Packages

The main feature of Python is its extensive collection of packages that enhance its functionality. On this page, we will explore various aspects of working with some of these packages.

Note: Some packages have dedicated sections.

The table below lists the most useful python packages whose names are easy to forgot.

Package	Purpose / Goal	Install Command	Docs Link
pyperclip	Copy/paste clipboard cross-platform	pip install pyperclip	Docs
python-dotenv	Load .env file into environment variables	pip install python-dotenv	Docs
rich	Pretty terminal output (text, tables, markdown)	pip install rich	Docs
typer	Build beautiful CLI apps with minimal code	pip install typer	Docs
pydantic	Data validation and parsing using Python types	pip install pydantic	Docs
httpx	Modern async/sync HTTP client	pip install httpx	Docs
schedule	Lightweight job scheduler	pip install schedule	Docs
watchdog	Watch for file system changes	pip install watchdog	Docs
click	Create command-line interfaces	pip install click	Docs
loguru	Super simple logging	pip install loguru	Docs
tabulate	Format tables in plain text	pip install tabulate	Docs
tqdm	Add progress bars to loops	pip install tqdm	Docs
colorama	Cross-platform terminal colors	pip install colorama	Docs
validators	Validate emails, URLs, IPs, etc.	pip install validators	Docs
filetype	Guess file type by content, not extension	pip install filetype	Docs
humanize	Convert numbers/dates to human-readable formats	pip install humanize	Docs
yaspin	Pretty terminal spinners	pip install yaspin	Docs
icecream	Debug print with context (like ic())	pip install icecream	Docs
dateparser	Parse natural language dates	pip install dateparser	Docs
shortuuid	Short, URL-safe UUIDs	pip install shortuuid	Docs

tqdm

tqdm is a package that allows to create progress bars in the terminal using pseudographics. For more details check specific page.

---

The following cell demonstrates typical usage of tqdm. The tqdm.tqdm object wraps an iterable in a for loop, so each iteration of the loop increments the progress bar by 1.

import tqdm
for _ in tqdm.tqdm(range(10), ncols=100, desc="my bar"): pass

my bar: 100%|███████████████████████████████████████████████████| 10/10 [00:00<00:00, 120180.63it/s]

zipfile

zipfile is a package that allows you to operate on ZIP archives in Python.

---

As an example, we’ll create a ZIP file using a Linux command line, and then unpack that archive using Python.

The following cell is creating file that we’ll use as example.

%%bash
echo "zipped message" > zip_example_file
zip -r archive.zip zip_example_file
rm zip_example_file

  adding: zip_example_file (stored 0%)

Now, using zipfile.ZipFile, we’ll open the created archive and extract all its contents to the current directory.

from zipfile import ZipFile
with ZipFile("archive.zip") as f:
    f.extractall(".")

Now we can check the contents of the extracted file - it the same like during creation.

!cat zip_example_file

zipped message

After everything is done, don’t forget to clear the environment.

%%bash
rm -r zip_example_file archive.zip

Psutil

psutil is a package that allows you to get in python program information about the harware resources used by a Python program. Check details in the documentation of the tool.

---

The following cell shows code to get information about the RAM used by the program.

import psutil
process = psutil.Process()
(process.memory_info().rss / (1024**2))

62.97265625

To make sure that it really works, the following cell creates a large object and counts the RAM used by the process.

ans = [0 for _ in range(2**25)]
(process.memory_full_info().rss / (1024**2))

318.97265625

Now the number is much larger. Finally, delete the large object and measure the RAM again.

del ans
(process.memory_full_info().rss / (1024**2))

63.171875

APScheduler

APScheduler is a package that allows to start and controll processes that will do some jobs according to some triggers.

For more check:

• Official APScheduler documentation.

• Corresponding page on this website.

---

The following cell shows how to run the simplest BackgroundScheduler. This code will start a process that prints the current time to the console every 3 seconds.

from datetime import datetime
from apscheduler.schedulers.background import BackgroundScheduler
def periodic_task():
    print(datetime.now().time())

scheduler = BackgroundScheduler()
scheduler.add_job(periodic_task, "interval", seconds=3)
scheduler.start()

11:54:38.109044
11:54:41.108479
11:54:44.108156
11:54:47.108562
11:54:50.108621
11:54:53.108557

Stop the scheduler using the shutdown method.

scheduler.shutdown()

Drawsvg

drawsvg is a package that allows you to describe SVG graphics with python code. Check out the Drawsvg quick reference.

The following table represents the basic features of the drawsvg package:

Category	Function / Class	Description
Drawing	drawsvg.Drawing	Create a drawing container.
	Drawing.append	Append elements to the drawing.
	Drawing.to_svg	Export as an SVG string.
	Drawing.save_svg	Save SVG directly to file.
	Drawing.save_png	Export as PNG (requires cairosvg).
Composition	drawsvg.Group	Group multiple elements (<g>).
	drawsvg.Use	Reuse elements by reference (<use>).
Shapes	drawsvg.Rectangle	Draw rectangles.
	drawsvg.Circle	Draw circles.
	drawsvg.Ellipse	Draw ellipses.
	drawsvg.Line	Draw straight lines.
	drawsvg.Polygon	Draw polygons.
	drawsvg.Path	Freeform path drawing.
Text	drawsvg.Text	Add text elements.
Transformations	drawsvg.Translate	Apply translation transform.
	drawsvg.Scale	Apply scaling transform.
	drawsvg.Rotate	Apply rotation transform.
	element.set_transform	Set transform on an element.
Accessibility	element.append_title	Add title (tooltip).
	element.append_desc	Add description (for accessibility/metadata).

---

The following cell shows how to draw a cross on a red background using svg.

import drawsvg as dw
d = dw.Drawing(500, 500)
d.append(dw.Rectangle(10, 10, 490, 490, fill="red"))
d.append(dw.Line(50, 450, 450, 50, stroke='black', stroke_width=10))
d.append(dw.Line(50, 50, 450, 450, stroke='black', stroke_width=10))
d

As you can see, Jupyter knows how to represent the results of the drawsvg library.

For me, it’s typically important to be able to load the corresponding SVG code. Do that with as_svg method of the drawsvg canvas.

print(d.as_svg())

<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"
     width="500" height="500" viewBox="0 0 500 500">
<defs>
</defs>
<rect x="10" y="10" width="490" height="490" fill="red" />
<path d="M50,450 L450,50" stroke="black" stroke-width="10" />
<path d="M50,50 L450,450" stroke="black" stroke-width="10" />
</svg>

Viewbox

Applying a viewbox can be a bit confusing. This is because original SVG syntax allows for some flexibility. In drawsvg, you have to specify exactly viewBox name (camel naming) and enter the value as a sting with four numbers separated by the space.

---

The following cell illustrates the use of the viewBox. It sets the viewbox to display only a corner of the rectangle added later.

d = dw.Drawing(width=200, height=200, viewBox="-10 -10 100 100")
d.append(dw.Rectangle(
    0, 0, 200, 200,
    fill="none",
    stroke_width=20,
    stroke="black"
))
d

Gradio

Gradio is an instrument that allows you to build simple user interfaces for taking some input and immediate output. It is ideal for prototyping the machine leaning models, but can also be used in more generally.

---

The following cell shows the gradio hello world and an approach to visualizing the output in a jupyter notebook.

import gradio as gr

def greet(name: str) -> str:
    return f"Hello {name}!"

demo = gr.Interface(fn=greet, inputs="text", outputs="text")
demo.launch(inline=True)

* Running on local URL:  http://127.0.0.1:7860
* To create a public link, set `share=True` in `launch()`.

AWS SDK

The AWS SDK for Python is the boto3 package.

Obviously, you need a aws subscription to use the boto3. Fortunately, if you just familiarise yourself with the API you, can use the moto package, which mocks the behaviour of boto3.

The following table lists and describes the key components of the boto3 package.

Entity	Description	Abstraction Level	Typical Use	Notes
Client	Low-level interface that directly maps to AWS API operations.	Low	Perform AWS operations with full control over request parameters.	Follows AWS API 1:1; returns plain dictionaries (no objects).
Resource	Higher-level, object-oriented abstraction built on top of clients.	High	Work with AWS resources as Python objects instead of manual API calls.	Not all AWS services support resources; auto-manages pagination and references.
Collection	Iterable abstraction that groups related resources.	Medium	Iterate or filter groups of resources.	Returned by resource properties (e.g. buckets.all(), objects.filter()).
Resource Instance	Represents a single AWS resource (e.g., an S3 bucket or EC2 instance).	High	Access attributes or perform actions on that specific resource.	Exposes both attributes and actions of the AWS resource.
Resource Meta	Holds low-level metadata about a resource.	N/A	Access client and configuration of a resource.	Often used to access the underlying client for unsupported operations.
Waiter	Utility that polls AWS APIs until a resource reaches a desired state.	Medium	Wait for resources to be created, updated, or deleted.	Built into clients; waits for specific conditions.
Paginator	Handles paginated API responses automatically.	Medium	Iterate through results that span multiple pages.	Available via clients and some resources.

See the AWS SDK page for a more detailed description.

---

The following cell shows the region which is specified in my ~/.aws/config file.

import boto3
from moto import mock_aws

mock = mock_aws()
mock.start()

session = boto3.session.Session()
print(session.region_name)

mock.stop()

us-east-1