pyextremes 2.4.0__tar.gz → 2.6.0__tar.gz

{pyextremes-2.4.0 → pyextremes-2.6.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pyextremes
-Version: 2.4.0
+Version: 2.6.0
 Summary: Extreme Value Analysis (EVA) in Python
 Keywords: statistics,extreme,extreme value analysis,eva,coastal,ocean,oceanography,marine,environmental,engineering
 Author: George Bocharov
@@ -26,7 +26,7 @@ Classifier: Topic :: Scientific/Engineering :: Mathematics
 Classifier: Typing :: Typed
 Requires-Dist: numpy>=1.19.0,<3.0.0
 Requires-Dist: scipy>=1.5.0,<2.0.0
-Requires-Dist: pandas>=1.0.0,<3.0.0
+Requires-Dist: pandas>=1.0.0,<4.0.0
 Requires-Dist: emcee>=3.0.3,<4.0.0
 Requires-Dist: matplotlib>=3.3.0,!=3.9.1,<4.0.0
 Requires-Dist: lmoments3>=1.0.8 ; extra == 'full'

{pyextremes-2.4.0 → pyextremes-2.6.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "pyextremes"
-version = "2.4.0"
+version = "2.6.0"
 description = "Extreme Value Analysis (EVA) in Python"
 readme = "README.md"
 requires-python = ">=3.9"
@@ -42,7 +42,7 @@ urls = { "homepage" = "https://georgebv.github.io/pyextremes", "github" = "https
 dependencies = [
     "numpy>=1.19.0,<3.0.0",
     "scipy>=1.5.0,<2.0.0",
-    "pandas>=1.0.0,<3.0.0",
+    "pandas>=1.0.0,<4.0.0",
     "emcee>=3.0.3,<4.0.0",
     "matplotlib>=3.3.0,!=3.9.1,<4.0.0", # avoid yanked 3.9.1 release
 ]

{pyextremes-2.4.0 → pyextremes-2.6.0}/src/pyextremes/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "2.4.0"
+__version__ = "2.6.0"
 __all__ = [
     "EVA",
     "get_extremes",

{pyextremes-2.4.0 → pyextremes-2.6.0}/src/pyextremes/eva.py

@@ -13,6 +13,7 @@ import scipy.stats
 
 from pyextremes.extremes import ExtremesTransformer, get_extremes, get_return_periods
 from pyextremes.models import MLE, Distribution, Emcee, get_model
+from pyextremes.models.model_emcee import PoolType
 from pyextremes.plotting import (
     plot_corner,
     plot_extremes,
@@ -84,7 +85,7 @@ class EVA:
         data = data.copy(deep=True)
 
         # Ensure that `data` has correct index and value dtypes
-        if not np.issubdtype(data.dtype, np.number):
+        if not pd.api.types.is_numeric_dtype(data):
             try:
                 message = "`data` values are not numeric - converting to numeric"
                 logger.debug(message)
@@ -620,7 +621,7 @@ class EVA:
         extremes = extremes.copy(deep=True)
         if not isinstance(extremes.index, pd.DatetimeIndex):
             raise TypeError("invalid index type for `extremes`, must be date-time")
-        if not np.issubdtype(extremes.dtype, np.number):
+        if not pd.api.types.is_numeric_dtype(extremes):
             raise TypeError("`extremes` must have numeric values")
         if extremes.name is None:
             extremes.name = self.data.name
@@ -897,6 +898,7 @@ class EVA:
         n_walkers: int = 100,
         n_samples: int = 500,
         progress: bool = False,
+        pool: typing.Optional[PoolType] = None,
     ) -> None: ...
 
     def fit_model(
@@ -965,6 +967,10 @@ class EVA:
                     suitable for Jupyter notebooks.
                     If False (default), no progress bar will be shown.
                     This progress bar is a part of the `emcee` package.
+                pool : PoolType, optional
+                    A pool object from the multiprocessing module or similar.
+                    If provided, the sampler will use this pool to parallelize
+                    computations.
 
         """
         # Select default distribution

{pyextremes-2.4.0 → pyextremes-2.6.0}/src/pyextremes/models/model_emcee.py

@@ -10,6 +10,17 @@ import scipy.stats
 
 from pyextremes.models.model_base import AbstractModelBaseClass
 
+
+@typing.runtime_checkable
+class PoolType(typing.Protocol):
+    def map(
+        self,
+        func: typing.Callable[..., typing.Any],
+        iterable: typing.Iterable[typing.Any],
+        /,
+    ) -> typing.Iterable[typing.Any]: ...
+
+
 logger = logging.getLogger(__name__)
 
 
@@ -29,6 +40,7 @@ class Emcee(AbstractModelBaseClass):
         n_walkers: int = 100,
         n_samples: int = 500,
         progress: bool = False,
+        pool: typing.Optional[PoolType] = None,
     ) -> None:
         super().__init__(
             extremes=extremes,
@@ -37,6 +49,7 @@ class Emcee(AbstractModelBaseClass):
             n_walkers=n_walkers,
             n_samples=n_samples,
             progress=progress,
+            pool=pool,
         )
         self.n_walkers = n_walkers
         self.n_samples = n_samples
@@ -50,6 +63,7 @@ class Emcee(AbstractModelBaseClass):
         n_walkers: int = kwargs.pop("n_walkers")
         n_samples: int = kwargs.pop("n_samples")
         progress: bool = kwargs.pop("progress")
+        pool: typing.Optional[PoolType] = kwargs.pop("pool", None)
         if len(kwargs) != 0:
             raise TypeError(
                 f"unrecognized arguments passed in: {', '.join(kwargs.keys())}"
@@ -60,6 +74,7 @@ class Emcee(AbstractModelBaseClass):
             nwalkers=n_walkers,
             ndim=self.distribution.number_of_parameters,
             log_prob_fn=self.distribution.log_probability,
+            pool=pool,
         )
 
         # Run the ensemble sampler

{pyextremes-2.4.0 → pyextremes-2.6.0}/src/pyextremes/models/models.py

@@ -3,7 +3,7 @@ from typing import Literal, Optional, Union, overload
 import pandas as pd
 import scipy.stats
 
-from pyextremes.models.model_emcee import Emcee
+from pyextremes.models.model_emcee import Emcee, PoolType
 from pyextremes.models.model_lmoments import Lmoments
 from pyextremes.models.model_mle import MLE
 from pyextremes.models.model_mom import MOM
@@ -46,6 +46,7 @@ def get_model(
     n_walkers: int = 100,
     n_samples: int = 500,
     progress: bool = False,
+    pool: Optional[PoolType] = None,
 ) -> Emcee: ...
 
 
@@ -104,6 +105,10 @@ def get_model(
                 suitable for Jupyter notebooks.
                 If False (default), no progress bar will be shown.
                 This progress bar is a part of the `emcee` package.
+            pool : PoolType, optional
+                A pool object from the multiprocessing module or similar.
+                If provided, the sampler will use this pool to parallelize
+                computations.
 
     Returns
     -------