Merge pull request #1424 from lanzagar/timevar

[ENH] SqlTable: Automatically recognize date/time fields

Orange/data/sql/table.py

@@ -16,7 +16,7 @@
 psycopg2.pool = Orange.misc.import_late_warning("psycopg2.pool")
 
 from .. import domain, variable, table, instance, filter,\
-    DiscreteVariable, ContinuousVariable, StringVariable
+    DiscreteVariable, ContinuousVariable, StringVariable, TimeVariable
 from Orange.data.sql import filter as sql_filter
 
 
@@ -104,8 +104,12 @@ def get_domain(self, type_hints=None, guess_values=False):
 
         def add_to_sql(var, field_name):
             if var.is_continuous:
-                var.to_sql = ToSql("({})::double precision".format(
-                    self.quote_identifier(field_name)))
+                if isinstance(var, TimeVariable):
+                    var.to_sql = ToSql("extract(epoch from {})".format(
+                        self.quote_identifier(field_name)))
+                else:
+                    var.to_sql = ToSql("({})::double precision".format(
+                        self.quote_identifier(field_name)))
             elif var.is_discrete:
                 var.to_sql = ToSql("({})::text".format(
                     self.quote_identifier(field_name)))
@@ -137,10 +141,19 @@ def get_variable(self, field_name, type_code, inspect_values=False):
         INT_TYPES = (20, 21, 23)  # bigint, int, smallint
         CHAR_TYPES = (25, 1042, 1043,)  # text, char, varchar
         BOOLEAN_TYPES = (16,)  # bool
+        DATE_TYPES = (1082, 1114, 1184, )  # date, timestamp, timestamptz
+        # time, timestamp, timestamptz, timetz
+        TIME_TYPES = (1083, 1114, 1184, 1266,)
 
         if type_code in FLOATISH_TYPES:
             return ContinuousVariable(field_name)
 
+        if type_code in TIME_TYPES + DATE_TYPES:
+            tv = TimeVariable(field_name)
+            tv.have_date |= type_code in DATE_TYPES
+            tv.have_time |= type_code in TIME_TYPES
+            return tv
+
         if type_code in INT_TYPES:  # bigint, int, smallint
             if inspect_values:
                 values = self.get_distinct_values(field_name)

Orange/data/variable.py

@@ -986,3 +986,12 @@ def parse(self, datestr):
         try: return dt.timestamp()
         except OverflowError:
             return -(self.UNIX_EPOCH - dt).total_seconds()
+
+    def to_val(self, s):
+        """
+        Convert a value, given as an instance of an arbitrary type, to a float.
+        """
+        if isinstance(s, str):
+            return self.parse(s)
+        else:
+            return super().to_val(s)

Orange/tests/sql/test_sql_table.py

@@ -2,12 +2,13 @@
 # pylint: disable=missing-docstring
 
 import unittest
+import unittest.mock
 
 import numpy as np
 from numpy.testing import assert_almost_equal
 
 from Orange.data import filter, ContinuousVariable, DiscreteVariable, \
-    StringVariable, Table, Domain
+    StringVariable, TimeVariable, Table, Domain
 from Orange.data.sql.table import SqlTable
 from Orange.preprocess.discretize import EqualWidth
 from Orange.statistics.basic_stats import BasicStats, DomainBasicStats
@@ -367,16 +368,66 @@ def test_meta_varchar(self):
         sql_table = SqlTable(conn, table_name, inspect_values=True)
         self.assertFirstMetaIsInstance(sql_table, StringVariable)
 
-    def test_date(self):
+    def test_time_date(self):
         table = np.array(['2014-04-12', '2014-04-13', '2014-04-14',
                           '2014-04-15', '2014-04-16']).reshape(-1, 1)
         conn, table_name = self.create_sql_table(table, ['date'])
 
         sql_table = SqlTable(conn, table_name, inspect_values=False)
-        self.assertFirstMetaIsInstance(sql_table, StringVariable)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
 
         sql_table = SqlTable(conn, table_name, inspect_values=True)
-        self.assertFirstMetaIsInstance(sql_table, StringVariable)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+    def test_time_time(self):
+        table = np.array(['17:39:51', '11:51:48.46', '05:20:21.492149',
+                          '21:47:06', '04:47:35.8']).reshape(-1, 1)
+        conn, table_name = self.create_sql_table(table, ['time'])
+
+        sql_table = SqlTable(conn, table_name, inspect_values=False)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+        sql_table = SqlTable(conn, table_name, inspect_values=True)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+    def test_time_timetz(self):
+        table = np.array(['17:39:51+0200', '11:51:48.46+01', '05:20:21.4921',
+                          '21:47:06-0600', '04:47:35.8+0330']).reshape(-1, 1)
+        conn, table_name = self.create_sql_table(table, ['timetz'])
+
+        sql_table = SqlTable(conn, table_name, inspect_values=False)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+        sql_table = SqlTable(conn, table_name, inspect_values=True)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+    def test_time_timestamp(self):
+        table = np.array(['2014-07-15 17:39:51.348149',
+                          '2008-10-05 11:51:48.468149',
+                          '2008-11-03 05:20:21.492149',
+                          '2015-01-02 21:47:06.228149',
+                          '2016-04-16 04:47:35.892149']).reshape(-1, 1)
+        conn, table_name = self.create_sql_table(table, ['timestamp'])
+
+        sql_table = SqlTable(conn, table_name, inspect_values=False)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+        sql_table = SqlTable(conn, table_name, inspect_values=True)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+    def test_time_timestamptz(self):
+        table = np.array(['2014-07-15 17:39:51.348149+0200',
+                          '2008-10-05 11:51:48.468149+02',
+                          '2008-11-03 05:20:21.492149+01',
+                          '2015-01-02 21:47:06.228149+0100',
+                          '2016-04-16 04:47:35.892149+0330']).reshape(-1, 1)
+        conn, table_name = self.create_sql_table(table, ['timestamptz'])
+
+        sql_table = SqlTable(conn, table_name, inspect_values=False)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
+
+        sql_table = SqlTable(conn, table_name, inspect_values=True)
+        self.assertFirstAttrIsInstance(sql_table, TimeVariable)
 
     def test_double_precision(self):
         table = np.arange(25).reshape((-1, 1))

Orange/tests/test_variable.py

@@ -312,12 +312,14 @@ class TestTimeVariable(VariableTest):
         ('1969-12-31 23:59:59', -1, '1969-12-31 23:59:59'),
         ('1900-01-01', -2208988800, '1900-01-01'),
         ('nan', np.nan, '?'),
+        ('1444651991.81', 1444651991.81, '2015-10-12 12:13:11.810000'),
+        (1444651991.81, 1444651991.81, '2015-10-12 12:13:11.810000'),
     ]
 
     def test_parse_repr(self):
         for datestr, timestamp, outstr in self.TESTS:
             var = TimeVariable('time')
-            ts = var.parse(datestr)
+            ts = var.to_val(datestr)  # calls parse for strings
             if not np.isnan(ts):
                 self.assertEqual(ts, timestamp, msg=datestr)
             self.assertEqual(var.repr_val(ts), outstr, msg=datestr)