Added a walkaround for freezing parallel processing. Now when one spe…

…cifies n_jobs=1 it will not do anything in parallel and will not span new processes

requirements_stable.txt

@@ -1,46 +1,51 @@
-atomicwrites==1.3.0
-attrs==19.1.0
+appnope==0.1.0
+backcall==0.1.0
 bleach==3.1.0
-certifi==2019.6.16
-cffi==1.12.3
+certifi==2019.11.28
 chardet==3.0.4
 Click==7.0
-Cython==0.29.13
+Cython==0.29.14
+decorator==4.4.1
 docutils==0.15.2
 idna==2.8
-importlib-metadata==0.19
-Jinja2==2.10.1
-joblib==0.13.2
-MarkupSafe==1.1.1
-more-itertools==7.2.0
-numpy==1.17.0
-packaging==19.1
-pandas==0.25.0
+importlib-metadata==1.3.0
+ipykernel==5.1.3
+ipython==7.9.0
+ipython-genutils==0.2.0
+jedi==0.15.1
+joblib==0.14.1
+jupyter-client==5.3.4
+jupyter-core==4.6.1
+keyring==21.0.0
+more-itertools==8.0.2
+numpy==1.18.1
+pandas==0.25.3
+parso==0.5.1
 patsy==0.5.1
-pip-tools==4.0.0
+pexpect==4.7.0
+pickleshare==0.7.5
+pip-tools==4.3.0
 pkginfo==1.5.0.1
-pluggy==0.12.0
-pmdarima==1.2.1
-py==1.8.0
-pycparser==2.19
+pmdarima==1.5.2
+prompt-toolkit==2.0.10
+ptyprocess==0.6.0
 Pygments==2.4.2
-pyparsing==2.4.2
-pytest==5.0.1
-python-dateutil==2.8.0
-pytz==2019.2
+python-dateutil==2.8.1
+pytz==2019.3
+pyzmq==18.1.0
 readme-renderer==24.0
 requests==2.22.0
 requests-toolbelt==0.9.1
-rpy2==3.0.5
-scikit-learn==0.21.3
-scipy==1.3.0
-simplegeneric==0.8.1
+scikit-learn==0.22.1
+scipy==1.4.1
 six==1.12.0
-statsmodels==0.10.1
--e git+git@github.com:intive-DataScience/tbats.git@4e726919f08e39e74dd70a592b5258dfc7b25953#egg=tbats
-tqdm==4.32.2
-twine==1.13.0
-urllib3==1.25.3
+statsmodels==0.10.2
+-e git+git@github.com:intive-DataScience/tbats.git@75bc957e425c3b82dfe75373e2cea7aff17ecc12#egg=tbats
+tornado==6.0.3
+tqdm==4.41.1
+traitlets==4.3.3
+twine==3.1.1
+urllib3==1.25.7
 wcwidth==0.1.7
 webencodings==0.5.1
-zipp==0.5.2
+zipp==0.6.0

setup.py

@@ -43,7 +43,7 @@ def run_tests(self):
 
 setuptools.setup(
     name='tbats',
-    version='1.0.8',
+    version='1.0.9',
     packages=setuptools.find_packages(exclude=('test', 'test_R')),
     url='https://github.com/intive-DataScience/tbats',
     license='MIT License',

tbats/abstract/Estimator.py

@@ -1,5 +1,6 @@
 import numpy as np
-import multiprocessing
+import multiprocessing as actual_processing
+import multiprocessing.dummy as dummy_processing
 from sklearn.base import BaseEstimator
 from sklearn.utils.validation import check_array, column_or_1d as c1d
 from sklearn.model_selection import ParameterGrid
@@ -139,7 +140,7 @@ def _choose_model_from_possible_component_settings(self, y, components_grid):
         """
         self._y = y
         # note n_jobs = None means to use cpu_count()
-        pool = multiprocessing.pool.Pool(processes=self.n_jobs)
+        pool = self._prepare_pool(self.n_jobs)
         models = pool.map(self._case_fit, components_grid)
         pool.close()
         self._y = None  # clean-up
@@ -151,6 +152,11 @@ def _choose_model_from_possible_component_settings(self, y, components_grid):
                 best_model = model
         return best_model
 
+    def _prepare_pool(self, n_jobs=None):
+        if n_jobs == 1:
+            return dummy_processing.Pool(processes=n_jobs)
+        return actual_processing.Pool(processes=n_jobs)
+
     def _prepare_components_grid(self, seasonal_harmonics=None):
         """Provides a grid of all allowed model component combinations.
 

tbats/tbats/HarmonicsChoosingStrategy.py

@@ -1,14 +1,15 @@
 import numpy as np
 import fractions
-import multiprocessing
+import multiprocessing as actual_processing
+import multiprocessing.dummy as dummy_processing
 
 
 class HarmonicsChoosingStrategy(object):
 
     def __init__(self, context, n_jobs=None):
         self.n_jobs = n_jobs
         if n_jobs is None:
-            self.n_jobs = multiprocessing.cpu_count()
+            self.n_jobs = actual_processing.cpu_count()
         self.context = context
 
     def choose(self, y, components):
@@ -78,7 +79,7 @@ def choose_for_season(self, season_index, max_harmonic, best_model_so_far):
             self._season_index = season_index
             self._y = best_model_so_far.y
             self._components = best_model_so_far.params.components
-            pool = multiprocessing.pool.Pool(processes=self.n_jobs)
+            pool = self._prepare_pool(self.n_jobs)
             models = pool.map(self._fit_model, harmonics_range)
             pool.close()
             for model in models:
@@ -102,6 +103,11 @@ def choose_for_season(self, season_index, max_harmonic, best_model_so_far):
             return best_model_so_far
         return best_model
 
+    def _prepare_pool(self, n_jobs=None):
+        if n_jobs == 1:
+            return dummy_processing.Pool(processes=n_jobs)
+        return actual_processing.Pool(processes=n_jobs)
+
     def _fit_model(self, harmonic_to_check):
         components = self._components.with_harmonic_for_season(
             season_index=self._season_index, new_harmonic=harmonic_to_check