pysrat 0.1.0

Python package for software reliability analysis and modeling

pysrat

Python implementation of NHPP (non-homogeneous Poisson process) software reliability models with a C/pybind11 core and a Pythonic estimator API.

Features

• Scikit-learn–style estimators: model.fit(data) and params_/llf_/aic_
• NHPP data helpers: NHPPData.from_intervals(...), from_counts(...), from_fault_times(...)
• Fast EM updates in C via pybind11
• CF1 (canonical phase-type) NHPP model and distribution helpers
• Simple plotting helpers (plot_mvf, plot_dmvf, plot_rate)

Installation

python -m pip install -U pip
pip install -e .

Notes:

• A C++20-capable compiler is required to build the CF1 extension module.

Quick start

import numpy as np
from pysrat import NHPPData, ExponentialNHPP, plot_mvf

data = NHPPData.from_intervals(time=[1, 1, 1, 1], fault=[0, 1, 0, 5])

model = ExponentialNHPP().fit(data)
print(model.params_)
print(model.aic_)

plot_mvf(data, model)

CF1 example:

import numpy as np
from pysrat import NHPPData
from pysrat.models.cf1 import CanonicalPhaseTypeNHPP

data = NHPPData.from_intervals(time=[1, 2, 1.5], fault=[1, 0, 2], type=[0, 1, 0])
model = CanonicalPhaseTypeNHPP(3).fit(data)
print(model.params_)

API overview

Data

• NHPPData.from_intervals(time=..., fault=..., type=..., te=...)
• NHPPData.from_counts(fault=...)
• NHPPData.from_fault_times(times=..., te=...)

Models

• ExponentialNHPP (exponential)
• TruncatedNormalNHPP, Pareto2NHPP, GammaNHPP, LogNormalNHPP
• TruncatedLogisticNHPP, LogLogisticNHPP
• TruncatedExtremeValueMaxNHPP, LogExtremeValueMaxNHPP
• TruncatedExtremeValueMinNHPP, LogExtremeValueMinNHPP
• CanonicalPhaseTypeNHPP (CF1)

Model comparison

from pysrat import compare, ExponentialNHPP, NHPPData

data = NHPPData.from_counts([0, 1, 0, 5])
fitted, best = compare([ExponentialNHPP()], data, criterion="AIC")

Distributions

R-like helpers live under pysrat.dists, e.g.:

from pysrat.dists import dcf1, pcf1, rcf1

alpha = [0.4, 0.3, 0.3]
rate = [0.5, 1.0, 1.5]
print(dcf1([0.1, 0.5], alpha=alpha, rate=rate))
print(pcf1([0.1, 0.5], alpha=alpha, rate=rate))
print(rcf1(5, alpha=alpha, rate=rate))

Plotting

• plot_mvf(data, model_or_results)
• plot_dmvf(data, model_or_results)
• plot_rate(data, model_or_results)

Note: Only the plot_* functions above are exported; legacy aliases like mvfplot, dmvfplot, and rateplot are not provided.

model_or_results can be an NHPPModel, a dict[str, NHPPModel], or a list of models.

Hyperparameters

Models expose sklearn-style hyperparameters via get_params()/set_params(). For example:

model = ExponentialNHPP(omega0=1.0, rate0=1.0)
model.set_params(rate0=0.5)

Development

python -m pip install -U pip
pip install -e .[dev]
pytest -q

Release

Tag and push to trigger the PyPI release workflow:

git tag v0.1.0
git push --tags

Set either:

• PYPI_API_TOKEN secret for token-based upload, or
• Trusted Publishing (OIDC) in PyPI and keep the workflow permissions as-is.