Merge pull request #14 from tfm000/dev/module_names

Dev/module names
tfm000 · Dec 4, 2023 · 8bf091a · 8bf091a
2 parents ad8e1d9 + 0840a9e
commit 8bf091a
Show file tree

Hide file tree

Showing 70 changed files with 338 additions and 358 deletions.
diff --git a/docs/source/Copulas.rst b/docs/source/Copulas.rst
@@ -203,7 +203,7 @@ Producing a pairplot of the marginals::
 
     mfitter2.pairplot()
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/mfitter_pairplot.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/mfitter_pairplot.png?raw=true
    :alt: MarginalFitter Pair-Plot
    :scale: 60%
    :align: center
@@ -315,17 +315,17 @@ Plotting our fit::
     fitted_copula.mc_cdf_plot(show=False)
     plt.show()
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/PDF_Gh_PDF_Plot_Plot2.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/PDF_Gh_PDF_Plot_Plot2.png?raw=true
    :alt: Generalized Hyperbolic PDF
    :scale: 60%
    :align: center
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/Copula_PDF_Gh_Copula_PDF_Plot_Plot2.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/Copula_PDF_Gh_Copula_PDF_Plot_Plot2.png?raw=true
    :alt: Generalized Hyperbolic Copula PDF
    :scale: 60%
    :align: center
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/MC_CDF_Gh_MC_CDF_Plot_Plot2.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/MC_CDF_Gh_MC_CDF_Plot_Plot2.png?raw=true
    :alt: Generalized Hyperbolic CDF
    :scale: 60%
    :align: center

diff --git a/docs/source/Multivariate.rst b/docs/source/Multivariate.rst
@@ -131,17 +131,17 @@ Plotting our fitted distribution::
     fitted_msh.marginal_pairplot(show=False)
     plt.show()
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/PDF_Mvt_Shyperbolic_PDF_Plot_Plot.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/PDF_Mvt_Shyperbolic_PDF_Plot_Plot.png?raw=true
    :alt: Symmetric Hyperbolic PDF
    :scale: 60%
    :align: center
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/MC_CDF_Mvt_Shyperbolic_MC_CDF_Plot_Plot.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/MC_CDF_Mvt_Shyperbolic_MC_CDF_Plot_Plot.png?raw=true
    :alt: Symmetric Hyperbolic PDF
    :scale: 60%
    :align: center
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/mvt_shyperbolic_marginal_pair_plot.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/mvt_shyperbolic_marginal_pair_plot.png?raw=true
    :alt: Symmetric Hyperbolic PDF
    :scale: 60%
    :align: center

diff --git a/docs/source/Univariate.rst b/docs/source/Univariate.rst
@@ -184,7 +184,7 @@ And plot our fitted distribution::
 
     fitted_gamma.plot()
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/univariate_continuous_example_figure1.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/univariate_continuous_example_figure1.png?raw=true
     :alt: gamma plot
     :align: center
 
@@ -258,7 +258,7 @@ And plot our fitted distribution::
 
     fitted_poisson.plot()
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/univariate_discrete_example_figure1.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/univariate_discrete_example_figure1.png?raw=true
     :alt: poisson plot
     :align: center
 
@@ -347,7 +347,7 @@ finding our best fit::
     Sum of Squared Error                                           9.18869
     #Fitted Data Points                                               1000
 
-.. image:: https://github.com/tfm000/sklarpy/blob/docs/readthedocs/media/univariate_fitter_example_figure1.png?raw=true
+.. image:: https://github.com/tfm000/sklarpy/blob/main/media/univariate_fitter_example_figure1.png?raw=true
     :alt: poisson plot
     :align: center
 

diff --git a/sklarpy/__init__.py b/sklarpy/__init__.py
@@ -1,2 +1,2 @@
-from sklarpy._utils._serialize import load
-from sklarpy._utils._object_printing import print_full
+from sklarpy.utils._serialize import load
+from sklarpy.utils._object_printing import print_full
diff --git a/sklarpy/_plotting/__init__.py b/sklarpy/_plotting/__init__.py
diff --git a/sklarpy/_utils/__init__.py b/sklarpy/_utils/__init__.py
diff --git a/sklarpy/copulas/_distributions/_archimedean.py b/sklarpy/copulas/_distributions/_archimedean.py
@@ -5,7 +5,7 @@
 
 from sklarpy.copulas._prefit_dists import PreFitCopula
 from sklarpy.copulas._fitted_dists import FittedCopula
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 from sklarpy.copulas import MarginalFitter
 
 __all__ = ['clayton_copula_gen', 'gumbel_copula_gen', 'frank_copula_gen']

diff --git a/sklarpy/copulas/_distributions/_gaussian.py b/sklarpy/copulas/_distributions/_gaussian.py
@@ -6,7 +6,7 @@
 
 from sklarpy.copulas._prefit_dists import PreFitCopula
 from sklarpy.copulas._fitted_dists import FittedCopula
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 from sklarpy.copulas import MarginalFitter
 
 __all__ = ['gaussian_copula_gen']

diff --git a/sklarpy/copulas/_distributions/_gaussian_kde.py b/sklarpy/copulas/_distributions/_gaussian_kde.py
@@ -6,7 +6,7 @@
 from sklarpy.copulas._prefit_dists import PreFitCopula
 from sklarpy.copulas._fitted_dists import FittedCopula
 from sklarpy.copulas import MarginalFitter
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['gaussian_kde_copula_gen']
 

diff --git a/sklarpy/copulas/_distributions/_generalized_hyperbolic.py b/sklarpy/copulas/_distributions/_generalized_hyperbolic.py
@@ -5,7 +5,7 @@
 
 from sklarpy.copulas._prefit_dists import PreFitCopula
 from sklarpy.copulas._fitted_dists import FittedCopula
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 from sklarpy.copulas import MarginalFitter
 from sklarpy.univariate import gh
 

diff --git a/sklarpy/copulas/_distributions/_hyperbolics.py b/sklarpy/copulas/_distributions/_hyperbolics.py
@@ -4,7 +4,7 @@
 
 from sklarpy.copulas._distributions._generalized_hyperbolic import \
     gen_hyperbolic_copula_gen
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['marginal_hyperbolic_copula_gen', 'hyperbolic_copula_gen',
            'nig_copula_gen']

diff --git a/sklarpy/copulas/_distributions/_skewed_t.py b/sklarpy/copulas/_distributions/_skewed_t.py
@@ -4,7 +4,7 @@
 
 from sklarpy.copulas._distributions._generalized_hyperbolic import \
     gen_hyperbolic_copula_gen
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 from sklarpy.univariate.distributions import skewed_t
 
 __all__ = ['skewed_t_copula_gen']

diff --git a/sklarpy/copulas/_distributions/_student_t.py b/sklarpy/copulas/_distributions/_student_t.py
@@ -6,7 +6,7 @@
 
 from sklarpy.copulas._prefit_dists import PreFitCopula
 from sklarpy.copulas._fitted_dists import FittedCopula
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 from sklarpy.copulas import MarginalFitter
 
 __all__ = ['student_t_copula_gen']

diff --git a/sklarpy/copulas/_distributions/_symmetric_generalized_hyperbolic.py b/sklarpy/copulas/_distributions/_symmetric_generalized_hyperbolic.py
@@ -4,7 +4,7 @@
 
 from sklarpy.copulas._distributions._generalized_hyperbolic import \
     gen_hyperbolic_copula_gen
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['sym_gen_hyperbolic_copula_gen']
 

diff --git a/sklarpy/copulas/_fitted_dists.py b/sklarpy/copulas/_fitted_dists.py
@@ -3,7 +3,10 @@
 import pandas as pd
 from typing import Union, Iterable
 
-from sklarpy._utils import TypeKeeper, Savable, Copyable, Params
+from sklarpy.utils._type_keeper import TypeKeeper
+from sklarpy.utils._serialize import Savable
+from sklarpy.utils._copy import Copyable
+from sklarpy.utils._params import Params
 
 __all__ = ['FittedCopula']
 

diff --git a/sklarpy/copulas/_prefit_dists.py b/sklarpy/copulas/_prefit_dists.py
@@ -2,21 +2,27 @@
 from typing import Union, Iterable, Callable, Dict, List
 import numpy as np
 import pandas as pd
+from collections import deque
 
 from sklarpy.copulas import MarginalFitter
-from sklarpy._utils import check_multivariate_data, TypeKeeper, Params, \
-    NotImplementedBase, get_mask
+from sklarpy.utils._input_handlers import check_multivariate_data, get_mask
+from sklarpy.utils._type_keeper import TypeKeeper
+from sklarpy.utils._params import Params
+from sklarpy.utils._not_implemented import NotImplementedBase
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate, \
     FittedContinuousMultivariate
 from sklarpy.univariate._fitted_dists import FittedUnivariateBase
-from sklarpy._plotting import pair_plot, threeD_plot
+from sklarpy.plotting._pair_plot import pair_plot
+from sklarpy.plotting._threeD_plot import threeD_plot
 from sklarpy.copulas._fitted_dists import FittedCopula
 
 __all__ = ['PreFitCopula']
 
 
 class PreFitCopula(NotImplementedBase):
     """A pre-fit copula model"""
+    __MAX_RVS_LOOPS: int = 100
+
     def __init__(self, name: str, mv_object: PreFitContinuousMultivariate):
         """A pre-fit copula model.
 
@@ -796,17 +802,31 @@ def copula_rvs(self, size: int, copula_params: Union[Params, tuple],
             distribution. These correspond to randomly sampled cdf /
             pseudo-observation values of the univariate marginals.
         """
-        # generating random variables from multivariate distribution
-        raw_mv_rvs: np.ndarray = self._mv_object.rvs(size, copula_params)
-
-        # bounding these above and below
-        eps: float = 10 ** -5
-        rvs_df: pd.DataFrame = pd.DataFrame(raw_mv_rvs)
-        rvs_df[rvs_df < 0] = eps
-        rvs_df[rvs_df > 1] = 1 - eps
-        mv_rvs: np.ndarray = rvs_df.to_numpy()
-
-        return self._g_to_u(mv_rvs, copula_params)
+        num_loops: int = 0
+        d: int = self._mv_object._get_dim(
+            self._mv_object._get_params(copula_params))
+        valid_copula_rvs: deque = deque()
+        while size > 0:
+            # generating random variables from multivariate distribution
+            mv_rvs: np.ndarray = self._mv_object.rvs(size, copula_params)
+
+            # converting to copula rvs
+            raw_copula_rvs: np.ndarray = self._g_to_u(mv_rvs, copula_params)
+
+            # filtering out invalid copula rvs (not in [0, 1]^d)
+            mask: np.ndarray = ((raw_copula_rvs > 0) & (raw_copula_rvs < 1)
+                                ).sum(axis=1) == d
+            copula_rvs = raw_copula_rvs[mask]
+            valid_copula_rvs.append(copula_rvs)
+
+            # repeating until sample size reached
+            size -= copula_rvs.shape[0]
+            num_loops += 1
+            if num_loops > self.__MAX_RVS_LOOPS:
+                raise ArithmeticError(f"Unable to generate valid copula rvs. "
+                                      f"Max number of retries reached: "
+                                      f"{self.__MAX_RVS_LOOPS}")
+        return np.concatenate(valid_copula_rvs, axis=0)
 
     def _get_components_summary(self,
                                 fitted_mv_object: FittedContinuousMultivariate,
@@ -1008,11 +1028,20 @@ def fit(self, data: Union[pd.DataFrame, np.ndarray, None] = None,
                              "params do not match.")
 
         # generating data to use when calculating statistics
-        data_array: np.ndarray = self.rvs(
-            size=10**3, copula_params=fitted_mv_object.params,
-            mdists=mdists_dict, ppf_approx=True
-        ) if data is None \
-            else check_multivariate_data(data, allow_1d=True, allow_nans=True)
+        try:
+            data_array: np.ndarray = self.rvs(
+                size=10**3, copula_params=fitted_mv_object.params,
+                mdists=mdists_dict, ppf_approx=True) if data is None \
+                else check_multivariate_data(
+                data, allow_1d=True, allow_nans=True)
+
+        except ArithmeticError as e:
+            if str(e) != (f"Unable to generate valid copula rvs. Max number "
+                          f"of retries reached: {self.__MAX_RVS_LOOPS}"):
+                raise
+            else:
+                data_array: np.ndarray = np.full((10**3, d), np.nan,
+                                                 dtype=float)
 
         # fitting TypeKeeper object
         type_keeper: TypeKeeper = TypeKeeper(data_array)

diff --git a/sklarpy/copulas/marginal_fitter.py b/sklarpy/copulas/marginal_fitter.py
@@ -4,10 +4,13 @@
 import pandas as pd
 from typing import Union, Iterable
 
-from sklarpy._utils import FitError, TypeKeeper, check_multivariate_data, \
-    get_iterator, Savable
+from sklarpy.utils._errors import FitError
+from sklarpy.utils._type_keeper import TypeKeeper
+from sklarpy.utils._input_handlers import check_multivariate_data
+from sklarpy.utils._iterator import get_iterator
+from sklarpy.utils._serialize import Savable
 from sklarpy.univariate import UnivariateFitter
-from sklarpy._plotting import pair_plot
+from sklarpy.plotting._pair_plot import pair_plot
 
 __all__ = ['MarginalFitter']
 

diff --git a/sklarpy/misc/correlation.py b/sklarpy/misc/correlation.py
@@ -4,7 +4,7 @@
 import warnings
 from typing import Tuple, Union
 
-from sklarpy._utils import check_multivariate_data
+from sklarpy.utils._input_handlers import check_multivariate_data
 
 __all__ = ['CorrelationMatrix']
 

diff --git a/sklarpy/misc/gradient.py b/sklarpy/misc/gradient.py
@@ -4,7 +4,7 @@
 from typing import Callable, Union, Iterable
 from collections import deque
 
-from sklarpy._utils import check_univariate_data
+from sklarpy.utils._input_handlers import check_univariate_data
 
 __all__ = ['gradient_1d']
 

diff --git a/sklarpy/multivariate/_distributions/_archimedean.py b/sklarpy/multivariate/_distributions/_archimedean.py
@@ -10,7 +10,9 @@
 
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate
 from sklarpy.multivariate._fitted_dists import FittedContinuousMultivariate
-from sklarpy._utils import get_iterator, FitError, Params
+from sklarpy.utils._params import Params
+from sklarpy.utils._iterator import get_iterator
+from sklarpy.utils._errors import FitError
 from sklarpy.misc import debye
 
 __all__ = ['multivariate_clayton_gen', 'multivariate_gumbel_gen',

diff --git a/sklarpy/multivariate/_distributions/_gaussian_kde.py b/sklarpy/multivariate/_distributions/_gaussian_kde.py
@@ -6,7 +6,7 @@
 
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate
 from sklarpy.multivariate._fitted_dists import FittedContinuousMultivariate
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['multivariate_gaussian_kde_gen']
 

diff --git a/sklarpy/multivariate/_distributions/_generalized_hyperbolic.py b/sklarpy/multivariate/_distributions/_generalized_hyperbolic.py
@@ -10,7 +10,7 @@
 from sklarpy.univariate import gig
 from sklarpy.univariate._distributions import _gh
 from sklarpy.misc import CorrelationMatrix, kv
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['multivariate_gen_hyperbolic_gen']
 

diff --git a/sklarpy/multivariate/_distributions/_hyperbolics.py b/sklarpy/multivariate/_distributions/_hyperbolics.py
@@ -7,7 +7,7 @@
 from sklarpy.multivariate._distributions._generalized_hyperbolic import \
     multivariate_gen_hyperbolic_gen
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['multivariate_marginal_hyperbolic_gen',
            'multivariate_hyperbolic_gen', 'multivariate_nig_gen']

diff --git a/sklarpy/multivariate/_distributions/_normal.py b/sklarpy/multivariate/_distributions/_normal.py
@@ -7,7 +7,7 @@
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate
 from sklarpy.misc import CorrelationMatrix
 from sklarpy.multivariate._fitted_dists import FittedContinuousMultivariate
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['multivariate_normal_gen']
 

diff --git a/sklarpy/multivariate/_distributions/_skewed_t.py b/sklarpy/multivariate/_distributions/_skewed_t.py
@@ -9,7 +9,7 @@
     multivariate_gen_hyperbolic_gen
 from sklarpy.multivariate._distributions._student_t import \
     multivariate_student_t_gen
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 from sklarpy.misc import kv
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate
 from sklarpy.univariate import ig

diff --git a/sklarpy/multivariate/_distributions/_student_t.py b/sklarpy/multivariate/_distributions/_student_t.py
@@ -8,7 +8,7 @@
 
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate
 from sklarpy.multivariate._fitted_dists import FittedContinuousMultivariate
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 from sklarpy.misc import CorrelationMatrix
 
 __all__ = ['multivariate_student_t_gen']

diff --git a/sklarpy/multivariate/_distributions/_symmetric_generalized_hyperbolic.py b/sklarpy/multivariate/_distributions/_symmetric_generalized_hyperbolic.py
@@ -5,7 +5,7 @@
 from sklarpy.multivariate._prefit_dists import PreFitContinuousMultivariate
 from sklarpy.multivariate._distributions._generalized_hyperbolic import \
     multivariate_gen_hyperbolic_gen
-from sklarpy._utils import Params
+from sklarpy.utils._params import Params
 
 __all__ = ['multivariate_sym_gen_hyperbolic_gen']