SE-for-Sci -- Homework 1

This problem has 3 parts.

Part I

In this repository, we have the following directories: In this part, we will set up the packaging for this repository.

• example - The source code for an example project. You will modify example/rk4.py in part II.
• tests - A directory containing tests, which may be helpful to you when modifying code. You do not need to modify this directory. Testing will be covered next week.

Task 1

Create a pyproject.toml file containing

• A build backend. hatchling is recommended, but any one should work.
• A project table with name, version, requires-python, and dependencies.
  • The name needs to be "example"; some backends (like hatchling) will look for a package with the same name.
  • There should only be one dependency: numpy.
• An dependency-groups table with a dev extra (dependencies pytest, and plotext).

This project now can be run with uv run if you have uv.

If you want to set up a .venv folder yourself, use uv sync (uv run does this for you). ? If you really want to do it manually without uv:

python -m venv .venv
source .venv/bin/activate
pip install -e. --group dev

(You might need to update pip to use --group, pip install -U pip)

The development group contains pytest, so you can do uv run pytest.

Task 2

Ruff's linter helps organize and debug code. We can write a ruleset into pyproject.toml so that every developer operates with the same set of rules.

Configure ruff by adding the following to your pyproject.toml.

[tool.ruff.lint]
extend-select = [
  "B",           # flake8-bugbear
  "I",           # isort
  "ARG",         # flake8-unused-arguments
  "C4",          # flake8-comprehensions
  "EM",          # flake8-errmsg
  "ICN",         # flake8-import-conventions
  "ISC",         # flake8-implicit-str-concat
  "G",           # flake8-logging-format
  "PGH",         # pygrep-hooks
  "PIE",         # flake8-pie
  "PL",          # pylint
  "PT",          # flake8-pytest-style
  "PTH",         # flake8-use-pathlib
  "RET",         # flake8-return
  "RUF",         # Ruff-specific
  "SIM",         # flake8-simplify
  "T20",         # flake8-print
  "UP",          # pyupgrade
  "YTT",         # flake8-2020
  "EXE",         # flake8-executable
  "NPY",         # NumPy specific rules
  "PD",          # pandas-vet
  "ERA",         # eradicate
]
ignore = [
  "PLR",    # Design related pylint codes
  "ISC001", # Conflicts with formatter
]
flake8-unused-arguments.ignore-variadic-names = true

[tool.ruff.lint.per-file-ignores]
"example/shm_error_analysis.py" = ["T201"]

We can run the linter with uvx ruff check (you can install ruff if you perfer, such as with uv tool install ruff). Some rules have automatic fixes (for example, isort will automatically sort imports) that can be employed with ruff check --fix. Additionally, some applications (like VS-Code) can be configured to highlight text where rules are broken. This is where those squiggly lines come from.

Task 3

Add a tool.pytest.ini_options table for pytest to pyproject.toml.

pytest tests should run without errors.

Instead of adding the arguments to the pytest command, you can specify them here. An example is given here, which is also listed below for convenience:

[tool.pytest.ini_options]
minversion = "7.0"
addopts = ["-ra", "--strict-markers", "--strict-config"]
xfail_strict = true
filterwarnings = ["error"]
log_cli_level = "INFO"
testpaths = ["tests"]

Part II

Fix the formatting of example/rk4.py. Before you submit, ruff check should not raise any issues. Additionally,

• Drop unneeded characters, like the . or .0 after values (unless you explicitly want an float, such as in a definition), or the 0, in a range(0, x), or parenthesis around return statements.
• Add docsstrings to functions.
• In solve(), _step() is called twice per loop. Reduce it to one by either using temporary variables or utilizing numpy vectorization syntax.

The code should still produce the same results. ruff format may also be of help. You may notice that ruff format behaves differently from ruff check --fix. They are different tools and not aliases of each other. If you are using VS-Code it is possible to configure ruff format as the formatter for python files, which can be run using a hotkey.

As you work, pytest may be helpful to verify that no code-breaking changes are made. example/shm_error_analysis.py is already formatted.

Hints

For the first parts, some good resources are:

• Week 5 of the course material
• Scientific Python Library Development Guide
• Packaging user guide
• INTERSECT Packaging workshop