Closed Datseris closed 1 year ago
Merging #220 (6e273fc) into main (cbb3319) will decrease coverage by
0.04%
. The diff coverage is75.75%
.:exclamation: Current head 6e273fc differs from pull request most recent head 3545738. Consider uploading reports for the commit 3545738 to get more accurate results
@@ Coverage Diff @@
## main #220 +/- ##
==========================================
- Coverage 84.47% 84.43% -0.05%
==========================================
Files 45 45
Lines 1140 1137 -3
==========================================
- Hits 963 960 -3
Misses 177 177
Impacted Files | Coverage Δ | |
---|---|---|
...e_dispersion_entropy/reverse_dispersion_entropy.jl | 100.00% <ø> (ø) |
|
src/entropies/convenience_definitions.jl | 100.00% <ø> (ø) |
|
src/entropies/curado.jl | 100.00% <ø> (ø) |
|
...estimators/nearest_neighbors/KozachenkoLeonenko.jl | 100.00% <ø> (ø) |
|
.../entropies/estimators/nearest_neighbors/Kraskov.jl | 100.00% <ø> (ø) |
|
src/entropies/estimators/nearest_neighbors/Zhu.jl | 100.00% <ø> (ø) |
|
...entropies/estimators/nearest_neighbors/ZhuSingh.jl | 100.00% <ø> (ø) |
|
...ies/estimators/order_statistics/AlizadehArghami.jl | 100.00% <ø> (ø) |
|
...rc/entropies/estimators/order_statistics/Correa.jl | 100.00% <ø> (ø) |
|
.../entropies/estimators/order_statistics/Ebrahimi.jl | 100.00% <ø> (ø) |
|
... and 18 more |
:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more
The analytical tests here (barely) fail due to #214. We can fix that in a separate PR.
What's the maximum entropy for:
? It's not mentione in the docstrings
What's the maximum entropy for Kaniadakis
There's an open issue (#182) on defining the maximum entropy for the Kaniadakis entropy. I haven't found anything in the literature on it yet, although I haven't searched very thoroughly. It might also be the case that nobody has defined it in the literature yet. It would be a fun exercise to figure it out, but I don't have time for that at the moment, so it just means that entropy_normalized
doesn't work for Kaniadakis for now.
What's the maximum entropy for Curado
For total outcomes L
, the maximum for Curado is L * (1 - exp(-b/L)) + exp(-b) - 1
Stretchd expnential
The maximum is defined in the source code. It involves various gamma functions, so better just look there.
Keyword metric
is not used anywhere in ApproxiamteEntropy as far as I can tell. Should we remove it?
Keyword metric is not used anywhere in ApproxiamteEntropy as far as I can tell. Should we remove it?
It should be used inside compute_ϕ
. That said, I don't know if we should document metric
in neither SampleEntropy
nor ApproximateEntropy
, because they both explicitly use the maximum norm. The field metric
is just there because, when implementing them, I thought the methods were generalizable to other metrics. Algorithmically speaking any metric-generalization is possible, but I don't know if theoretically speaking, it makes any sense to use other metrics.
So to not confuse anyone, let's just keep metric
as an internal implementation detail, and just say in the docstrings that the Chebyshev metric is used. That is, both for SampleEntropy
and ApproximateEntropy
.
I would say that we shouldn't have metric
. They use Chebyshev, we use Chebyshev. Overgeneralization is a problem when it serves no purpose.
They use Chebyshev, we use Chebyshev. Overgeneralization is a problem when it serves no purpose.
Sure, that's also fine.
this is good to go. many examples dont\' work at the moment btu this is for a different pr that fixes the examples
this is good to go. many examples dont' work at the moment btu this is for a different pr that fixes the examples
Ok, then I'll review this quickly and merge.
Will you fix the examples, or should I do it?
EDIT: I'll at least fix the tests while reviewing this. Maybe I'll do the examples too if it's nothing major. EDIT 2: I fixed the examples too. It wasn't too much work.
@Datseris Ok, I think I'm done here now. I did
docs/src/dev
folder. I'll merge when CI passes.
The tests and docs passed. The last commit is just a typo for the docs.
Closes #219 . Currently wip.
Changes done:
Entropy
toEntropyDefinition
.EntropyEstimator
toDifferentialEntropyEstimator
.