burning-cost 0.1.0__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (4) hide show
  1. burning_cost-0.1.0/PKG-INFO +120 -0
  2. burning_cost-0.1.0/README.md +88 -0
  3. burning_cost-0.1.0/pyproject.toml +52 -0
  4. burning_cost-0.1.0/src/burning_cost/__init__.py +74 -0
burning_cost-0.1.0/PKG-INFO ADDED
@@ -0,0 +1,120 @@
1
+ Metadata-Version: 2.4
2
+ Name: burning-cost
3
+ Version: 0.1.0
4
+ Summary: Meta-package installing the full Burning Cost insurance pricing toolkit
5
+ Project-URL: Homepage, https://burning-cost.github.io
6
+ Project-URL: Repository, https://github.com/burning-cost/burning-cost-meta
7
+ Project-URL: Documentation, https://burning-cost.github.io
8
+ Author-email: Burning Cost <pricing.frontier@gmail.com>
9
+ License: MIT
10
+ Keywords: actuarial,causal inference,insurance,machine learning,pricing
11
+ Classifier: Development Status :: 4 - Beta
12
+ Classifier: Intended Audience :: Financial and Insurance Industry
13
+ Classifier: License :: OSI Approved :: MIT License
14
+ Classifier: Programming Language :: Python :: 3
15
+ Classifier: Programming Language :: Python :: 3.10
16
+ Classifier: Programming Language :: Python :: 3.11
17
+ Classifier: Programming Language :: Python :: 3.12
18
+ Classifier: Topic :: Office/Business :: Financial
19
+ Classifier: Topic :: Scientific/Engineering
20
+ Requires-Python: >=3.10
21
+ Requires-Dist: insurance-causal
22
+ Requires-Dist: insurance-conformal
23
+ Requires-Dist: insurance-credibility
24
+ Requires-Dist: insurance-fairness
25
+ Requires-Dist: insurance-frequency-severity
26
+ Requires-Dist: insurance-gam
27
+ Requires-Dist: insurance-governance
28
+ Requires-Dist: insurance-monitoring
29
+ Requires-Dist: insurance-telematics
30
+ Requires-Dist: insurance-whittaker
31
+ Description-Content-Type: text/markdown
32
+
33
+ # burning-cost
34
+
35
+ [![PyPI](https://img.shields.io/pypi/v/burning-cost)](https://pypi.org/project/burning-cost/)
36
+ [![Python](https://img.shields.io/pypi/pyversions/burning-cost)](https://pypi.org/project/burning-cost/)
37
+ [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
38
+
39
+ **One install. The full Burning Cost toolkit.**
40
+
41
+ ```bash
42
+ pip install burning-cost
43
+ ```
44
+
45
+ That single command installs all 10 flagship libraries for insurance pricing with modern ML.
46
+
47
+ ---
48
+
49
+ ## The problem this solves
50
+
51
+ The Burning Cost ecosystem has 10 libraries. They are designed to work together — causal inference to estimate treatment effects, conformal to put intervals on predictions, monitoring to track whether those predictions are still valid in production. But you should not have to remember which libraries to include in a requirements file when setting up a new pricing project or Colab notebook.
52
+
53
+ `burning-cost` is a meta-package: it has no code of its own. It is a single dependency that pulls in everything.
54
+
55
+ ---
56
+
57
+ ## What you get
58
+
59
+ | Library | What it does |
60
+ |---|---|
61
+ | `insurance-causal` | Double Machine Learning for causal treatment effects on observational data |
62
+ | `insurance-fairness` | Fairness auditing under UK Equality Act 2010 |
63
+ | `insurance-monitoring` | PSI/CSI drift, A/E monitoring, Gini tests, anytime-valid A/B testing |
64
+ | `insurance-conformal` | Distribution-free prediction intervals with coverage guarantees |
65
+ | `insurance-whittaker` | Whittaker-Henderson graduation for triangles and rating factors |
66
+ | `insurance-telematics` | Trip-level feature engineering and behavioural risk scoring |
67
+ | `insurance-credibility` | Buhlmann-Straub and hierarchical Bayesian credibility models |
68
+ | `insurance-frequency-severity` | Two-part freq/sev modelling with proper exposure handling |
69
+ | `insurance-gam` | Generalised Additive Models for interpretable rating factor curves |
70
+ | `insurance-governance` | FCA/PRA-aligned model risk management documentation |
71
+
72
+ ---
73
+
74
+ ## Usage
75
+
76
+ You do not import from `burning_cost` directly. Each library has its own namespace:
77
+
78
+ ```python
79
+ from insurance_causal import CausalPricingModel
80
+ from insurance_conformal import InsuranceConformalPredictor
81
+ from insurance_monitoring import MonitoringReport
82
+ from insurance_fairness import FairnessAudit
83
+ ```
84
+
85
+ The meta-package just ensures they are all installed.
86
+
87
+ ---
88
+
89
+ ## For new pricing projects
90
+
91
+ Add to your `pyproject.toml`:
92
+
93
+ ```toml
94
+ [project]
95
+ dependencies = [
96
+ "burning-cost",
97
+ ]
98
+ ```
99
+
100
+ Or for a requirements file:
101
+
102
+ ```
103
+ burning-cost
104
+ catboost
105
+ polars
106
+ ```
107
+
108
+ ---
109
+
110
+ ## Resources
111
+
112
+ - Website: [burning-cost.github.io](https://burning-cost.github.io)
113
+ - Examples: [github.com/burning-cost/burning-cost-examples](https://github.com/burning-cost/burning-cost-examples)
114
+ - 30-minute demo: [Colab notebook](https://colab.research.google.com/github/burning-cost/burning-cost-examples/blob/main/notebooks/burning_cost_30_minutes.ipynb)
115
+
116
+ ---
117
+
118
+ ## License
119
+
120
+ MIT
burning_cost-0.1.0/README.md ADDED
@@ -0,0 +1,88 @@
1
+ # burning-cost
2
+
3
+ [![PyPI](https://img.shields.io/pypi/v/burning-cost)](https://pypi.org/project/burning-cost/)
4
+ [![Python](https://img.shields.io/pypi/pyversions/burning-cost)](https://pypi.org/project/burning-cost/)
5
+ [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
6
+
7
+ **One install. The full Burning Cost toolkit.**
8
+
9
+ ```bash
10
+ pip install burning-cost
11
+ ```
12
+
13
+ That single command installs all 10 flagship libraries for insurance pricing with modern ML.
14
+
15
+ ---
16
+
17
+ ## The problem this solves
18
+
19
+ The Burning Cost ecosystem has 10 libraries. They are designed to work together — causal inference to estimate treatment effects, conformal to put intervals on predictions, monitoring to track whether those predictions are still valid in production. But you should not have to remember which libraries to include in a requirements file when setting up a new pricing project or Colab notebook.
20
+
21
+ `burning-cost` is a meta-package: it has no code of its own. It is a single dependency that pulls in everything.
22
+
23
+ ---
24
+
25
+ ## What you get
26
+
27
+ | Library | What it does |
28
+ |---|---|
29
+ | `insurance-causal` | Double Machine Learning for causal treatment effects on observational data |
30
+ | `insurance-fairness` | Fairness auditing under UK Equality Act 2010 |
31
+ | `insurance-monitoring` | PSI/CSI drift, A/E monitoring, Gini tests, anytime-valid A/B testing |
32
+ | `insurance-conformal` | Distribution-free prediction intervals with coverage guarantees |
33
+ | `insurance-whittaker` | Whittaker-Henderson graduation for triangles and rating factors |
34
+ | `insurance-telematics` | Trip-level feature engineering and behavioural risk scoring |
35
+ | `insurance-credibility` | Buhlmann-Straub and hierarchical Bayesian credibility models |
36
+ | `insurance-frequency-severity` | Two-part freq/sev modelling with proper exposure handling |
37
+ | `insurance-gam` | Generalised Additive Models for interpretable rating factor curves |
38
+ | `insurance-governance` | FCA/PRA-aligned model risk management documentation |
39
+
40
+ ---
41
+
42
+ ## Usage
43
+
44
+ You do not import from `burning_cost` directly. Each library has its own namespace:
45
+
46
+ ```python
47
+ from insurance_causal import CausalPricingModel
48
+ from insurance_conformal import InsuranceConformalPredictor
49
+ from insurance_monitoring import MonitoringReport
50
+ from insurance_fairness import FairnessAudit
51
+ ```
52
+
53
+ The meta-package just ensures they are all installed.
54
+
55
+ ---
56
+
57
+ ## For new pricing projects
58
+
59
+ Add to your `pyproject.toml`:
60
+
61
+ ```toml
62
+ [project]
63
+ dependencies = [
64
+ "burning-cost",
65
+ ]
66
+ ```
67
+
68
+ Or for a requirements file:
69
+
70
+ ```
71
+ burning-cost
72
+ catboost
73
+ polars
74
+ ```
75
+
76
+ ---
77
+
78
+ ## Resources
79
+
80
+ - Website: [burning-cost.github.io](https://burning-cost.github.io)
81
+ - Examples: [github.com/burning-cost/burning-cost-examples](https://github.com/burning-cost/burning-cost-examples)
82
+ - 30-minute demo: [Colab notebook](https://colab.research.google.com/github/burning-cost/burning-cost-examples/blob/main/notebooks/burning_cost_30_minutes.ipynb)
83
+
84
+ ---
85
+
86
+ ## License
87
+
88
+ MIT
burning_cost-0.1.0/pyproject.toml ADDED
@@ -0,0 +1,52 @@
1
+ [build-system]
2
+ requires = ["hatchling"]
3
+ build-backend = "hatchling.build"
4
+
5
+ [project]
6
+ name = "burning-cost"
7
+ version = "0.1.0"
8
+ description = "Meta-package installing the full Burning Cost insurance pricing toolkit"
9
+ readme = "README.md"
10
+ license = { text = "MIT" }
11
+ requires-python = ">=3.10"
12
+ authors = [
13
+ { name = "Burning Cost", email = "pricing.frontier@gmail.com" },
14
+ ]
15
+ keywords = [
16
+ "insurance",
17
+ "actuarial",
18
+ "pricing",
19
+ "machine learning",
20
+ "causal inference",
21
+ ]
22
+ classifiers = [
23
+ "Development Status :: 4 - Beta",
24
+ "Intended Audience :: Financial and Insurance Industry",
25
+ "License :: OSI Approved :: MIT License",
26
+ "Programming Language :: Python :: 3",
27
+ "Programming Language :: Python :: 3.10",
28
+ "Programming Language :: Python :: 3.11",
29
+ "Programming Language :: Python :: 3.12",
30
+ "Topic :: Scientific/Engineering",
31
+ "Topic :: Office/Business :: Financial",
32
+ ]
33
+ dependencies = [
34
+ "insurance-causal",
35
+ "insurance-fairness",
36
+ "insurance-monitoring",
37
+ "insurance-conformal",
38
+ "insurance-whittaker",
39
+ "insurance-telematics",
40
+ "insurance-credibility",
41
+ "insurance-frequency-severity",
42
+ "insurance-gam",
43
+ "insurance-governance",
44
+ ]
45
+
46
+ [project.urls]
47
+ Homepage = "https://burning-cost.github.io"
48
+ Repository = "https://github.com/burning-cost/burning-cost-meta"
49
+ Documentation = "https://burning-cost.github.io"
50
+
51
+ [tool.hatch.build.targets.wheel]
52
+ packages = ["src/burning_cost"]
burning_cost-0.1.0/src/burning_cost/__init__.py ADDED
@@ -0,0 +1,74 @@
1
+ """
2
+ burning-cost: the full Burning Cost insurance pricing toolkit.
3
+
4
+ This is a meta-package. Installing it gives you all 10 flagship libraries
5
+ from the Burning Cost ecosystem in one go:
6
+
7
+ pip install burning-cost
8
+
9
+ Libraries included
10
+ ------------------
11
+ insurance-causal
12
+ Double Machine Learning for causal effect estimation. Isolates the causal
13
+ effect of price changes, telematics, and other treatments from selection bias
14
+ in observational insurance data.
15
+
16
+ insurance-fairness
17
+ Fairness auditing for insurance pricing models. Tests for indirect
18
+ discrimination on protected characteristics under UK Equality Act 2010.
19
+
20
+ insurance-monitoring
21
+ Model drift detection and monitoring. PSI/CSI feature drift, A/E ratio
22
+ monitoring, Gini discrimination tests, Murphy decomposition, anytime-valid
23
+ A/B testing, and PIT-based sequential calibration monitoring.
24
+
25
+ insurance-conformal
26
+ Distribution-free prediction intervals with coverage guarantees. Pearson-
27
+ weighted nonconformity scores for Poisson/Tweedie count data, locally-weighted
28
+ conformal, conformal risk control for premium sufficiency, and Solvency II
29
+ SCR bounds.
30
+
31
+ insurance-whittaker
32
+ Whittaker-Henderson graduation for claim development triangles and rating
33
+ factor smoothing. The actuarial workhorse, properly implemented.
34
+
35
+ insurance-telematics
36
+ Telematics trip-level feature engineering and behavioural risk scoring.
37
+ Speed profiles, harsh event detection, and mileage-adjusted frequency models.
38
+
39
+ insurance-credibility
40
+ Credibility weighting and hierarchical Bayesian models for thin-data
41
+ rating factors. Buhlmann-Straub and hierarchical GLM approaches.
42
+
43
+ insurance-frequency-severity
44
+ Two-part frequency/severity modelling. Poisson/Tweedie GLMs with proper
45
+ exposure handling, offset terms, and model diagnostics.
46
+
47
+ insurance-gam
48
+ Generalised Additive Models for insurance pricing. Interpretable smooth
49
+ terms for continuous rating factors, with visualisation utilities.
50
+
51
+ insurance-governance
52
+ Model governance documentation and audit trail generation. FCA/PRA-aligned
53
+ model risk management outputs from fitted pricing models.
54
+
55
+ Usage
56
+ -----
57
+ You do not import from ``burning_cost`` directly — import from the individual
58
+ libraries as needed::
59
+
60
+ from insurance_causal import CausalPricingModel
61
+ from insurance_conformal import InsuranceConformalPredictor
62
+ from insurance_monitoring import MonitoringReport
63
+
64
+ See https://burning-cost.github.io for full documentation and examples.
65
+ """
66
+
67
+ from importlib.metadata import version, PackageNotFoundError
68
+
69
+ try:
70
+ __version__ = version("burning-cost")
71
+ except PackageNotFoundError:
72
+ __version__ = "0.0.0"
73
+
74
+ __all__ = ["__version__"]