---

DerivKit Demos

Runnable demo collection for the DerivKit ecosystem.
This repository is non-installable by design — just clone it, create an environment, and run the demos in place.
It contains ready-to-run Python scripts and optional notebooks showcasing core DerivKit modules such as DerivativeKit, CalculusKit, and ForecastKit.

---

Repository structure

derivkit-demos/
├─ scripts/  # Runnable demo scripts
│  ├─ calculus-kit-advanced.py
│  ├─ calculus-kit-simple.py
│  ├─ derivative-kit-advanced.py
│  ├─ derivative-kit-simple.py
│  ├─ derivative-kit-tabulated-advanced.py
│  ├─ derivative-kit-tabulated-simple.py
│  ├─ forecast-kit-cosmic-shear-ccl.py
│  ├─ forecast-kit-cluster-counts-forecast.py
│  ├─ forecast-kit-cluster-counts-sweep.py
│  ├─ forecast-kit-dali-1d.py
│  ├─ forecast-kit-dali-2d.py
│  ├─ forecast-kit-fisher-bias.py
│  └─ forecast-kit-fisher.py
│
├─ notebooks/  # Optional notebooks
├─ utils/
│  └─ style.py  # Shared Matplotlib style and palette
│
├─ plots/  # Auto-generated figures (git-ignored)
├─ run_demo.py  # Launcher utility for demos
├─ requirements.txt  # Minimal dependencies
├─ environment.yaml  # Conda environment file
├─ pyproject.toml  # Metadata only (repo is not installable)
├─ LICENSE
└─ README.md

---

Demos included

DerivativeKit

• Derivative methods backends with analytic comparisons (finite, poly, adaptive).
• Derivatives of tabulated functions with interpolation schemes.

CalculusKit

• Gradients, Hessians, and Jacobians with analytic comparisons.

ForecastKit

• Fisher Information — analytic vs. numerical ellipses.
• Fisher Bias — systematic offset → parameter shift (Δθ).
• DALI — nonlinear likelihood expansions beyond Fisher.
• Cluster counts — forecasting and parameter sweeps for cluster-count examples
• Cosmic shear with pyCCL — forecast for cosmic shear with pyccl

Figures are automatically saved to plots/ as both .pdf and .png unless --plot is used to display them interactively.

---

Quick start

You do not need to install this repo — just clone and run inside a Python or Conda environment.

1 Create an environment

With pip:

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

With conda:

conda env create -f environment.yaml
conda activate derivkit-demos

---

2 Run demos

List all available demos:

python run_demo.py --list

Run a specific demo:

python run_demo.py forecast-kit-fisher
python run_demo.py forecast-kit-fisher-bias
python run_demo.py forecast-kit-dali-1d
python run_demo.py forecast-kit-dali-2d

Run by fuzzy name:

python run_demo.py fisher
python run_demo.py fisher-bias
python run_demo.py dali
python run_demo.py cluster

Pass extra arguments through to the demo with --:

python run_demo.py dali -- --plot
python run_demo.py forecast-kit-dali-2d -- --method adaptive
python run_demo.py forecast-kit-fisher-bias -- --plot

Run directly:

python -m scripts.forecast-kit-fisher.py --plot

The -- separator passes additional arguments directly to the demo.

---

Styling

All demos use a shared visual style from:

from utils.style import apply_plot_style, DEFAULT_COLORS
apply_plot_style()

Available colors:
DEFAULT_COLORS = {"blue": "#3b9ab2", "yellow": "#e1af00", "red": "#f21901"}

---

Tips

• ImportError: utils.style
  Always run scripts from the repo root or via run_demo.py.

• No install step
  This repo is meant to be run in place; avoid pip install ..

• Clean working tree
  Figures and notebook outputs are ignored via .gitignore.

---

Development tips

If you use ruff:

ruff check --fix .

Add or modify demos under demo-scripts/, following the numbered naming scheme (NN-topic-name.py) and including a top-level docstring explaining purpose, usage, and key results.

---

Citation

If you use derivkit in your research, please cite:

@software{sarcevic2025derivkit,
  author       = {Nikolina Šarčević and Matthijs van der Wild and Cynthia Trendafilova and Bastien Carreres},
  title        = {derivkit: A Python Toolkit for Numerical Derivatives},
  year         = {2025},
  publisher    = {GitHub},
  journal      = {GitHub Repository},
  howpublished = {\url{https://github.com/derivkit/derivkit}},
}

---

License

MIT — see LICENSE.

Acknowledgements

Developed as part of the DerivKit ecosystem (CalculusKit, ForecastKit, and friends).
Thanks to all contributors and collaborators supporting derivative-based cosmological inference tools.

Useful links

• DerivKit documentation
• DerivKit website
• DerivKit GitHub repository
• DerivKit organization