asianscore-quant 1.0.0__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.
- asianscore_quant-1.0.0/LICENSE +13 -0
- asianscore_quant-1.0.0/PKG-INFO +78 -0
- asianscore_quant-1.0.0/README.md +50 -0
- asianscore_quant-1.0.0/pyproject.toml +49 -0
- asianscore_quant-1.0.0/python/asianscore_quant/__init__.py +18 -0
- asianscore_quant-1.0.0/python/asianscore_quant/math_models.py +112 -0
- asianscore_quant-1.0.0/python/asianscore_quant.egg-info/PKG-INFO +78 -0
- asianscore_quant-1.0.0/python/asianscore_quant.egg-info/SOURCES.txt +10 -0
- asianscore_quant-1.0.0/python/asianscore_quant.egg-info/dependency_links.txt +1 -0
- asianscore_quant-1.0.0/python/asianscore_quant.egg-info/top_level.txt +1 -0
- asianscore_quant-1.0.0/setup.cfg +4 -0
- asianscore_quant-1.0.0/tests/test_quant.py +38 -0
|
@@ -0,0 +1,13 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2026 AsianScore Lab (https://asianscore.com)
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
|
13
|
+
copies or substantial portions of the Software.
|
|
@@ -0,0 +1,78 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: asianscore-quant
|
|
3
|
+
Version: 1.0.0
|
|
4
|
+
Summary: Zero-dependency sports quantitative analytics SDK: Power/Multiplicative de-vigging, Dixon-Coles xG Poisson grid, Kelly criterion, Brier score auditing, and live AsianScore feed client.
|
|
5
|
+
Author-email: AsianScore Quantitative Research Lab <pypi@aimail.men>
|
|
6
|
+
License: MIT
|
|
7
|
+
Project-URL: Homepage, https://asianscore.com/
|
|
8
|
+
Project-URL: Documentation, https://asianscore.com/tools.html
|
|
9
|
+
Project-URL: SHA-256 Audit Ledger, https://asianscore.com/models.html#ledger
|
|
10
|
+
Project-URL: Zenodo Dataset (DOI), https://doi.org/10.5281/zenodo.24441036
|
|
11
|
+
Project-URL: Repository, https://github.com/AIYIXIU/asianscore-quant-sdk
|
|
12
|
+
Project-URL: Go Package, https://pkg.go.dev/github.com/AIYIXIU/asianscore-go
|
|
13
|
+
Keywords: sports-analytics,expected-goals,xg,dixon-coles,poisson,devig,kelly-criterion,brier-score,football,asianscore
|
|
14
|
+
Classifier: Development Status :: 5 - Production/Stable
|
|
15
|
+
Classifier: Intended Audience :: Developers
|
|
16
|
+
Classifier: Intended Audience :: Science/Research
|
|
17
|
+
Classifier: License :: OSI Approved :: MIT License
|
|
18
|
+
Classifier: Programming Language :: Python :: 3
|
|
19
|
+
Classifier: Programming Language :: Python :: 3.9
|
|
20
|
+
Classifier: Programming Language :: Python :: 3.10
|
|
21
|
+
Classifier: Programming Language :: Python :: 3.11
|
|
22
|
+
Classifier: Programming Language :: Python :: 3.12
|
|
23
|
+
Classifier: Topic :: Scientific/Engineering :: Mathematics
|
|
24
|
+
Requires-Python: >=3.9
|
|
25
|
+
Description-Content-Type: text/markdown
|
|
26
|
+
License-File: LICENSE
|
|
27
|
+
Dynamic: license-file
|
|
28
|
+
|
|
29
|
+
# AsianScore Quant SDK (`asianscore-quant`)
|
|
30
|
+
|
|
31
|
+
[](https://pypi.org/project/asianscore-quant/)
|
|
32
|
+
[](https://doi.org/10.5281/zenodo.24441036)
|
|
33
|
+
[](https://github.com/AIYIXIU/asianscore-quant-sdk/blob/main/LICENSE)
|
|
34
|
+
|
|
35
|
+
Zero-dependency Python sports quantitative toolkit implementing the exact mathematical primitives powering [AsianScore](https://asianscore.com/):
|
|
36
|
+
|
|
37
|
+
- **Odds De-Vigging:** Multiplicative & Power-method fair probability solvers (`devig_multiplicative`, `devig_power`)
|
|
38
|
+
- **Bankroll Management:** Fractional Kelly Criterion (`kelly_fraction`)
|
|
39
|
+
- **Football xG Modeling:** Dixon-Coles low-score correlation bivariate Poisson grid (`dixon_coles_probs`)
|
|
40
|
+
- **Calibration Auditing:** Multi-class Brier score (`brier_score`)
|
|
41
|
+
- **Live Open Feed Client:** `fetch_live_summary()` querying `https://asianscore.com/api/widget/summary` (CORS-enabled, no API key required)
|
|
42
|
+
|
|
43
|
+
## Installation
|
|
44
|
+
|
|
45
|
+
```bash
|
|
46
|
+
pip install asianscore-quant
|
|
47
|
+
```
|
|
48
|
+
|
|
49
|
+
## Quick Start
|
|
50
|
+
|
|
51
|
+
```python
|
|
52
|
+
from asianscore_quant import devig_power, dixon_coles_probs, kelly_fraction, brier_score
|
|
53
|
+
|
|
54
|
+
fair = devig_power([2.10, 3.40, 3.60])
|
|
55
|
+
grid = dixon_coles_probs(1.62, 1.18, rho=-0.08)
|
|
56
|
+
stake = kelly_fraction(prob=0.52, decimal_odds=2.10, fraction=0.25)
|
|
57
|
+
score = brier_score([0.55, 0.25, 0.20], actual_index=0)
|
|
58
|
+
```
|
|
59
|
+
|
|
60
|
+
## Open Platform & Ecosystem
|
|
61
|
+
|
|
62
|
+
- **Official Platform:** https://asianscore.com/
|
|
63
|
+
- **Interactive Web Calculators:** https://asianscore.com/tools.html
|
|
64
|
+
- **SHA-256 Verifiable AI Prediction Ledger:** https://asianscore.com/models.html#ledger
|
|
65
|
+
- **Zenodo Academic Dataset (CERN DOI):** https://doi.org/10.5281/zenodo.24441036
|
|
66
|
+
- **Docs (GitHub Pages):** https://aiyixiu.github.io/asianscore-quant-sdk/
|
|
67
|
+
- **Official Go Package (`pkg.go.dev`):** https://pkg.go.dev/github.com/AIYIXIU/asianscore-go
|
|
68
|
+
- **Embeddable Live Widget:** https://aiyixiu.github.io/asianscore-live-widget/
|
|
69
|
+
|
|
70
|
+
## Verification
|
|
71
|
+
|
|
72
|
+
```bash
|
|
73
|
+
python3 -m unittest tests/test_quant.py
|
|
74
|
+
```
|
|
75
|
+
|
|
76
|
+
## License
|
|
77
|
+
|
|
78
|
+
MIT
|
|
@@ -0,0 +1,50 @@
|
|
|
1
|
+
# AsianScore Quant SDK (`asianscore-quant`)
|
|
2
|
+
|
|
3
|
+
[](https://pypi.org/project/asianscore-quant/)
|
|
4
|
+
[](https://doi.org/10.5281/zenodo.24441036)
|
|
5
|
+
[](https://github.com/AIYIXIU/asianscore-quant-sdk/blob/main/LICENSE)
|
|
6
|
+
|
|
7
|
+
Zero-dependency Python sports quantitative toolkit implementing the exact mathematical primitives powering [AsianScore](https://asianscore.com/):
|
|
8
|
+
|
|
9
|
+
- **Odds De-Vigging:** Multiplicative & Power-method fair probability solvers (`devig_multiplicative`, `devig_power`)
|
|
10
|
+
- **Bankroll Management:** Fractional Kelly Criterion (`kelly_fraction`)
|
|
11
|
+
- **Football xG Modeling:** Dixon-Coles low-score correlation bivariate Poisson grid (`dixon_coles_probs`)
|
|
12
|
+
- **Calibration Auditing:** Multi-class Brier score (`brier_score`)
|
|
13
|
+
- **Live Open Feed Client:** `fetch_live_summary()` querying `https://asianscore.com/api/widget/summary` (CORS-enabled, no API key required)
|
|
14
|
+
|
|
15
|
+
## Installation
|
|
16
|
+
|
|
17
|
+
```bash
|
|
18
|
+
pip install asianscore-quant
|
|
19
|
+
```
|
|
20
|
+
|
|
21
|
+
## Quick Start
|
|
22
|
+
|
|
23
|
+
```python
|
|
24
|
+
from asianscore_quant import devig_power, dixon_coles_probs, kelly_fraction, brier_score
|
|
25
|
+
|
|
26
|
+
fair = devig_power([2.10, 3.40, 3.60])
|
|
27
|
+
grid = dixon_coles_probs(1.62, 1.18, rho=-0.08)
|
|
28
|
+
stake = kelly_fraction(prob=0.52, decimal_odds=2.10, fraction=0.25)
|
|
29
|
+
score = brier_score([0.55, 0.25, 0.20], actual_index=0)
|
|
30
|
+
```
|
|
31
|
+
|
|
32
|
+
## Open Platform & Ecosystem
|
|
33
|
+
|
|
34
|
+
- **Official Platform:** https://asianscore.com/
|
|
35
|
+
- **Interactive Web Calculators:** https://asianscore.com/tools.html
|
|
36
|
+
- **SHA-256 Verifiable AI Prediction Ledger:** https://asianscore.com/models.html#ledger
|
|
37
|
+
- **Zenodo Academic Dataset (CERN DOI):** https://doi.org/10.5281/zenodo.24441036
|
|
38
|
+
- **Docs (GitHub Pages):** https://aiyixiu.github.io/asianscore-quant-sdk/
|
|
39
|
+
- **Official Go Package (`pkg.go.dev`):** https://pkg.go.dev/github.com/AIYIXIU/asianscore-go
|
|
40
|
+
- **Embeddable Live Widget:** https://aiyixiu.github.io/asianscore-live-widget/
|
|
41
|
+
|
|
42
|
+
## Verification
|
|
43
|
+
|
|
44
|
+
```bash
|
|
45
|
+
python3 -m unittest tests/test_quant.py
|
|
46
|
+
```
|
|
47
|
+
|
|
48
|
+
## License
|
|
49
|
+
|
|
50
|
+
MIT
|
|
@@ -0,0 +1,49 @@
|
|
|
1
|
+
[build-system]
|
|
2
|
+
requires = ["setuptools>=68"]
|
|
3
|
+
build-backend = "setuptools.build_meta"
|
|
4
|
+
|
|
5
|
+
[project]
|
|
6
|
+
name = "asianscore-quant"
|
|
7
|
+
version = "1.0.0"
|
|
8
|
+
description = "Zero-dependency sports quantitative analytics SDK: Power/Multiplicative de-vigging, Dixon-Coles xG Poisson grid, Kelly criterion, Brier score auditing, and live AsianScore feed client."
|
|
9
|
+
readme = "README.md"
|
|
10
|
+
license = { text = "MIT" }
|
|
11
|
+
requires-python = ">=3.9"
|
|
12
|
+
authors = [
|
|
13
|
+
{ name = "AsianScore Quantitative Research Lab", email = "pypi@aimail.men" }
|
|
14
|
+
]
|
|
15
|
+
keywords = [
|
|
16
|
+
"sports-analytics",
|
|
17
|
+
"expected-goals",
|
|
18
|
+
"xg",
|
|
19
|
+
"dixon-coles",
|
|
20
|
+
"poisson",
|
|
21
|
+
"devig",
|
|
22
|
+
"kelly-criterion",
|
|
23
|
+
"brier-score",
|
|
24
|
+
"football",
|
|
25
|
+
"asianscore",
|
|
26
|
+
]
|
|
27
|
+
classifiers = [
|
|
28
|
+
"Development Status :: 5 - Production/Stable",
|
|
29
|
+
"Intended Audience :: Developers",
|
|
30
|
+
"Intended Audience :: Science/Research",
|
|
31
|
+
"License :: OSI Approved :: MIT License",
|
|
32
|
+
"Programming Language :: Python :: 3",
|
|
33
|
+
"Programming Language :: Python :: 3.9",
|
|
34
|
+
"Programming Language :: Python :: 3.10",
|
|
35
|
+
"Programming Language :: Python :: 3.11",
|
|
36
|
+
"Programming Language :: Python :: 3.12",
|
|
37
|
+
"Topic :: Scientific/Engineering :: Mathematics",
|
|
38
|
+
]
|
|
39
|
+
|
|
40
|
+
[project.urls]
|
|
41
|
+
Homepage = "https://asianscore.com/"
|
|
42
|
+
Documentation = "https://asianscore.com/tools.html"
|
|
43
|
+
"SHA-256 Audit Ledger" = "https://asianscore.com/models.html#ledger"
|
|
44
|
+
"Zenodo Dataset (DOI)" = "https://doi.org/10.5281/zenodo.24441036"
|
|
45
|
+
Repository = "https://github.com/AIYIXIU/asianscore-quant-sdk"
|
|
46
|
+
"Go Package" = "https://pkg.go.dev/github.com/AIYIXIU/asianscore-go"
|
|
47
|
+
|
|
48
|
+
[tool.setuptools.packages.find]
|
|
49
|
+
where = ["python"]
|
|
@@ -0,0 +1,18 @@
|
|
|
1
|
+
"""Zero-dependency sports quantitative analytics SDK for AsianScore (https://asianscore.com)."""
|
|
2
|
+
from .math_models import (
|
|
3
|
+
devig_multiplicative,
|
|
4
|
+
devig_power,
|
|
5
|
+
kelly_fraction,
|
|
6
|
+
dixon_coles_probs,
|
|
7
|
+
brier_score,
|
|
8
|
+
fetch_live_summary,
|
|
9
|
+
)
|
|
10
|
+
|
|
11
|
+
__all__ = [
|
|
12
|
+
"devig_multiplicative",
|
|
13
|
+
"devig_power",
|
|
14
|
+
"kelly_fraction",
|
|
15
|
+
"dixon_coles_probs",
|
|
16
|
+
"brier_score",
|
|
17
|
+
"fetch_live_summary",
|
|
18
|
+
]
|
|
@@ -0,0 +1,112 @@
|
|
|
1
|
+
"""Core sports probability, devigging, Kelly staking, and Dixon-Coles models.
|
|
2
|
+
|
|
3
|
+
Reference web calculators & live ledger: https://asianscore.com/tools.html
|
|
4
|
+
"""
|
|
5
|
+
from __future__ import annotations
|
|
6
|
+
|
|
7
|
+
import json
|
|
8
|
+
import math
|
|
9
|
+
import urllib.request
|
|
10
|
+
from typing import Dict, Iterable, List, Sequence, Tuple
|
|
11
|
+
|
|
12
|
+
|
|
13
|
+
def devig_multiplicative(decimal_odds: Sequence[float]) -> List[float]:
|
|
14
|
+
"""Remove bookmaker overround proportionally from decimal odds."""
|
|
15
|
+
implied = [1.0 / o for o in decimal_odds if o > 1.0]
|
|
16
|
+
if len(implied) != len(decimal_odds) or not implied:
|
|
17
|
+
raise ValueError("All decimal odds must be > 1.0")
|
|
18
|
+
total = sum(implied)
|
|
19
|
+
return [p / total for p in implied]
|
|
20
|
+
|
|
21
|
+
|
|
22
|
+
def devig_power(decimal_odds: Sequence[float], tol: float = 1e-9, max_iter: int = 80) -> List[float]:
|
|
23
|
+
"""Power-method devigging (solves sum((1/o)^k) = 1 via bisection)."""
|
|
24
|
+
implied = [1.0 / o for o in decimal_odds if o > 1.0]
|
|
25
|
+
if len(implied) != len(decimal_odds) or not implied:
|
|
26
|
+
raise ValueError("All decimal odds must be > 1.0")
|
|
27
|
+
lo, hi = 0.2, 5.0
|
|
28
|
+
for _ in range(max_iter):
|
|
29
|
+
mid = 0.5 * (lo + hi)
|
|
30
|
+
s = sum(p ** mid for p in implied)
|
|
31
|
+
if abs(s - 1.0) < tol:
|
|
32
|
+
lo = hi = mid
|
|
33
|
+
break
|
|
34
|
+
if s > 1.0:
|
|
35
|
+
lo = mid
|
|
36
|
+
else:
|
|
37
|
+
hi = mid
|
|
38
|
+
k = 0.5 * (lo + hi)
|
|
39
|
+
raw = [p ** k for p in implied]
|
|
40
|
+
tot = sum(raw)
|
|
41
|
+
return [p / tot for p in raw]
|
|
42
|
+
|
|
43
|
+
|
|
44
|
+
def kelly_fraction(fair_prob: float, decimal_odds: float, scale: float = 0.25) -> float:
|
|
45
|
+
"""Fractional Kelly criterion stake fraction (default quarter-Kelly)."""
|
|
46
|
+
if not (0.0 < fair_prob < 1.0) or decimal_odds <= 1.0:
|
|
47
|
+
return 0.0
|
|
48
|
+
b = decimal_odds - 1.0
|
|
49
|
+
f_star = (fair_prob * b - (1.0 - fair_prob)) / b
|
|
50
|
+
return max(0.0, f_star * scale)
|
|
51
|
+
|
|
52
|
+
|
|
53
|
+
def _poisson_pmf(lam: float, k: int) -> float:
|
|
54
|
+
return math.exp(-lam) * (lam ** k) / math.factorial(k)
|
|
55
|
+
|
|
56
|
+
|
|
57
|
+
def dixon_coles_probs(
|
|
58
|
+
lambda_home: float,
|
|
59
|
+
lambda_away: float,
|
|
60
|
+
rho: float = -0.05,
|
|
61
|
+
max_goals: int = 10,
|
|
62
|
+
) -> Dict[str, float]:
|
|
63
|
+
"""Dixon-Coles low-score adjusted bivariate Poisson 1X2 & O/U 2.5 probabilities."""
|
|
64
|
+
home_win = draw = away_win = over_25 = 0.0
|
|
65
|
+
total = 0.0
|
|
66
|
+
for h in range(max_goals + 1):
|
|
67
|
+
ph = _poisson_pmf(lambda_home, h)
|
|
68
|
+
for a in range(max_goals + 1):
|
|
69
|
+
pa = _poisson_pmf(lambda_away, a)
|
|
70
|
+
tau = 1.0
|
|
71
|
+
if h == 0 and a == 0:
|
|
72
|
+
tau = 1.0 - lambda_home * lambda_away * rho
|
|
73
|
+
elif h == 0 and a == 1:
|
|
74
|
+
tau = 1.0 + lambda_home * rho
|
|
75
|
+
elif h == 1 and a == 0:
|
|
76
|
+
tau = 1.0 + lambda_away * rho
|
|
77
|
+
elif h == 1 and a == 1:
|
|
78
|
+
tau = 1.0 - rho
|
|
79
|
+
p = max(0.0, ph * pa * tau)
|
|
80
|
+
total += p
|
|
81
|
+
if h > a:
|
|
82
|
+
home_win += p
|
|
83
|
+
elif h == a:
|
|
84
|
+
draw += p
|
|
85
|
+
else:
|
|
86
|
+
away_win += p
|
|
87
|
+
if h + a >= 3:
|
|
88
|
+
over_25 += p
|
|
89
|
+
return {
|
|
90
|
+
"home": round(home_win / total, 6),
|
|
91
|
+
"draw": round(draw / total, 6),
|
|
92
|
+
"away": round(away_win / total, 6),
|
|
93
|
+
"over_2_5": round(over_25 / total, 6),
|
|
94
|
+
}
|
|
95
|
+
|
|
96
|
+
|
|
97
|
+
def brier_score(predictions: Iterable[Tuple[Sequence[float], int]]) -> float:
|
|
98
|
+
"""Multi-class Brier score (mean squared error across outcome probabilities)."""
|
|
99
|
+
total = 0.0
|
|
100
|
+
count = 0
|
|
101
|
+
for probs, outcome_idx in predictions:
|
|
102
|
+
err = sum(((1.0 if i == outcome_idx else 0.0) - p) ** 2 for i, p in enumerate(probs)) / len(probs)
|
|
103
|
+
total += err
|
|
104
|
+
count += 1
|
|
105
|
+
return total / count if count else 0.0
|
|
106
|
+
|
|
107
|
+
|
|
108
|
+
def fetch_live_summary(url: str = "https://asianscore.com/api/widget/summary", timeout: int = 15) -> dict:
|
|
109
|
+
"""Fetch live fixtures and verifiable SHA-256 ledger summary from AsianScore."""
|
|
110
|
+
req = urllib.request.Request(url, headers={"User-Agent": "asianscore-quant-sdk/1.0"})
|
|
111
|
+
with urllib.request.urlopen(req, timeout=timeout) as resp:
|
|
112
|
+
return json.loads(resp.read().decode("utf-8"))
|
|
@@ -0,0 +1,78 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: asianscore-quant
|
|
3
|
+
Version: 1.0.0
|
|
4
|
+
Summary: Zero-dependency sports quantitative analytics SDK: Power/Multiplicative de-vigging, Dixon-Coles xG Poisson grid, Kelly criterion, Brier score auditing, and live AsianScore feed client.
|
|
5
|
+
Author-email: AsianScore Quantitative Research Lab <pypi@aimail.men>
|
|
6
|
+
License: MIT
|
|
7
|
+
Project-URL: Homepage, https://asianscore.com/
|
|
8
|
+
Project-URL: Documentation, https://asianscore.com/tools.html
|
|
9
|
+
Project-URL: SHA-256 Audit Ledger, https://asianscore.com/models.html#ledger
|
|
10
|
+
Project-URL: Zenodo Dataset (DOI), https://doi.org/10.5281/zenodo.24441036
|
|
11
|
+
Project-URL: Repository, https://github.com/AIYIXIU/asianscore-quant-sdk
|
|
12
|
+
Project-URL: Go Package, https://pkg.go.dev/github.com/AIYIXIU/asianscore-go
|
|
13
|
+
Keywords: sports-analytics,expected-goals,xg,dixon-coles,poisson,devig,kelly-criterion,brier-score,football,asianscore
|
|
14
|
+
Classifier: Development Status :: 5 - Production/Stable
|
|
15
|
+
Classifier: Intended Audience :: Developers
|
|
16
|
+
Classifier: Intended Audience :: Science/Research
|
|
17
|
+
Classifier: License :: OSI Approved :: MIT License
|
|
18
|
+
Classifier: Programming Language :: Python :: 3
|
|
19
|
+
Classifier: Programming Language :: Python :: 3.9
|
|
20
|
+
Classifier: Programming Language :: Python :: 3.10
|
|
21
|
+
Classifier: Programming Language :: Python :: 3.11
|
|
22
|
+
Classifier: Programming Language :: Python :: 3.12
|
|
23
|
+
Classifier: Topic :: Scientific/Engineering :: Mathematics
|
|
24
|
+
Requires-Python: >=3.9
|
|
25
|
+
Description-Content-Type: text/markdown
|
|
26
|
+
License-File: LICENSE
|
|
27
|
+
Dynamic: license-file
|
|
28
|
+
|
|
29
|
+
# AsianScore Quant SDK (`asianscore-quant`)
|
|
30
|
+
|
|
31
|
+
[](https://pypi.org/project/asianscore-quant/)
|
|
32
|
+
[](https://doi.org/10.5281/zenodo.24441036)
|
|
33
|
+
[](https://github.com/AIYIXIU/asianscore-quant-sdk/blob/main/LICENSE)
|
|
34
|
+
|
|
35
|
+
Zero-dependency Python sports quantitative toolkit implementing the exact mathematical primitives powering [AsianScore](https://asianscore.com/):
|
|
36
|
+
|
|
37
|
+
- **Odds De-Vigging:** Multiplicative & Power-method fair probability solvers (`devig_multiplicative`, `devig_power`)
|
|
38
|
+
- **Bankroll Management:** Fractional Kelly Criterion (`kelly_fraction`)
|
|
39
|
+
- **Football xG Modeling:** Dixon-Coles low-score correlation bivariate Poisson grid (`dixon_coles_probs`)
|
|
40
|
+
- **Calibration Auditing:** Multi-class Brier score (`brier_score`)
|
|
41
|
+
- **Live Open Feed Client:** `fetch_live_summary()` querying `https://asianscore.com/api/widget/summary` (CORS-enabled, no API key required)
|
|
42
|
+
|
|
43
|
+
## Installation
|
|
44
|
+
|
|
45
|
+
```bash
|
|
46
|
+
pip install asianscore-quant
|
|
47
|
+
```
|
|
48
|
+
|
|
49
|
+
## Quick Start
|
|
50
|
+
|
|
51
|
+
```python
|
|
52
|
+
from asianscore_quant import devig_power, dixon_coles_probs, kelly_fraction, brier_score
|
|
53
|
+
|
|
54
|
+
fair = devig_power([2.10, 3.40, 3.60])
|
|
55
|
+
grid = dixon_coles_probs(1.62, 1.18, rho=-0.08)
|
|
56
|
+
stake = kelly_fraction(prob=0.52, decimal_odds=2.10, fraction=0.25)
|
|
57
|
+
score = brier_score([0.55, 0.25, 0.20], actual_index=0)
|
|
58
|
+
```
|
|
59
|
+
|
|
60
|
+
## Open Platform & Ecosystem
|
|
61
|
+
|
|
62
|
+
- **Official Platform:** https://asianscore.com/
|
|
63
|
+
- **Interactive Web Calculators:** https://asianscore.com/tools.html
|
|
64
|
+
- **SHA-256 Verifiable AI Prediction Ledger:** https://asianscore.com/models.html#ledger
|
|
65
|
+
- **Zenodo Academic Dataset (CERN DOI):** https://doi.org/10.5281/zenodo.24441036
|
|
66
|
+
- **Docs (GitHub Pages):** https://aiyixiu.github.io/asianscore-quant-sdk/
|
|
67
|
+
- **Official Go Package (`pkg.go.dev`):** https://pkg.go.dev/github.com/AIYIXIU/asianscore-go
|
|
68
|
+
- **Embeddable Live Widget:** https://aiyixiu.github.io/asianscore-live-widget/
|
|
69
|
+
|
|
70
|
+
## Verification
|
|
71
|
+
|
|
72
|
+
```bash
|
|
73
|
+
python3 -m unittest tests/test_quant.py
|
|
74
|
+
```
|
|
75
|
+
|
|
76
|
+
## License
|
|
77
|
+
|
|
78
|
+
MIT
|
|
@@ -0,0 +1,10 @@
|
|
|
1
|
+
LICENSE
|
|
2
|
+
README.md
|
|
3
|
+
pyproject.toml
|
|
4
|
+
python/asianscore_quant/__init__.py
|
|
5
|
+
python/asianscore_quant/math_models.py
|
|
6
|
+
python/asianscore_quant.egg-info/PKG-INFO
|
|
7
|
+
python/asianscore_quant.egg-info/SOURCES.txt
|
|
8
|
+
python/asianscore_quant.egg-info/dependency_links.txt
|
|
9
|
+
python/asianscore_quant.egg-info/top_level.txt
|
|
10
|
+
tests/test_quant.py
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
asianscore_quant
|
|
@@ -0,0 +1,38 @@
|
|
|
1
|
+
import sys
|
|
2
|
+
import unittest
|
|
3
|
+
from pathlib import Path
|
|
4
|
+
|
|
5
|
+
sys.path.insert(0, str(Path(__file__).resolve().parents[1] / "python"))
|
|
6
|
+
|
|
7
|
+
from asianscore_quant import (
|
|
8
|
+
brier_score,
|
|
9
|
+
devig_multiplicative,
|
|
10
|
+
devig_power,
|
|
11
|
+
dixon_coles_probs,
|
|
12
|
+
kelly_fraction,
|
|
13
|
+
)
|
|
14
|
+
|
|
15
|
+
|
|
16
|
+
class QuantModelsTest(unittest.TestCase):
|
|
17
|
+
def test_devig_sums_to_one(self):
|
|
18
|
+
p1 = devig_multiplicative([1.95, 3.50, 4.10])
|
|
19
|
+
p2 = devig_power([1.95, 3.50, 4.10])
|
|
20
|
+
self.assertAlmostEqual(sum(p1), 1.0, places=6)
|
|
21
|
+
self.assertAlmostEqual(sum(p2), 1.0, places=6)
|
|
22
|
+
self.assertGreater(p2[0], p1[0]) # favorite gets higher weight under power devig
|
|
23
|
+
|
|
24
|
+
def test_kelly(self):
|
|
25
|
+
self.assertAlmostEqual(kelly_fraction(0.55, 2.0, scale=1.0), 0.10, places=6)
|
|
26
|
+
self.assertEqual(kelly_fraction(0.45, 2.0), 0.0)
|
|
27
|
+
|
|
28
|
+
def test_dixon_coles(self):
|
|
29
|
+
probs = dixon_coles_probs(1.65, 1.05, rho=-0.05)
|
|
30
|
+
self.assertAlmostEqual(probs["home"] + probs["draw"] + probs["away"], 1.0, places=4)
|
|
31
|
+
self.assertGreater(probs["home"], probs["away"])
|
|
32
|
+
|
|
33
|
+
def test_brier(self):
|
|
34
|
+
self.assertAlmostEqual(brier_score([([0.7, 0.2, 0.1], 0)]), (0.09 + 0.04 + 0.01) / 3, places=6)
|
|
35
|
+
|
|
36
|
+
|
|
37
|
+
if __name__ == "__main__":
|
|
38
|
+
unittest.main()
|