Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Feedback fixes #25

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Feedback fixes #25

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
9f8538e
shorten epsiode and update layout
christopher-wild Aug 14, 2024
048e823
accessing packages episode grammar and suggestions
christopher-wild Aug 14, 2024
e91e733
creating-packages grammar and suggestions
christopher-wild Aug 14, 2024
074927a
Update from PR feedback comments
christopher-wild Sep 30, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
39 changes: 28 additions & 11 deletions episodes/accessing-packages.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@ exercises: 2

:::::::::::::::::::::::::::::::::::::: questions

- What are the different ways of downloading python packages?
- What are the different ways of downloading Python packages?
- What are package managers?
- How can I access my own package?

Expand Down Expand Up @@ -39,7 +39,7 @@ Some of the most popular packages you may have heard of are:

:::

To use a package that is installed you use the key word `import` in python.
To use a package that is installed you use the key word `import` in Python.

```python
# This imports the pandas package and gives it a new name 'pd'.
Expand All @@ -51,7 +51,7 @@ pd.read_csv("/my_data.csv")

## Python Package Index (PyPI)

The Python Package Index or PyPI is an online repository of Python packages hosting over 500,000 packages! While there are alternatives such as [conda-forge](https://conda-forge.org), PyPI is by far the most commonly used and likely to have all you need.
The Python Package Index or PyPI is an online repository of Python packages hosting over 500,000 packages! While there are alternatives such as [conda-forge](https://conda-forge.org), PyPI is by far the most commonly used and likely to have all you need.

::: challenge

Expand All @@ -72,7 +72,7 @@ You can think of PyPI being the supermarket full of packages and pip being the d

### Using pip

pip itself is a python package that can be found on [PyPI](https://pypi.org/project/pip/). It however comes preinstalled with most python installations, for example [python.org](https://python.org) and inside virtual environments.
pip itself is a Python package that can be found on [PyPI](https://pypi.org/project/pip/). It however comes preinstalled with most Python installations, for example [python.org](https://python.org) and inside virtual environments.

The most common way to use pip is from the command line. At the top of a package page on PyPI will be the example line you need to install the package

Expand All @@ -82,6 +82,10 @@ py -m pip install numpy

The above will install [numpy](https://pypi.org/project/numpy/) from PyPI, a popular scientific computing package enabling a wide range of mathematical and scientific functions.

::: callout
You may notice a wheel file download during the pip install, for example `Downloading numpy-2.0.0-cp312-cp312-win_amd64.whl`. A wheel in Python is a prebuilt package format that allows for quicker and more efficient installation, so when it is downloaded your local computer doesn't need to do any building. The alternative is source files which often take the form `.zip` or `.tar.gz`, which when downloaded will then need to be built then installed, which is often far slower.

:::

::: challenge
### Exercise 2: Create venv and install Numpy
Copy link
Collaborator

@f-allian f-allian Sep 17, 2024 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The Windows/Mac tabs look absolutely massive here and in another one below! Might be because of the number of #?

image

Copy link
Owner Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

They do look commically big on yours! The amount of #'s is the same as the workbench recommends and looks okay on mine
image.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@christopher-wild that's strange! Never mind then!

Expand All @@ -103,6 +107,8 @@ Step 1: Create a venv in the .venv directory using the command `py -m venv .venv

:::

When activated you should see the name of your environment in brackets at the start of your terminal line

Step 2: Install Numpy into your new environment

Step 3: Check your results with `py -m pip list`
Expand All @@ -122,13 +128,24 @@ Check out [this documentation](https://docs.python.org/3/l[PyPI](https://pypi.or

pip can also be used to install packages from source. This means that the package file structure (source) is on your local computer and pip installs it using the instructions from the `setup.py` or `pyproject.toml` file. This is especially handy for packages either not on PyPI, like ones downloaded from github, or for your own packages you're developing.

```
py -m pip install .
```
::: tab

### Windows

`py -m pip install .`



### macOS / Linux

`python3 -m pip install .`


:::

::: instructor

The above command should be universal on both windows and mac/unix setups. It may be worth checking with the class at this point that they are all familiar with the -m notation, and what the above command does exactly
It may be worth checking with the class at this point that they are all familiar with the -m notation, and what the above command does exactly

:::

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Again I think the key points below can be made more informative

Expand All @@ -137,9 +154,9 @@ Here the `.` means to install your current directory as a Python package. For th


::: keypoints
- pip can be used to download and install Python packages
- PyPI is an online package repository which pip downloads from
- pip can also install local packages like your own
- pip is the most common tool used to download and access python packages from PyPI.
- PyPI is an online package repository which users can choose to upload their packages to for others to use.
- pip can also be used to install packages on your local system (installing from source)
:::


Expand Down
26 changes: 13 additions & 13 deletions episodes/creating-packages.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@

## Introduction

This episode will see us creating our own Python project from scratch and installing it into a python virtual environment ready for use. Feel free if you're feeling adventurous to create your own package content or follow along with this example of a Fibonacci counter.
This episode will see us creating our own Python project from scratch and installing it into a Python virtual environment ready for use. Feel free if you're feeling adventurous to create your own package content or follow along with this example of a Fibonacci counter.


## Fibonacci Counter
Expand All @@ -43,7 +43,7 @@
Think back to the earlier episodes and try to recall all the things that can go into a package.

::: solution
1. Python Module - This is the directory with the python code that does the work.
1. Python Module - This is the directory with the Python code that does the work.
2. Configuration File - e.g. your pyproject.toml file
3. Other metadata files - e.g. LICENCE, README.md, citation.cff
4. Python Tests - A directory full of unit-tests and other tests
Expand All @@ -61,12 +61,12 @@
└── 📄 test_fibonacci.py
```

The first thing we will do in this project is create the python module (the actual code!).
The first thing we will do in this project is create the Python module (the actual code!).

::: challenge
### Creating Python module

1. Create a python file called `fibonacci.py` as shown in the structure above.
1. Create a Python file called `fibonacci.py` as shown in the structure above.
2. Add the following code which contains a function that returns the Fibonacci sequence
```python
def fibonacci(n_terms):
Expand All @@ -86,11 +86,11 @@
::: challenge
### Using your Python module

Create a script in your project directory that imports and uses your fibonacci script. This will serve as a good quick test that it works.
Create a script in your project directory that imports and uses your Fibonacci script. This will serve as a good quick test that it works.

::: solution
1. Create the file in `/my_project`, for example `fibonacci_test.py`.
2. Import and run the fibonacci function:
2. Import and run the Fibonacci function:
```python
from my_package.fibonacci import fibonacci

Expand All @@ -110,9 +110,9 @@


[project]
name = "my_cool_package"
version = "0.0.0"
description = "A package to do awesome things"
name = "fibonacci"
version = "0.1.0"
description = "A package to calculate the fibonacci sequence"
dependencies = ["pandas", "numpy"]
```

Expand All @@ -125,7 +125,7 @@
::: callout
### pyproject.toml documentation

The full list of accepted keys can be found [here](https://packaging.python.org/en/latest/specifications/pyproject-toml/) in the documentation

Check warning on line 128 in episodes/creating-packages.md

View workflow job for this annotation

GitHub Actions / Build markdown source files if valid 

 [uninformative link text]: [here](https://packaging.python.org/en/latest/specifications/pyproject-toml/)
:::

::: challenge
Expand All @@ -147,7 +147,7 @@

[project]
name = "fibonacci"
version = "0.0.0"
version = "0.1.0"
description = "A package which can produce the Fibonacci sequence"
authors = [{name = "your_name", email="[email protected]"}]
keywords = ["fibonacci", "maths"]
Expand All @@ -162,16 +162,16 @@
When installing your own package locally, there is an option called editable or `-e` for short.
`py -m pip install -e .`

With a default installation (without -e), any changes to your source package will only appear in your python environment when your package is rebuilt and reinstalled. The editable option allows for quick development of a package by removing that need to be reinstalled, for this reason it is sometimes called development mode!
With a default installation (without -e), any changes to your source package will only appear in your Python environment when your package is rebuilt and reinstalled. The editable option allows for quick development of a package by removing that need to be reinstalled, for this reason it is sometimes called development mode!



:::
::::::::::::::::::::::::::::::::::::: keypoints

- A package can be built with as little as 2 files, a python script and a configuration file
- A package can be built with as little as 2 files, a Python script and a configuration file
- pyproject.toml files have 2 key tables, [build-system] and [project]
- Editable instals allow for quick and easy package development
- Editable installs allow for quick and easy package development

::::::::::::::::::::::::::::::::::::::::::::::::

75 changes: 19 additions & 56 deletions episodes/package-file-history.Rmd
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
title: 'Package File History'
teaching: 10
exercises: 2
exercises: 1
editor_options:
markdown:
wrap: 100
Expand Down Expand Up @@ -31,8 +31,7 @@ This is to both build an understanding of why the current standard is what it is

1. requirements.txt
2. setup.py
3. setup.cfg
4. pyproject.toml
3. pyproject.toml

:::

Expand All @@ -58,11 +57,15 @@ Before the introduction of `pyproject.toml` the main tool supported for installi

Q: Discuss with each other what problems if any you think there may be with using a python file to create python packages

`Hint: Think about the differences between a code file and a text file`
::: hint
### Hint

Think about the differences between a code file and a text file
:::

::: solution

Some potential issues are:

1. As `setup.py` is a code file, there is a potential for malicious code to be hidden in them if the file comes from an untrusted source
2. There is quite a bit of 'boilerplate' in each file
3. The syntax has to be precise and may be confusing to those not familiar with Python
Expand All @@ -74,66 +77,20 @@ Some potential issues are:
from setuptools import setup

setup(
name='my_cool_package',
name='fibonacci',
version='1.0.0',
description='A package to do awesome things',
description='A package to calculate the fibonacci sequence',
long_description=open('README.md').read(),
author='John Doe',
author_email='[email protected]',
license='MIT',
)
```

## Setup.cfg

To tackle some of the problems with `setup.py`, another standard file was introduced called `setup.cfg` (cfg stands for config).

The task of a `setup.cfg` file is to declare metadata and settings required for a package in a simple manner. Unlike a setup.py which requires code imports and functions, the `setup.cfg` only has headers and key value pairs.

::: callout
### Key Value Pair

A key value pair is a fundamental way of storing data which is used across many languages and formats. Here's how it works:

- Key: Is the unique identifier, like the label on a file in a filing cabinet
- Value: Is the actual data that needs storing. It can be a number, text or many other things.

An example would be name = Jane

:::

```
[metadata]

name = my_cool_package
description = A package to do awesome things
long_description = file: README.md
author = John Doe
author_email = [email protected]
keywords = data, analysis, python
license = MIT

[options]
# Specify libraries your project needs (dependencies)
install_requires = pandas numpy

# Python version compatibility (optional)
python_requires = >=3.7
```

When using a `setup.cfg` however, a dummy `setup.py` was still required to build the package. This looked like:

```python
from setuptools import setup

if __name__ == "__main__":
setup()
```

## Pyproject.toml

Introduced in (PEP517)[https://peps.python.org/pep-0517/], the latest file for packaging a python project is the `pyproject.toml` file.
Like a `.cfg` file, a `toml` file is designed to be easy to read and declarative.
Introduced in [PEP517](https://peps.python.org/pep-0517/), the latest file for packaging a python project is the `pyproject.toml` file.
Like a `.cfg` file, a `toml` file is designed to be easy to read and declarative. It is the current recommended way to package your Python

::: callout
TOML stands for Tom's Obvious Minimal Language!
Expand All @@ -150,7 +107,7 @@ requires = ["setuptools", "wheel"]

Project metadata however was still being specified either in a `setup.py` or a `setup.cfg`, the latter being preferred.

With the introduction of (PEP621)[https://peps.python.org/pep-0621/] in 2020, project metadata could also be stored in the `pyproject.toml` files, meaning you only now need the single file to specify all the build requirements and metadata required for your package! This is still the preferred way in the community.
With the introduction of [PEP621](https://peps.python.org/pep-0621/) in 2020, project metadata could also be stored in the `pyproject.toml` files, meaning you only now need the single file to specify all the build requirements and metadata required for your package! This is still the preferred way in the community.

We will be going into how to make a `pyproject.toml` file in more detail in one of the next episodes.

Expand All @@ -170,3 +127,9 @@ dependencies = ["pandas", "numpy"]
[tool.black]
line-length = 98
```

::: keypoints
- Python packages make code easier to install, reuse and maintain.
- A single pyproject.toml file is all that is required to package your Python project.
- There are multiple standards out there for Python packaging, but pyproject.toml is the current recommended way.
:::
Loading