Merge pull request #197 from UDST/sort_bug

Update Pandas sort methods

HISTORY.rst

@@ -1,3 +1,10 @@
+v3.1.1
+======
+
+2017/5/9
+
+* Updated deprecated `sort` method for Pandas Series and DataFrames
+
 v3.1.0
 ======
 

docs/conf.py

@@ -59,7 +59,7 @@
 # The short X.Y version.
 version = '3.1'
 # The full version, including alpha/beta/rc tags.
-release = '3.1.0'
+release = '3.1.1'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.

docs/gettingstarted.rst

@@ -34,7 +34,7 @@ UrbanSim depends on the following libraries, most of which are in Anaconda:
 * `matplotlib <http://matplotlib.org>`__ >= 1.3.1
 * `numpy <http://numpy.org>`__ >= 1.8.0
 * `orca <https://github.com/UDST/orca>`__ >= 1.1
-* `pandas <http://pandas.pydata.org>`__ >= 0.15.
+* `pandas <http://pandas.pydata.org>`__ >= 0.17.0
 * `patsy <http://patsy.readthedocs.org/en/latest/>`__ >= 0.3.0
 * `prettytable <https://code.google.com/p/prettytable/>`__ >= 0.7.2
 * `pyyaml <http://pyyaml.org/>`__ >= 3.10

setup.py

@@ -14,7 +14,7 @@
 
 setup(
     name='urbansim',
-    version='3.1.0',
+    version='3.1.1',
     description='Tool for modeling metropolitan real estate markets',
     long_description=long_description,
     author='UrbanSim Inc.',
@@ -37,7 +37,7 @@
         'matplotlib>=1.3.1',
         'numpy>=1.8.0',
         'orca>=1.1',
-        'pandas>=0.15.0',
+        'pandas>=0.17.0',
         'patsy>=0.3.0',
         'prettytable>=0.7.2',
         'pyyaml>=3.10',

urbansim/__init__.py

@@ -1 +1 @@
-__version__ = version = '3.1.0'
+__version__ = version = '3.1.1'

urbansim/models/tests/test_util.py

@@ -152,7 +152,7 @@ def test_sorted_groupby():
     df = pd.DataFrame(
         {'alpha': np.random.choice(list(letters), 100),
          'num': np.random.randint(100)})
-    sorted_df = df.sort('alpha')
+    sorted_df = df.sort_values('alpha')
 
     expected = {name: d.to_dict() for name, d in df.groupby('alpha')}
     test = {name: d.to_dict()

urbansim/utils/sampling.py

@@ -142,7 +142,7 @@ def accounting_sample_no_replace(total, data, accounting_column, prob_column=Non
     else:
         # weighted shuffle
         ran_p = pd.Series(np.power(np.random.rand(len(p)), 1.0 / p), index=data.index)
-        ran_p.sort(ascending=False)
+        ran_p.sort_values(ascending=False)
         shuff_idx = ran_p.index.values
 
     # get the initial sample