jump-diffusion-estimation 0.2.0__py3-none-any.whl

jump_diffusion_estimation-0.2.0.dist-info/METADATA

@@ -0,0 +1,241 @@
+Metadata-Version: 2.4
+Name: jump-diffusion-estimation
+Version: 0.2.0
+Summary: Simulation and maximum-likelihood estimation of jump-diffusion processes with pluggable asymmetric, heavy-tailed jump distributions, plus likelihood-based inference and a test for the presence of jumps.
+Author-email: Juan David Ospina Arango <jdospina@gmail.com>
+License: MIT
+Project-URL: Homepage, https://github.com/jdospina/jump-diffusion-estimation
+Project-URL: Bug Tracker, https://github.com/jdospina/jump-diffusion-estimation/issues
+Project-URL: Documentation, https://jump-diffusion-estimation.readthedocs.io/
+Keywords: jump-diffusion,stochastic-processes,maximum-likelihood,differential-evolution,profile-likelihood,parametric-bootstrap,quantitative-finance,parameter-estimation
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Intended Audience :: Financial and Insurance Industry
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.8
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Topic :: Scientific/Engineering :: Mathematics
+Classifier: Topic :: Office/Business :: Financial
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy>=1.20.0
+Requires-Dist: scipy>=1.7.0
+Requires-Dist: matplotlib>=3.3.0
+Requires-Dist: seaborn>=0.11.0
+Requires-Dist: pandas>=1.3.0
+Requires-Dist: scikit-learn>=1.0.0
+Requires-Dist: tqdm>=4.60.0
+Provides-Extra: dev
+Requires-Dist: pytest>=6.0; extra == "dev"
+Requires-Dist: pytest-cov; extra == "dev"
+Requires-Dist: black; extra == "dev"
+Requires-Dist: flake8; extra == "dev"
+Requires-Dist: mypy; extra == "dev"
+Requires-Dist: jupyter; extra == "dev"
+Requires-Dist: sphinx; extra == "dev"
+Requires-Dist: sphinx-rtd-theme; extra == "dev"
+Requires-Dist: pandas-stubs; python_version >= "3.10" and extra == "dev"
+Requires-Dist: scipy-stubs; python_version >= "3.10" and extra == "dev"
+Provides-Extra: tutorials
+Requires-Dist: jupyter; extra == "tutorials"
+Requires-Dist: matplotlib; extra == "tutorials"
+Requires-Dist: seaborn; extra == "tutorials"
+Requires-Dist: plotly; extra == "tutorials"
+Provides-Extra: docs
+Requires-Dist: sphinx; extra == "docs"
+Requires-Dist: sphinx-rtd-theme; extra == "docs"
+Dynamic: license-file
+
+# Jump-Diffusion Parameter Estimation
+
+[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.21305522.svg)](https://doi.org/10.5281/zenodo.21305522)
+
+A comprehensive Python library for simulating and estimating parameters of jump-diffusion processes with asymmetric jump distributions.
+
+## 🚀 Features
+
+- **Flexible Simulation**: Generate jump-diffusion paths with customizable parameters
+- **Maximum Likelihood Estimation**: Robust parameter estimation using mixture distributions
+- **Pluggable Jump Distributions**: Skew-normal, Normal (Merton), and the Skewed Generalized Error Distribution (SGED) built in, with a simple interface (`jump_distribution=`) to add more
+- **Goodness-of-Fit Comparison**: Rank candidate jump distributions on the same data via AIC/BIC and a simulation-based Kolmogorov-Smirnov test
+- **Validation Tools**: Monte Carlo experiments for method validation
+- **Extensible Architecture**: Easy to add new models, jump distributions, and estimation methods
+- **Educational Focus**: Comprehensive documentation and tutorials
+
+## 📊 Model
+
+Our implementation focuses on jump-diffusion processes of the form:
+
+```
+dX_t = μ dt + σ dW_t + J_t dN_t
+```
+
+Where:
+- `μ`: drift parameter
+- `σ`: diffusion volatility
+- `W_t`: Brownian motion
+- `J_t`: jump sizes (asymmetrically distributed)
+- `N_t`: jump arrival times (Bernoulli approximation)
+
+## 🛠️ Installation
+
+```bash
+# Clone the repository
+git clone https://github.com/jdospina/jump-diffusion-estimation.git
+cd jump-diffusion-estimation
+
+# Install the package
+pip install -e .
+
+# Or install from PyPI (when available)
+pip install jump-diffusion-estimation
+```
+
+## 🎯 Quick Start
+
+```python
+import numpy as np
+from jump_diffusion import JumpDiffusionSimulator, JumpDiffusionEstimator
+
+# Create simulator
+simulator = JumpDiffusionSimulator(
+    mu=0.05,           # 5% annual drift
+    sigma=0.2,         # 20% annual volatility
+    jump_prob=0.1,     # 10% jump probability per period
+    jump_scale=0.15,   # jump magnitude scale
+    jump_skew=2.0      # positive skewness
+)
+
+# Simulate a path
+times, path, jumps = simulator.simulate_path(T=1.0, n_steps=252)
+
+# Estimate parameters
+increments = np.diff(path)
+dt = times[1] - times[0]
+estimator = JumpDiffusionEstimator(increments, dt)
+results = estimator.estimate()
+
+print(f"Estimated drift: {results['parameters']['mu']:.4f}")
+print(f"Estimated volatility: {results['parameters']['sigma']:.4f}")
+```
+
+### Using a different jump distribution
+
+Jumps follow a skew-normal distribution by default. Other distributions can be plugged in via `jump_distribution`, both when simulating and when estimating:
+
+```python
+from jump_diffusion.distributions import SGEDJump
+
+simulator = JumpDiffusionSimulator(
+    mu=0.05, sigma=0.2, jump_prob=0.1,
+    jump_distribution=SGEDJump(),
+    jump_loc=0.0, jump_scale=0.15, jump_nu=1.5, jump_xi=2.0,
+)
+times, path, jumps = simulator.simulate_path(T=1.0, n_steps=252)
+
+increments = np.diff(path)
+dt = times[1] - times[0]
+estimator = JumpDiffusionEstimator(increments, dt, jump_distribution=SGEDJump())
+results = estimator.estimate()
+```
+
+Distributions without a known closed-form likelihood (like SGED) fall back to a generic FFT-based convolution to approximate the mixture density, so adding a new distribution only requires implementing its `pdf`.
+
+### Robust estimation with differential evolution
+
+The default `L-BFGS-B` optimizer needs a reasonable initial guess and can stall on harder mixture likelihoods (SGED in particular). Differential evolution searches globally instead, needing no initial guess — the applied finding of the thesis this library is based on:
+
+```python
+results = estimator.estimate(method="differential_evolution", seed=42)
+```
+
+It costs thousands of likelihood evaluations (seconds instead of milliseconds), with defaults ported from the thesis (rand/1 strategy, `DEoptim`-style population sizing, early stopping on convergence).
+
+### Standard errors via Likelihood Profiling
+
+In complex jump-diffusion mixture models, the numerical Hessian is often unstable or ill-conditioned. Standard errors and 95% confidence intervals can be robustly calculated using Profile Likelihood. After estimating the parameters (preferably with global optimization), you can run:
+
+```python
+# Compute standard errors and confidence intervals using a Wilks' theorem threshold
+se_results = estimator.estimate_standard_errors(n_points=5, confidence_level=0.95)
+
+# The results table now includes standard errors and CI bounds
+estimator.diagnostics()
+
+# Visualize the profile log-likelihood curves
+estimator.plot_profiles()
+```
+
+### Comparing jump distributions
+
+`JumpDistributionComparison` fits several candidate jump distributions to the same data and ranks them by AIC/BIC plus a simulation-based Kolmogorov-Smirnov test:
+
+```python
+from jump_diffusion.distributions import NormalJump, SGEDJump, SkewNormalJump
+from jump_diffusion.validation import JumpDistributionComparison
+
+comparison = JumpDistributionComparison(increments, dt)
+comparison.fit("Normal", NormalJump())
+comparison.fit("SkewNormal", SkewNormalJump())
+comparison.fit("SGED", SGEDJump())
+
+print(comparison.compare())  # ranked by AIC, includes KS statistic/p-value
+comparison.plot_comparison()
+```
+
+## 📚 Examples
+
+Ready-to-run scripts are available in the `examples/` directory:
+
+- **[tutorial_completo.ipynb](notebooks/tutorial_completo.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/jdospina/jump-diffusion-estimation/blob/main/notebooks/tutorial_completo.ipynb) (español) · [tutorial_completo_en.ipynb](notebooks/tutorial_completo_en.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/jdospina/jump-diffusion-estimation/blob/main/notebooks/tutorial_completo_en.ipynb) (English) – the canonical, end-to-end tutorial: simulate, estimate (L-BFGS-B and Differential Evolution), quantify uncertainty via all three inference routes (profile / Wald / bootstrap), test for jumps, and compare jump distributions. Start here.**
+- [basic_usage.py](examples/basic_usage.py) – demonstrates basic library usage
+- [validation_experiment.py](examples/validation_experiment.py) – runs Monte Carlo validation experiments
+- [jump_diffusion_playground.ipynb](notebooks/jump_diffusion_playground.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/jdospina/jump-diffusion-estimation/blob/main/notebooks/jump_diffusion_playground.ipynb) – interactive playground: pick a jump distribution (Normal, Skew-Normal, SGED, Kou, Student-t), simulate, and try the "guess the parameters" game
+- [differential_evolution_showcase.ipynb](notebooks/differential_evolution_showcase.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/jdospina/jump-diffusion-estimation/blob/main/notebooks/differential_evolution_showcase.ipynb) – showcases the power of Differential Evolution (DE) compared to L-BFGS-B on the multimodal mixture likelihood of the SGED jump-diffusion model
+- [sp500_jump_diffusion_example.ipynb](notebooks/sp500_jump_diffusion_example.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/jdospina/jump-diffusion-estimation/blob/main/notebooks/sp500_jump_diffusion_example.ipynb) – applies the model to real S&P 500 data: parameter estimation, simulated-vs-real comparison, and ranking all five jump distributions by AIC/BIC/KS
+
+🌐 **Language / Idioma:** every notebook has an English counterpart with the `_en` suffix — [`jump_diffusion_playground_en.ipynb`](notebooks/jump_diffusion_playground_en.ipynb), [`differential_evolution_showcase_en.ipynb`](notebooks/differential_evolution_showcase_en.ipynb), [`sp500_jump_diffusion_example_en.ipynb`](notebooks/sp500_jump_diffusion_example_en.ipynb). Cada notebook tiene su versión en español (sin sufijo).
+
+### Notebook setup
+
+Install optional dependencies and launch Jupyter to explore the notebook:
+
+```bash
+pip install notebook ipywidgets matplotlib
+jupyter notebook
+```
+
+## 📖 Referencias Académicas
+
+Los métodos numéricos y modelos estadísticos implementados en esta librería están fundamentados en la siguiente literatura:
+
+1. **Calibración con Evolución Diferencial y SGED:**
+   - Ospina Arango, J. D. (2009). *Tesis de Maestría*. Universidad Nacional de Colombia. (Fundamentos de la aplicación de SGED y Evolución Diferencial a procesos de Salto-Difusión).
+   - Ardia, D., Ospina, J. D., & Giraldo, N. D. (2011). *Jump-diffusion calibration using differential evolution.* Wilmott, 2011(55), 76-79.
+   - Storn, R., & Price, K. (1997). *Differential evolution–a simple and efficient heuristic for global optimization over continuous spaces.* Journal of global optimization, 11(4), 341-359.
+
+2. **Modelos de Salto-Difusión y Distribuciones:**
+   - Merton, R. C. (1976). *Option pricing when underlying stock returns are discontinuous.* Journal of financial economics, 3(1-2), 125-144.
+   - Theodossiou, P. (2015). *Skewed Generalized Error Distribution of Financial Assets and Option Pricing.* Multinational Finance Journal, 19(4), 223-266.
+
+3. **Inferencia Estadística (Perfilado de Verosimilitud):**
+   - Wilks, S. S. (1938). *The large-sample distribution of the likelihood ratio for testing composite hypotheses.* The Annals of Mathematical Statistics, 9(1), 60-62.
+
+## 🤝 Contributing
+
+We welcome contributions! Please see our [Contributing Guidelines](CONTRIBUTING.md) for details.
+
+## 📄 License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
+
+## 🙏 Acknowledgments
+
+- Inspired by classical jump-diffusion literature
+- Built with love for the quantitative finance community
+- Special thanks to contributors and users

jump_diffusion_estimation-0.2.0.dist-info/RECORD

@@ -0,0 +1,29 @@
+jump_diffusion/__init__.py,sha256=E_wGyZb1jQBt8Bc7ykp3hQw-o3y7aAdxd8F_gBD7Fh0,1057
+jump_diffusion/distributions/__init__.py,sha256=RXXJsSuiFbVj8JCmLopv3uX7quv-uHK0ztc-Zx73q8I,501
+jump_diffusion/distributions/base.py,sha256=NtAY2nYAX_uNbku6sgkz0Rd9-yOSScLmPwf2a5zZheA,5750
+jump_diffusion/distributions/kou.py,sha256=a2OFcxBC0lm0zijc2iLY0ujatlO0VSAJQ3q9Q7fixCw,3896
+jump_diffusion/distributions/normal.py,sha256=AMBAFmdbKPUphQnJCQ0ir-xMDIzY8eRcYP_YvZAkj3g,2119
+jump_diffusion/distributions/sged.py,sha256=UZPYX4_u901QPMHAXv7f19UDcR_48KrTuqjK7gT9Mto,3949
+jump_diffusion/distributions/skew_normal.py,sha256=xAnQbPw2hfh2ELfX1e_ey_lm-b2tAJmjKEdBP_GluYI,3044
+jump_diffusion/distributions/student_t.py,sha256=tkca7XFrEsU_IEmy-QhMm9kihuyYlIuzJpSb-T1-HdE,2727
+jump_diffusion/estimation/__init__.py,sha256=IIKjIMnFFdTi4MnYTmltaMBN3tjtFGM-uxVyO5t19Fw,502
+jump_diffusion/estimation/base_estimator.py,sha256=HhcyzpKAuswXvsfMcRun5Ceg6hBXMEP2SEXw_u1tWPE,1412
+jump_diffusion/estimation/maximum_likelihood.py,sha256=e1E9cQz6eLZGVHTbTWcNo_gcFVjgaS3mdFhprQk-NFI,50048
+jump_diffusion/models/__init__.py,sha256=VKM-u4icKvUlewXswfMJMM87umWwPPnyrGRKpMxQySc,294
+jump_diffusion/models/base_model.py,sha256=6TSRcfUGIYLckaT-IR_gYSkt60RbTky_ZjI73TPOZSs,2253
+jump_diffusion/models/jump_diffusion.py,sha256=8FHJMEjGRfGS8BTukFalGrm_XwvbqAhzzolyoVWxvxY,5404
+jump_diffusion/scripts/__init__.py,sha256=Mp7GHPmlTv4ZlALlk54yusZFasvthZ9b6swupTlt9SI,90
+jump_diffusion/scripts/benchmark.py,sha256=exGXAWMv7o65Me8EVByj54e5StjTSQAb1Eldf9SiCdk,236
+jump_diffusion/scripts/validate.py,sha256=KOypZd2iQJqNCJwYJpGPC0h0SeLiFrlWoFtufp6G1m0,1992
+jump_diffusion/simulation/__init__.py,sha256=2OKJwsrg4vzSuigBRqyYMIBUiEExk4gvH9Ij5WvPuZw,358
+jump_diffusion/simulation/base_simulator.py,sha256=i9fnDveAi5gcJ5bUzbzOP4JITO72EX6RHQ9N-Nbqo38,1059
+jump_diffusion/simulation/jump_diffusion_simulator.py,sha256=BfPN6oDexyjBZ8Jewzqj2exNknj556BmAAUzrRtGhyQ,8872
+jump_diffusion/validation/__init__.py,sha256=WLC5GgcBDWxtMxf05oOJ6P7Vm8PPwhW9egvLKVnMzbU,341
+jump_diffusion/validation/distribution_comparison.py,sha256=U_4LWqRemAx8JuGJSEwYXH5kL34CQ7LbY89YidRf6zk,9894
+jump_diffusion/validation/monte_carlo.py,sha256=soROGlKdkJtMRFYzipCAoH1fkwPpCdQBWZe3nW3hzXs,11331
+jump_diffusion_estimation-0.2.0.dist-info/licenses/LICENSE,sha256=S7sgMJD5RbKFH4hZgD9HYYE-eEwH7qhOttKO50tuPxE,1224
+jump_diffusion_estimation-0.2.0.dist-info/METADATA,sha256=cdLu8ZpcOjIyyJhNveCKCPp5co8igu5B_H9XiMVCDVo,12472
+jump_diffusion_estimation-0.2.0.dist-info/WHEEL,sha256=K260EYznzXsJYBQGqmI8VTxEdiZYNvDZwW9cBh9-_MA,91
+jump_diffusion_estimation-0.2.0.dist-info/entry_points.txt,sha256=A2NPhQ5UMpwC5pYeJQ_AdFVSsHIVmqtB0WFHJiQRZfY,134
+jump_diffusion_estimation-0.2.0.dist-info/top_level.txt,sha256=JLdR-Z4TRQBJbIg1S8-5_buwdYPPNu687GG9iMKUauM,15
+jump_diffusion_estimation-0.2.0.dist-info/RECORD,,

jump_diffusion_estimation-0.2.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (83.0.0)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

jump_diffusion_estimation-0.2.0.dist-info/entry_points.txt

@@ -0,0 +1,3 @@
+[console_scripts]
+jumpdiff-benchmark = jump_diffusion.scripts.benchmark:main
+jumpdiff-validate = jump_diffusion.scripts.validate:main

jump_diffusion_estimation-0.2.0.dist-info/licenses/LICENSE

@@ -0,0 +1,25 @@
+# ==========================================
+# LICENSE
+# ==========================================
+
+MIT License
+
+Copyright (c) 2024 Juan David OSPINA ARANGO and Jump-Diffusion Estimation Contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

jump_diffusion_estimation-0.2.0.dist-info/top_level.txt

@@ -0,0 +1 @@
+jump_diffusion