Merge pull request pydata#270 from davidastephens/EDGAR

TST: Skip EDGAR tests until rewrite

Author: davidastephens

docs/source/remote_data.rst

@@ -471,6 +471,9 @@ You can specify dataset ID "tran_sf_railac" to get corresponding data via ``Data
 EDGAR Index
 ===========
 
+** As of December 31st, the SEC disabled access via FTP. EDGAR support
+currently broken until re-write to use HTTPS. **
+
 Company filing index from EDGAR (SEC).
 
 The daily indices get large quickly (i.e. the set of daily indices from 1994
@@ -480,17 +483,17 @@ If the FTP server starts refusing your connections, you should be able to
 reconnect after waiting a few minutes.
 
 
-.. ipython:: python
+.. .. ipython:: python
 
-    import pandas_datareader.data as web
-    ed = web.DataReader('full', 'edgar-index')
-    ed[:5]
+..     import pandas_datareader.data as web
+..     ed = web.DataReader('full', 'edgar-index')
+..     ed[:5]
 
-.. ipython:: python
+.. .. ipython:: python
 
-    import pandas_datareader.data as web
-    ed = web.DataReader('daily', 'edgar-index', '1998-05-18', '1998-05-18')
-    ed[:5]
+..     import pandas_datareader.data as web
+..     ed = web.DataReader('daily', 'edgar-index', '1998-05-18', '1998-05-18')
+..     ed[:5]
 
 .. _remote_data.tsp:
 

pandas_datareader/google/options.py

@@ -137,6 +137,11 @@ def _process_data(self, jd, expiry):
         indexes = ['Strike', 'Expiry', 'Type', 'Symbol']
         rows_list, index = self._process_rows(jd, now, expiry)
         df = DataFrame(rows_list, columns=columns, index=MultiIndex.from_tuples(index, names=indexes))
+
+        # Make dtype consistent, requires float64 as there can be NaNs
+        df['Vol'] = df['Vol'].astype('float64')
+        df['Open_Int'] = df['Open_Int'].astype('float64')
+
         return df.sort_index()
 
     def _process_rows(self, jd, now, expiry):

pandas_datareader/tests/test_edgar.py

@@ -10,6 +10,8 @@
 class TestEdgarIndex(tm.TestCase):
 
     def test_get_full_index(self):
+        # As of December 31, SEC has disabled ftp access to EDGAR.  Disabling tests until re-write
+        raise nose.SkipTest()
         try:
             ed = web.DataReader('full', 'edgar-index')
             assert len(ed) > 1000
@@ -22,6 +24,8 @@ def test_get_full_index(self):
             raise nose.SkipTest(e)
 
     def test_get_nonzip_index_and_low_date(self):
+        # As of December 31, SEC has disabled ftp access to EDGAR.  Disabling tests until re-write
+        raise nose.SkipTest()
         try:
             ed = web.DataReader('daily', 'edgar-index', '1994-06-30',
                                 '1994-07-02')
@@ -48,6 +52,8 @@ def test_get_gz_index_and_no_date(self):
             raise nose.SkipTest(e)
 
     def test_6_digit_date(self):
+        # As of December 31, SEC has disabled ftp access to EDGAR.  Disabling tests until re-write
+        raise nose.SkipTest()
         try:
             ed = web.DataReader('daily', 'edgar-index', start='1998-05-18',
                                 end='1998-05-18')

pandas_datareader/tests/test_google_options.py

@@ -31,7 +31,7 @@ def assert_option_result(self, df):
         tm.assert_index_equal(df.columns, exp_columns)
         tm.assert_equal(df.index.names, [u'Strike', u'Expiry', u'Type', u'Symbol'])
 
-        dtypes = ['float64'] * 6 + ['int64', 'object', 'float64', 'datetime64[ns]']
+        dtypes = ['float64'] * 7 + ['object', 'float64', 'datetime64[ns]']
         dtypes = [np.dtype(x) for x in dtypes]
         tm.assert_series_equal(df.dtypes, pd.Series(dtypes, index=exp_columns))
 
@@ -54,7 +54,7 @@ def test_expiry_dates(self):
         except RemoteDataError as e:  # pragma: no cover
             raise nose.SkipTest(e)
 
-        self.assertTrue(len(dates) >= 5)
+        self.assertTrue(len(dates) >= 4)
         self.assertIsInstance(dates, list)
         self.assertTrue(all(isinstance(dt, date) for dt in dates))
 

pandas_datareader/tests/test_yahoo_options.py

@@ -50,8 +50,8 @@ def assert_option_result(self, df):
         tm.assert_index_equal(df.columns, exp_columns)
         tm.assert_equal(df.index.names, [u'Strike', u'Expiry', u'Type', u'Symbol'])
 
-        dtypes = [np.dtype(x) for x in ['float64'] * 5 +
-                  ['int64', 'int64', 'float64', 'object', 'bool', 'object', 'float64', 'datetime64[ns]',
+        dtypes = [np.dtype(x) for x in ['float64'] * 7 +
+                  ['float64', 'object', 'bool', 'object', 'float64', 'datetime64[ns]',
                    'datetime64[ns]', 'object']]
         tm.assert_series_equal(df.dtypes, pd.Series(dtypes, index=exp_columns))
 

pandas_datareader/yahoo/options.py

@@ -684,6 +684,10 @@ def _process_data(self, jd):
 
         df['IsNonstandard'] = df['Root'] != self.symbol.replace('-', '')
 
+        # Make dtype consistent, requires float64 as there can be NaNs
+        df['Vol'] = df['Vol'].astype('float64')
+        df['Open_Int'] = df['Open_Int'].astype('float64')
+
         return df.sort_index()
 
     def _process_rows(self, jd):