dumps(obj=dict(id=str(id), data=dict(n=n, total=total))) + + requests.post(url=url, data=json_data, headers={"Content-Type": "application/json"}) + + +def _run_sleep_tasks_in_subprocess(task_times: List[float], iteration_index: int, id: int, url: str): + """ + Run a 'task' that takes a certain amount of time to run on each worker. + + In this case that task is simply time.sleep. + + Parameters + ---------- + sleep_time : float + The amount of time this task emulates having taken to complete. + iteration_index : int + The index of this task in the list of all tasks from the buffer map. + Each index would map to a different tqdm position. + id : int + Identifier of ??. + url : str + The localhost URL to sent progress updates to. + """ sub_progress_bar = TQDMPublisher( - iterable=iterable, + iterable=task_times, position=iteration_index + 1, desc=f"Progress on iteration {iteration_index} ({id})", leave=False, ) if url: - - def to_main_process(n: int, total: int, **kwargs): - - json_data = json.dumps(dict(id=str(id), data=dict(n=n, total=total))) - - requests.post(url, data=json_data, headers={"Content-Type": "application/json"}) - sub_progress_bar.subscribe(to_main_process) - for _ in sub_progress_bar: - time.sleep(0.5) + for sleep_time in sub_progress_bar: + time.sleep(sleep_time) if __name__ == "__main__": number_of_jobs = 3 - repeats = [4, 6, 8, 10, 4, 8, 20, 10, 5, 12, 5, 4, 5, 5, 5] + + # Each outer entry is a list of 'tasks' to perform on a particular worker + # For demonstration purposes, each in the list of tasks is the length of time in seconds + # that each iteration of the task takes to run and update the progress bar (emulated by sleeping) + all_task_times: List[List[float]] = [ + [4.2, 6.7, 8.5, 10.3, 4.2, 8.1], + [20.5, 10.7, 5.3], + [12.4, 5.2, 4.9, 5.1], + [5.7, 5.8], + ] flags_list = sys.argv[1:] @@ -67,12 +86,16 @@ def to_main_process(n: int, total: int, **kwargs): URL = f"http://{HOST}:{PORT}" if port_flag else None - # Start the parallel jobs with ProcessPoolExecutor(max_workers=number_of_jobs) as executor: - + # Assign the parallel jobs job_map = executor.map( - _run_sleep_in_subprocess, - [(repeat, iteration_index, uuid.uuid4(), URL) for iteration_index, repeat in enumerate(repeats)], + _run_sleep_tasks_in_subprocess, + [ + (task_times, iteration_index, uuid.uuid4(), URL) + for iteration_index, task_times in enumerate(all_task_times) + ], ) - [_ for _ in job_map] # perform iteration to deploy jobs + # Perform iteration to deploy jobs + for _ in job_map: + pass