Python教程-Python tqdm 模块 | 如何创建终端进度条

每当我们安装任何 Python 库、模块或安装软件时，都会在屏幕上看到进度条，表示小进度条，它估计了进程完成或渲染需要多长时间。它给我们一种活动的印象，可以使人放心。我们都熟悉各种不同的进度条。有些吸引人，有些则比较无聊。

在本文中，我们将创建进度条，而不涉及 Python 的核心日志库。

进度条根据完成任务所需进度的百分比填充。进度可以通过将 已处理的项目数 除以 总输入项目数 来计算。各种因素会影响进度条，例如网络速度、延迟，以及将持久数据写入本地存储以获得更准确的预计到达时间（ETA）。

我们可以使用名为 tqdm 的 Python 外部库创建简单而无麻烦的进度条。我们可以将其添加到代码中，使其看起来更漂亮。

tqdm 代表阿拉伯语中的 taqadum，意思是进度。Python tqdm 模块可以在各种平台上工作，如 Linux、Windows、Mac 等，它还与 IPython/Jupyter 笔记本兼容。

进度条的需求

如果我们处理较小的数据集，进度不会影响我们的工作流程。然而，进度条可以用于迭代数据集、训练模型或编码大量信息集。

• 进度条为我们提供了进程的估计，给出了它可能需要更多时间的时间近似值。
• 它为我们提供了进度仍在运行并且没有被粗鲁地终止的信息。

先决条件

系统必须安装 Python 3，并且我们还可以创建一个虚拟环境来安装 tqdm 库。

安装

打开命令行终端，输入以下内容。

pip install tqdm

或

pip3 install tqdm

以上命令将成功在系统中安装 tqdm。我们可以使用以下语句验证它。

import tqdm

如果没有错误，这意味着该库已成功安装。

为循环添加进度条

程序

import time  
import sys  
from tqdm import trange  
  
  
def do_something():  
    time.sleep(1)  
  
def do_another_something():  
    time.sleep(1)  
  
  
for i in trange(10, file=sys.stdout, desc='outer loop'):  
    do_something()  
  
    for j in trange(100,file=sys.stdout, leave=False, unit_scale=True, desc='inner loop'):  
        do_another_something()  

输出:

outer loop:   0%|          | 0/10 [00:00<?, ?it/s]
inner loop:   0%|          | 0.00/100 [00:00<?, ?it/s]
inner loop:   1%|          | 1.00/100 [00:01<01:40, 1.02s/it]
inner loop:   2%|▏         | 2.00/100 [00:02<01:39, 1.02s/it]
inner loop:   3%|▎         | 3.00/100 [00:03<01:38, 1.01s/it]
inner loop:   4%|▍         | 4.00/100 [00:04<01:37, 1.01s/it]
inner loop:   5%|▌         | 5.00/100 [00:05<01:35, 1.01s/it]
inner loop:   6%|▌         | 6.00/100 [00:06<01:34, 1.01s/it]
inner loop:   7%|▋         | 7.00/100 [00:07<01:34, 1.01s/it]
inner loop:   8%|▊         | 8.00/100 [00:08<01:33, 1.02s/it]
inner loop:   9%|▉         | 9.00/100 [00:09<01:32, 1.02s/it]
inner loop:  10%|█         | 10.0/100 [00:10<01:31, 1.02s/it]
inner loop:  11%|█         | 11.0/100 [00:11<01:30, 1.01s/it]
inner loop:  12%|█▏        | 12.0/100 [00:12<01:29, 1.01s/it]
inner loop:  13%|█▎        | 13.0/100 [00:13<01:28, 1.01s/it]
inner loop:  14%|█▍        | 14.0/100 [00:14<01:29, 1.04s/it]
inner loop:  15%|█▌        | 15.0/100 [00:15<01:27, 1.03s/it]
inner loop:  16%|█▌        | 16.0/100 [00:16<01:26, 1.02s/it]
inner loop:  17%|█▋        | 17.0/100 [00:17<01:24, 1.02s/it]
inner loop:  18%|█▊        | 18.0/100 [00:18<01:23, 1.02s/it]
inner loop:  19%|█▉        | 19.0/100 [00:19<01:22, 1.02s/it]
…………………………………………………………………………………………….
……………………………………………………………………………………………..
It will go till it reached at 100%.

我们再了解一个例子。

示例-1

# importing modules  
from tqdm import trange  
from time import sleep  
  
# creating loop  
for i in trange(10, desc="loop "):  
    # slowing the for loop  
    sleep(0.1)  

输出:

loop : 100%|██████████| 10/10 [00:01<00:00, 9.08it/s]

示例 -2

# importing modules  
from tqdm import tnrange  
from time import sleep  
  
# creating loop  
for i in tnrange(2, dec="loop 1"):  
  
    # creating nested loop  
    for j in tnrange(5, dec="loop 2"):  
        # slowing the for loop  
        sleep(0.3)  

输出:

outer loop:   0%|          | 0/10 [00:00<?, ?it/s]
inner loop:   0%|          | 0.00/100 [00:00<?, ?it/s]
inner loop:   1%|          | 1.00/100 [00:01<01:39, 1.00s/it]
inner loop:   2%|▏         | 2.00/100 [00:02<01:38, 1.00s/it]
inner loop:   3%|▎         | 3.00/100 [00:03<01:37, 1.00s/it]
inner loop:   4%|▍         | 4.00/100 [00:04<01:36, 1.00s/it]
inner loop:   5%|▌         | 5.00/100 [00:05<01:35, 1.00s/it]
inner loop:   6%|▌         | 6.00/100 [00:06<01:34, 1.01s/it]
inner loop:   7%|▋         | 7.00/100 [00:07<01:33, 1.01s/it]
inner loop:   8%|▊         | 8.00/100 [00:08<01:32, 1.01s/it]
inner loop:   9%|▉         | 9.00/100 [00:09<01:31, 1.00s/it]
inner loop:  10%|█         | 10.0/100 [00:10<01:30, 1.01s/it]

示例 -3

# importing modules  
import time  
import sys  
from tqdm import trange  
  
  
# creating random function  
def random_function():  
    time.sleep(0.5)  
  
  
# another random function  
def another_random_function():  
    time.sleep(0.2)  
  
  
# defining outer loop  
for i in trange(3, file=sys.stdout, desc='Outer loop'):  
    random_function()  
  
    # inner loop  
    for j in trange(5, file=sys.stdout, desc='Inner loop'):  
        another_random_function()  

输出:

Outer loop:   0%|          | 0/3 [00:00<?, ?it/s]
Inner loop:   0%|          | 0/5 [00:00<?, ?it/s]
Inner loop:  20%|██        | 1/5 [00:00<00:00,  4.62it/s]
Inner loop:  40%|████      | 2/5 [00:00<00:00,  4.64it/s]
Inner loop:  60%|██████    | 3/5 [00:00<00:00,  4.63it/s]
Inner loop:  80%|████████  | 4/5 [00:00<00:00,  4.65it/s]
Outer loop:  33%|███▎      | 1/3 [00:01<00:03,  1.58s/it]
Inner loop:   0%|          | 0/5 [00:00<?, ?it/s]
Inner loop:  20%|██        | 1/5 [00:00<00:00,  4.91it/s]
Inner loop:  40%|████      | 2/5 [00:00<00:00,  4.85it/s]
Inner loop:  60%|██████    | 3/5 [00:00<00:00,  4.79it/s]
Inner loop:  80%|████████  | 4/5 [00:00<00:00,  4.77it/s]
Outer loop:  67%|██████▋   | 2/3 [00:03<00:01,  1.58s/it]
Inner loop:   0%|          | 0/5 [00:00<?, ?it/s]
Inner loop:  20%|██        | 1/5 [00:00<00:00,  4.43it/s]
Inner loop:  40%|████      | 2/5 [00:00<00:00,  4.56it/s]
Inner loop:  60%|██████    | 3/5 [00:00<00:00,  4.65it/s]
Inner loop:  80%|████████  | 4/5 [00:00<00:00,  4.72it/s]
Outer loop: 100%|██████████| 3/3 [00:04<00:00,  1.57s/it]

预测性手动更新进度条

tqdm 模块提供了手动在特定间隔更新进度条的功能。当我们以多部分方式下载文件或处理流数据时，可以手动更新进度条函数。让我们了解以下示例。

示例 -

# importing modules  
import time  
import sys  
from tqdm import tqdm  
  
  
def task():  
    time.sleep(1)  
  
  
with tqdm(total=100, file=sys.stdout) as pbar:  
    for i in range(10):  
        task()  
        # Here we are updating progress bar manually, useful for streams such as reading files.  
        pbar.update(10)  
        # Updates in increments of 10 stops at 100  

输出:

100%|██████████| 100/100 [00:10<00:00, 9.93it/s]

解释：

在上面的代码中，我们设置了总数属性为100。调用的函数在每次迭代中递增十次，直到达到100%。我们可以将任何值传递给 update() 方法。

多线程进度条

我们还可以将 Python tqdm 包嵌套到 Python 线程中。多进程是使用所有核心的最佳方式。tqdm 的位置参数允许我们指定要打印此进度条的行偏移。在未编写脚本的情况下，默认设置为自动。让我们了解以下示例。必须指定该参数以管理多个进度条。如果忽略此参数，我们的进度条将被不同线程覆盖。

示例 -

import time  
  
from random import randrange  
from multiprocessing.pool import ThreadPool  
  
from tqdm import tqdm  
  
  
def func_call(position, total):  
    text = 'progressbar #{position}'.format(position=position)  
    with  tqdm(total=total, position=position, desc=text) as progress:  
        for _ in range(0, total, 5):  
            progress.update(5)  
            time.sleep(randrange(3))  
  
  
pool = ThreadPool(10)  
tasks = range(5)  
for i, url in enumerate(tasks, 1):  
    pool.apply_async(func_call, args=(i, 100))  
pool.close()  
pool.join()  

输出:

progressbar #5:   0%|          | 0/100 [00:00<?, ?it/s]
progressbar #2:   0%|          | 0/100 [00:00<?, ?it/s]
progressbar #1:   0%|          | 0/100 [00:00<?, ?it/s]
progressbar #4:   0%|          | 0/100 [00:00<?, ?it/s]
progressbar #3:   0%|          | 0/100 [00:00<?, ?it/s]
progressbar #3:  10%|█         | 10/100 [00:01<00:09,  9.91it/s]
progressbar #5:  10%|█         | 10/100 [00:01<00:09,  9.91it/s]
progressbar #1:  10%|█         | 10/100 [00:02<00:18,  4.97it/s]
progressbar #2:  10%|█         | 10/100 [00:02<00:18,  4.96it/s]
progressbar #4:  10%|█         | 10/100 [00:02<00:18,  4.96it/s]
progressbar #5:  15%|█▌        | 15/100 [00:02<00:11,  7.64it/s]
progressbar #3:  15%|█▌        | 15/100 [00:02<00:11,  7.64it/s]
progressbar #4:  15%|█▌        | 15/100 [00:03<00:17,  4.96it/s]
progressbar #2:  15%|█▌        | 15/100 [00:03<00:17,  4.95it/s]
progressbar #3:  20%|██        | 20/100 [00:04<00:16,  4.71it/s]
progressbar #5:  25%|██▌       | 25/100 [00:04<00:11,  6.58it/s]
progressbar #1:  20%|██        | 20/100 [00:04<00:16,  4.97it/s]
progressbar #2:  20%|██        | 20/100 [00:04<00:16,  4.96it/s]

在 Tqdm 进度条中添加颜色

颜色可以使进度条变得非常吸引人。但是，它不会为进度条的工作方式添加任何新功能。tqdm 可以与 Colorama 一起使用，Colorama 是 Python 中的简单跨平台颜色终端文本。让我们了解以下示例。

示例 -

from tqdm import trange  
from colorama import Fore  
  
# Cross-platform colored terminal text.  
color_bars = [Fore.BLACK,  
    Fore.RED,  
    Fore.GREEN,  
    Fore.YELLOW,  
    Fore.BLUE,  
    Fore.MAGENTA,  
    Fore.CYAN,  
    Fore.WHITE]  
  
for color in color_bars:  
    for i in trange(int(7e7),  
                    bar_format="{l_bar}%s{bar}%s{r_bar}" % (color, Fore.RESET)):  
        pass  

输出:

结论

我们已经讨论了与进度条相关的所有基本概念。Python 自带了 tqdm 模块，可以帮助我们手动设计进度条。我们为重要操作定义了合适的示例，可以在 tqdm 模块中执行这些操作。tqdm 模块可以与子进程和线程协作，我们可以同时运行多个进度条。