tracegamma 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.
- tracegamma-0.1.0/LICENSE +21 -0
- tracegamma-0.1.0/PKG-INFO +102 -0
- tracegamma-0.1.0/README.md +71 -0
- tracegamma-0.1.0/pyproject.toml +49 -0
- tracegamma-0.1.0/setup.cfg +4 -0
- tracegamma-0.1.0/src/tracegamma/__init__.py +80 -0
- tracegamma-0.1.0/src/tracegamma/bb.py +52 -0
- tracegamma-0.1.0/src/tracegamma/core.py +328 -0
- tracegamma-0.1.0/src/tracegamma/correction.py +91 -0
- tracegamma-0.1.0/src/tracegamma/data/parity.json +594 -0
- tracegamma-0.1.0/src/tracegamma/defects.py +114 -0
- tracegamma-0.1.0/src/tracegamma/exact.py +179 -0
- tracegamma-0.1.0/src/tracegamma.egg-info/PKG-INFO +102 -0
- tracegamma-0.1.0/src/tracegamma.egg-info/SOURCES.txt +19 -0
- tracegamma-0.1.0/src/tracegamma.egg-info/dependency_links.txt +1 -0
- tracegamma-0.1.0/src/tracegamma.egg-info/requires.txt +5 -0
- tracegamma-0.1.0/src/tracegamma.egg-info/top_level.txt +1 -0
- tracegamma-0.1.0/tests/test_core.py +79 -0
- tracegamma-0.1.0/tests/test_deposited.py +61 -0
- tracegamma-0.1.0/tests/test_exact.py +52 -0
- tracegamma-0.1.0/tests/test_identities.py +48 -0
tracegamma-0.1.0/LICENSE
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2026 William J. Dwyer
|
|
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.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
21
|
+
SOFTWARE.
|
|
@@ -0,0 +1,102 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: tracegamma
|
|
3
|
+
Version: 0.1.0
|
|
4
|
+
Summary: The trace-gamma distribution: a three-cumulant calibrated reference for quadratic-form tests
|
|
5
|
+
Author: William J. Dwyer
|
|
6
|
+
License: MIT
|
|
7
|
+
Project-URL: Homepage, https://github.com/wjdwyer/tracegamma
|
|
8
|
+
Project-URL: Repository, https://github.com/wjdwyer/tracegamma
|
|
9
|
+
Project-URL: Issues, https://github.com/wjdwyer/tracegamma/issues
|
|
10
|
+
Project-URL: Changelog, https://github.com/wjdwyer/tracegamma/blob/main/python/CHANGELOG.md
|
|
11
|
+
Project-URL: Deposit, https://doi.org/10.5281/zenodo.22060095
|
|
12
|
+
Keywords: quadratic form,generalized chi-square,Satterthwaite,Welch,Behrens-Fisher,heteroscedasticity,reference distribution
|
|
13
|
+
Classifier: Development Status :: 4 - Beta
|
|
14
|
+
Classifier: Intended Audience :: Science/Research
|
|
15
|
+
Classifier: License :: OSI Approved :: MIT License
|
|
16
|
+
Classifier: Programming Language :: Python :: 3
|
|
17
|
+
Classifier: Programming Language :: Python :: 3.10
|
|
18
|
+
Classifier: Programming Language :: Python :: 3.11
|
|
19
|
+
Classifier: Programming Language :: Python :: 3.12
|
|
20
|
+
Classifier: Programming Language :: Python :: 3.13
|
|
21
|
+
Classifier: Programming Language :: Python :: 3.14
|
|
22
|
+
Classifier: Topic :: Scientific/Engineering :: Mathematics
|
|
23
|
+
Requires-Python: >=3.10
|
|
24
|
+
Description-Content-Type: text/markdown
|
|
25
|
+
License-File: LICENSE
|
|
26
|
+
Requires-Dist: numpy>=1.21
|
|
27
|
+
Requires-Dist: scipy>=1.7
|
|
28
|
+
Provides-Extra: test
|
|
29
|
+
Requires-Dist: pytest>=7; extra == "test"
|
|
30
|
+
Dynamic: license-file
|
|
31
|
+
|
|
32
|
+
# tracegamma (Python)
|
|
33
|
+
|
|
34
|
+
The **trace-gamma** is the closed-form shifted, scaled chi-square matched to the
|
|
35
|
+
first three cumulants of a quadratic form in near-normal variables. Those
|
|
36
|
+
cumulants are exact traces of one matrix product (Lancaster's identity), and the
|
|
37
|
+
three-moment fit is Solomon and Stephens's. Welch's F is the **two-cumulant
|
|
38
|
+
member** of the family and Pearson's chi-square the **equal-eigenvalue member**,
|
|
39
|
+
so the family contains the classical references rather than competing with them.
|
|
40
|
+
|
|
41
|
+
The package also provides the exact generalized chi-square by Imhof inversion,
|
|
42
|
+
so the closed form can be checked against the truth rather than against
|
|
43
|
+
simulation; the kurtosis and tail-rate defects that govern the error; and an
|
|
44
|
+
estimable correction for non-normal inputs.
|
|
45
|
+
|
|
46
|
+
## Two things worth knowing before you use it
|
|
47
|
+
|
|
48
|
+
**The two-cumulant branch is requested, not discovered.** The fit falls back to
|
|
49
|
+
Satterthwaite's two-moment form when the standardized skewness does not exceed
|
|
50
|
+
`1e-6`. No spectrum of positive eigenvalues reaches that in practice: the
|
|
51
|
+
skewness is `2*sqrt(2)*p3/p2^1.5`, minimised at equal eigenvalues where it is
|
|
52
|
+
`2*sqrt(2)/sqrt(m)`, so `1e-6` would need about 8e12 eigenvalues. You reach the
|
|
53
|
+
branch by supplying `k3 = 0` — which is exactly how the paper states the Welch
|
|
54
|
+
special case — or a negative `k3`, which is discarded with a warning. Inspect
|
|
55
|
+
the fitted object's `branch` to see which fit you have.
|
|
56
|
+
|
|
57
|
+
**The exact-inversion method matters for small spectra.** `method = "quad"`, the
|
|
58
|
+
single adaptive call used in the deposited research code, is accurate for ten or
|
|
59
|
+
more eigenvalues and **unusable below three**: measured against the exact oracle
|
|
60
|
+
`sum of m unit chi-squares = chi-square on m`, its worst relative error is about
|
|
61
|
+
0.69 at m = 1 and 1.0e-2 at m = 2. `method = "panel"`, the default, integrates
|
|
62
|
+
between successive zeros and stays below 5e-9 throughout. It is slower. Both are
|
|
63
|
+
kept: `"quad"` reproduces the published cross-check numbers exactly.
|
|
64
|
+
|
|
65
|
+
## Provenance and honest limits
|
|
66
|
+
|
|
67
|
+
Ported from the research code deposited with the paper (Zenodo concept DOI
|
|
68
|
+
`10.5281/zenodo.22060095`). The port changes no formula and no constant; the
|
|
69
|
+
Python test suite pins it to the deposited module value for value. Two
|
|
70
|
+
departures are deliberate and documented in the source: the fitted branch is
|
|
71
|
+
recorded and a negative third cumulant warns, and the exact-inversion default is
|
|
72
|
+
the repaired panel method rather than the single quad call.
|
|
73
|
+
|
|
74
|
+
MIT licensed.
|
|
75
|
+
|
|
76
|
+
## Install
|
|
77
|
+
|
|
78
|
+
```
|
|
79
|
+
pip install tracegamma
|
|
80
|
+
```
|
|
81
|
+
|
|
82
|
+
## Use
|
|
83
|
+
|
|
84
|
+
```python
|
|
85
|
+
from tracegamma import TraceGamma, imhof_ppf, kurtosis_defect
|
|
86
|
+
|
|
87
|
+
lam = [4.0, 2.0, 1.0, 1.0] # eigenvalues of M^{1/2} D M^{1/2}
|
|
88
|
+
tg = TraceGamma.from_eigs(lam)
|
|
89
|
+
tg.cutoff(0.05) # 21.249374...
|
|
90
|
+
tg.branch # 'three-cumulant'
|
|
91
|
+
|
|
92
|
+
imhof_ppf(0.95, lam) # 20.919385... the exact quantile
|
|
93
|
+
kurtosis_defect(lam) # what drives the shoulder error
|
|
94
|
+
```
|
|
95
|
+
|
|
96
|
+
Also `TraceGamma.from_quadform(M, D)` and `TraceGamma(k1, k2, k3)`.
|
|
97
|
+
|
|
98
|
+
## Test
|
|
99
|
+
|
|
100
|
+
```
|
|
101
|
+
pytest --doctest-modules src/tracegamma tests
|
|
102
|
+
```
|
|
@@ -0,0 +1,71 @@
|
|
|
1
|
+
# tracegamma (Python)
|
|
2
|
+
|
|
3
|
+
The **trace-gamma** is the closed-form shifted, scaled chi-square matched to the
|
|
4
|
+
first three cumulants of a quadratic form in near-normal variables. Those
|
|
5
|
+
cumulants are exact traces of one matrix product (Lancaster's identity), and the
|
|
6
|
+
three-moment fit is Solomon and Stephens's. Welch's F is the **two-cumulant
|
|
7
|
+
member** of the family and Pearson's chi-square the **equal-eigenvalue member**,
|
|
8
|
+
so the family contains the classical references rather than competing with them.
|
|
9
|
+
|
|
10
|
+
The package also provides the exact generalized chi-square by Imhof inversion,
|
|
11
|
+
so the closed form can be checked against the truth rather than against
|
|
12
|
+
simulation; the kurtosis and tail-rate defects that govern the error; and an
|
|
13
|
+
estimable correction for non-normal inputs.
|
|
14
|
+
|
|
15
|
+
## Two things worth knowing before you use it
|
|
16
|
+
|
|
17
|
+
**The two-cumulant branch is requested, not discovered.** The fit falls back to
|
|
18
|
+
Satterthwaite's two-moment form when the standardized skewness does not exceed
|
|
19
|
+
`1e-6`. No spectrum of positive eigenvalues reaches that in practice: the
|
|
20
|
+
skewness is `2*sqrt(2)*p3/p2^1.5`, minimised at equal eigenvalues where it is
|
|
21
|
+
`2*sqrt(2)/sqrt(m)`, so `1e-6` would need about 8e12 eigenvalues. You reach the
|
|
22
|
+
branch by supplying `k3 = 0` — which is exactly how the paper states the Welch
|
|
23
|
+
special case — or a negative `k3`, which is discarded with a warning. Inspect
|
|
24
|
+
the fitted object's `branch` to see which fit you have.
|
|
25
|
+
|
|
26
|
+
**The exact-inversion method matters for small spectra.** `method = "quad"`, the
|
|
27
|
+
single adaptive call used in the deposited research code, is accurate for ten or
|
|
28
|
+
more eigenvalues and **unusable below three**: measured against the exact oracle
|
|
29
|
+
`sum of m unit chi-squares = chi-square on m`, its worst relative error is about
|
|
30
|
+
0.69 at m = 1 and 1.0e-2 at m = 2. `method = "panel"`, the default, integrates
|
|
31
|
+
between successive zeros and stays below 5e-9 throughout. It is slower. Both are
|
|
32
|
+
kept: `"quad"` reproduces the published cross-check numbers exactly.
|
|
33
|
+
|
|
34
|
+
## Provenance and honest limits
|
|
35
|
+
|
|
36
|
+
Ported from the research code deposited with the paper (Zenodo concept DOI
|
|
37
|
+
`10.5281/zenodo.22060095`). The port changes no formula and no constant; the
|
|
38
|
+
Python test suite pins it to the deposited module value for value. Two
|
|
39
|
+
departures are deliberate and documented in the source: the fitted branch is
|
|
40
|
+
recorded and a negative third cumulant warns, and the exact-inversion default is
|
|
41
|
+
the repaired panel method rather than the single quad call.
|
|
42
|
+
|
|
43
|
+
MIT licensed.
|
|
44
|
+
|
|
45
|
+
## Install
|
|
46
|
+
|
|
47
|
+
```
|
|
48
|
+
pip install tracegamma
|
|
49
|
+
```
|
|
50
|
+
|
|
51
|
+
## Use
|
|
52
|
+
|
|
53
|
+
```python
|
|
54
|
+
from tracegamma import TraceGamma, imhof_ppf, kurtosis_defect
|
|
55
|
+
|
|
56
|
+
lam = [4.0, 2.0, 1.0, 1.0] # eigenvalues of M^{1/2} D M^{1/2}
|
|
57
|
+
tg = TraceGamma.from_eigs(lam)
|
|
58
|
+
tg.cutoff(0.05) # 21.249374...
|
|
59
|
+
tg.branch # 'three-cumulant'
|
|
60
|
+
|
|
61
|
+
imhof_ppf(0.95, lam) # 20.919385... the exact quantile
|
|
62
|
+
kurtosis_defect(lam) # what drives the shoulder error
|
|
63
|
+
```
|
|
64
|
+
|
|
65
|
+
Also `TraceGamma.from_quadform(M, D)` and `TraceGamma(k1, k2, k3)`.
|
|
66
|
+
|
|
67
|
+
## Test
|
|
68
|
+
|
|
69
|
+
```
|
|
70
|
+
pytest --doctest-modules src/tracegamma tests
|
|
71
|
+
```
|
|
@@ -0,0 +1,49 @@
|
|
|
1
|
+
[build-system]
|
|
2
|
+
requires = ["setuptools>=61", "wheel"]
|
|
3
|
+
build-backend = "setuptools.build_meta"
|
|
4
|
+
|
|
5
|
+
[project]
|
|
6
|
+
name = "tracegamma"
|
|
7
|
+
version = "0.1.0"
|
|
8
|
+
description = "The trace-gamma distribution: a three-cumulant calibrated reference for quadratic-form tests"
|
|
9
|
+
readme = "README.md"
|
|
10
|
+
requires-python = ">=3.10"
|
|
11
|
+
license = {text = "MIT"}
|
|
12
|
+
authors = [{name = "William J. Dwyer"}]
|
|
13
|
+
keywords = ["quadratic form", "generalized chi-square", "Satterthwaite",
|
|
14
|
+
"Welch", "Behrens-Fisher", "heteroscedasticity", "reference distribution"]
|
|
15
|
+
classifiers = [
|
|
16
|
+
"Development Status :: 4 - Beta",
|
|
17
|
+
"Intended Audience :: Science/Research",
|
|
18
|
+
"License :: OSI Approved :: MIT License",
|
|
19
|
+
"Programming Language :: Python :: 3",
|
|
20
|
+
# One classifier per version the CI matrix actually runs. If a version is
|
|
21
|
+
# added or removed here it must be added or removed in python.yml too: an
|
|
22
|
+
# untested classifier is a claim, not a fact.
|
|
23
|
+
"Programming Language :: Python :: 3.10",
|
|
24
|
+
"Programming Language :: Python :: 3.11",
|
|
25
|
+
"Programming Language :: Python :: 3.12",
|
|
26
|
+
"Programming Language :: Python :: 3.13",
|
|
27
|
+
"Programming Language :: Python :: 3.14",
|
|
28
|
+
"Topic :: Scientific/Engineering :: Mathematics",
|
|
29
|
+
]
|
|
30
|
+
dependencies = ["numpy>=1.21", "scipy>=1.7"]
|
|
31
|
+
|
|
32
|
+
[project.optional-dependencies]
|
|
33
|
+
test = ["pytest>=7"]
|
|
34
|
+
|
|
35
|
+
[project.urls]
|
|
36
|
+
Homepage = "https://github.com/wjdwyer/tracegamma"
|
|
37
|
+
Repository = "https://github.com/wjdwyer/tracegamma"
|
|
38
|
+
Issues = "https://github.com/wjdwyer/tracegamma/issues"
|
|
39
|
+
Changelog = "https://github.com/wjdwyer/tracegamma/blob/main/python/CHANGELOG.md"
|
|
40
|
+
Deposit = "https://doi.org/10.5281/zenodo.22060095"
|
|
41
|
+
|
|
42
|
+
[tool.setuptools.packages.find]
|
|
43
|
+
where = ["src"]
|
|
44
|
+
|
|
45
|
+
[tool.setuptools.package-data]
|
|
46
|
+
tracegamma = ["data/*.json"]
|
|
47
|
+
|
|
48
|
+
[tool.pytest.ini_options]
|
|
49
|
+
testpaths = ["tests"]
|
|
@@ -0,0 +1,80 @@
|
|
|
1
|
+
# (c) 2026 William J. Dwyer. MIT License.
|
|
2
|
+
r"""tracegamma -- the three-cumulant calibrated reference for quadratic-form tests.
|
|
3
|
+
|
|
4
|
+
The trace-gamma is the closed-form shifted, scaled chi-square matched to the
|
|
5
|
+
first three cumulants of a quadratic form in near-normal variables. Welch's F is
|
|
6
|
+
its two-cumulant member and Pearson's chi-square its equal-eigenvalue member, so
|
|
7
|
+
the family *contains* the classical references rather than competing with them.
|
|
8
|
+
|
|
9
|
+
Quick start
|
|
10
|
+
-----------
|
|
11
|
+
>>> from tracegamma import TraceGamma
|
|
12
|
+
>>> tg = TraceGamma.from_eigs([4.0, 2.0, 1.0, 1.0])
|
|
13
|
+
>>> round(tg.cutoff(0.05), 6)
|
|
14
|
+
21.249374
|
|
15
|
+
>>> tg.branch
|
|
16
|
+
'three-cumulant'
|
|
17
|
+
|
|
18
|
+
Checking the closed form against the exact law:
|
|
19
|
+
|
|
20
|
+
>>> from tracegamma import imhof_ppf
|
|
21
|
+
>>> lam = [4.0, 2.0, 1.0, 1.0]
|
|
22
|
+
>>> round(TraceGamma.from_eigs(lam).ppf(0.95) - imhof_ppf(0.95, lam), 4)
|
|
23
|
+
0.33
|
|
24
|
+
|
|
25
|
+
Citation
|
|
26
|
+
--------
|
|
27
|
+
Dwyer, W. J. (2026). *The trace-gamma distribution: a three-cumulant calibrated
|
|
28
|
+
reference for quadratic-form tests, with Welch's F and Pearson's chi-square as
|
|
29
|
+
special cases.* Zenodo concept DOI 10.5281/zenodo.22060095.
|
|
30
|
+
"""
|
|
31
|
+
from .core import (
|
|
32
|
+
SKEW_FLOOR,
|
|
33
|
+
VAR_FLOOR,
|
|
34
|
+
TraceGamma,
|
|
35
|
+
cumulants_from_eigs,
|
|
36
|
+
cumulants_from_traces,
|
|
37
|
+
params_from_power_sums,
|
|
38
|
+
traces_from_quadform,
|
|
39
|
+
)
|
|
40
|
+
from .exact import imhof_cdf, imhof_ppf, imhof_sf
|
|
41
|
+
from .defects import (
|
|
42
|
+
cf_modulus,
|
|
43
|
+
kurtosis_defect,
|
|
44
|
+
rho_ratio,
|
|
45
|
+
standardized_cumulant,
|
|
46
|
+
tail_rate_defect,
|
|
47
|
+
)
|
|
48
|
+
from .correction import (
|
|
49
|
+
estimate_moments,
|
|
50
|
+
kurtosis_widening,
|
|
51
|
+
marginal_cumulants,
|
|
52
|
+
tracegamma_from_marginal,
|
|
53
|
+
)
|
|
54
|
+
from .bb import bb_scale_mixture
|
|
55
|
+
|
|
56
|
+
__version__ = "0.1.0"
|
|
57
|
+
|
|
58
|
+
__all__ = [
|
|
59
|
+
"TraceGamma",
|
|
60
|
+
"cumulants_from_traces",
|
|
61
|
+
"traces_from_quadform",
|
|
62
|
+
"cumulants_from_eigs",
|
|
63
|
+
"params_from_power_sums",
|
|
64
|
+
"SKEW_FLOOR",
|
|
65
|
+
"VAR_FLOOR",
|
|
66
|
+
"imhof_sf",
|
|
67
|
+
"imhof_cdf",
|
|
68
|
+
"imhof_ppf",
|
|
69
|
+
"kurtosis_defect",
|
|
70
|
+
"tail_rate_defect",
|
|
71
|
+
"standardized_cumulant",
|
|
72
|
+
"rho_ratio",
|
|
73
|
+
"cf_modulus",
|
|
74
|
+
"marginal_cumulants",
|
|
75
|
+
"tracegamma_from_marginal",
|
|
76
|
+
"estimate_moments",
|
|
77
|
+
"kurtosis_widening",
|
|
78
|
+
"bb_scale_mixture",
|
|
79
|
+
"__version__",
|
|
80
|
+
]
|
|
@@ -0,0 +1,52 @@
|
|
|
1
|
+
# (c) 2026 William J. Dwyer. MIT License.
|
|
2
|
+
r"""The Berger-Boos scale mixture: the guaranteed test's oracle law.
|
|
3
|
+
|
|
4
|
+
The guaranteed statistic ``T_BB`` dilates the trace-gamma cutoff by a
|
|
5
|
+
data-dependent log-variance scale. Equivalently, the deployed statistic is the
|
|
6
|
+
trace-gamma core *divided* by an independent scale, so its law is a scale
|
|
7
|
+
mixture of the core over the log-chi-square dispersion -- the object the paper
|
|
8
|
+
names the **Berger-Boos scale mixture**.
|
|
9
|
+
|
|
10
|
+
Scope warning
|
|
11
|
+
-------------
|
|
12
|
+
This module is here because the scope chosen for the package was "everything in
|
|
13
|
+
the project tree". The statistic itself belongs to a different paper in the
|
|
14
|
+
series (m01A / T_BB), not to the trace-gamma paper. Nothing in the trace-gamma
|
|
15
|
+
propositions depends on it. Treat it as a convenience, and read that paper
|
|
16
|
+
before relying on the constant ``ks``.
|
|
17
|
+
"""
|
|
18
|
+
from __future__ import annotations
|
|
19
|
+
|
|
20
|
+
import numpy as np
|
|
21
|
+
|
|
22
|
+
__all__ = ["bb_scale_mixture"]
|
|
23
|
+
|
|
24
|
+
|
|
25
|
+
def bb_scale_mixture(core_cutoff, ks, rbar, k):
|
|
26
|
+
r"""Inflate a trace-gamma cutoff to the guaranteed ``T_BB`` cutoff.
|
|
27
|
+
|
|
28
|
+
``T_BB`` rejects when ``Q_obs > exp(ks sqrt(rbar/k)) * core_cutoff``.
|
|
29
|
+
|
|
30
|
+
Parameters
|
|
31
|
+
----------
|
|
32
|
+
core_cutoff : float or array_like
|
|
33
|
+
The trace-gamma cutoff, e.g. from
|
|
34
|
+
:meth:`tracegamma.core.TraceGamma.cutoff`.
|
|
35
|
+
ks : float
|
|
36
|
+
The Berger-Boos constant for the design.
|
|
37
|
+
rbar : float
|
|
38
|
+
The design radius (log-variance dispersion).
|
|
39
|
+
k : int
|
|
40
|
+
Number of groups.
|
|
41
|
+
|
|
42
|
+
Returns
|
|
43
|
+
-------
|
|
44
|
+
float or ndarray
|
|
45
|
+
The inflated cutoff.
|
|
46
|
+
|
|
47
|
+
Examples
|
|
48
|
+
--------
|
|
49
|
+
>>> round(float(bb_scale_mixture(10.0, 0.0, 1.0, 4)), 12)
|
|
50
|
+
10.0
|
|
51
|
+
"""
|
|
52
|
+
return np.exp(ks * np.sqrt(rbar / k)) * core_cutoff
|