def test_merge_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='foo')
table = tableset.merge()
self.assertColumnNames(table, ['foo', 'letter', 'number'])
self.assertColumnTypes(table, [Text, Text, Number])
def test_from_json_file(self):
tableset1 = TableSet(self.tables.values(), self.tables.keys())
tableset2 = TableSet.from_json("examples/test_tableset.json")
with open("examples/test_tableset.json") as f:
filelike = StringIO(f.read())
tableset3 = TableSet.from_json(filelike)
self.assertSequenceEqual(tableset1.column_names, tableset2.column_names, tableset3.column_names)
self.assertSequenceEqual(
[type(t) for t in tableset1.column_types],
[type(t) for t in tableset2.column_types],
[type(t) for t in tableset3.column_types],
)
self.assertEqual(len(tableset1), len(tableset2), len(tableset3))
for name in ["table1", "table2", "table3"]:
self.assertEqual(len(tableset1[name].columns), len(tableset2[name].columns), len(tableset3[name].columns))
self.assertEqual(len(tableset1[name].rows), len(tableset2[name].rows), len(tableset3[name].rows))
self.assertSequenceEqual(tableset1[name].rows[0], tableset2[name].rows[0], tableset3[name].rows[0])
self.assertSequenceEqual(tableset1[name].rows[1], tableset2[name].rows[1], tableset3[name].rows[1])
self.assertSequenceEqual(tableset1[name].rows[2], tableset2[name].rows[2], tableset3[name].rows[2])
def test_merge_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name="foo")
table = tableset.merge()
self.assertColumnNames(table, ["foo", "letter", "number"])
self.assertColumnTypes(table, [Text, Text, Number])
def test_from_json_file(self):
tableset1 = TableSet(self.tables.values(), self.tables.keys())
tableset2 = TableSet.from_json('examples/test_tableset.json')
with open('examples/test_tableset.json') as f:
filelike = StringIO(f.read())
tableset3 = TableSet.from_json(filelike)
self.assertSequenceEqual(tableset1.column_names,
tableset2.column_names,
tableset3.column_names)
self.assertSequenceEqual([type(t) for t in tableset1.column_types],
[type(t) for t in tableset2.column_types],
[type(t) for t in tableset3.column_types])
self.assertEqual(len(tableset1), len(tableset2), len(tableset3))
for name in ['table1', 'table2', 'table3']:
self.assertEqual(len(tableset1[name].columns),
len(tableset2[name].columns),
len(tableset3[name].columns))
self.assertEqual(len(tableset1[name].rows),
len(tableset2[name].rows),
len(tableset3[name].rows))
self.assertSequenceEqual(tableset1[name].rows[0],
tableset2[name].rows[0],
tableset3[name].rows[0])
self.assertSequenceEqual(tableset1[name].rows[1],
tableset2[name].rows[1],
tableset3[name].rows[1])
self.assertSequenceEqual(tableset1[name].rows[2],
tableset2[name].rows[2],
tableset3[name].rows[2])
def test_nested_aggregation(self):
tableset = TableSet(self.tables.values(),
self.tables.keys(),
key_name='test')
nested = tableset.group_by('letter')
results = nested.aggregate([('letter', Length(), 'count'),
('number', Sum(), 'number_sum')])
self.assertIsInstance(results, Table)
self.assertEqual(len(results.rows), 7)
self.assertEqual(len(results.columns), 4)
self.assertSequenceEqual(results._column_names,
('test', 'letter', 'count', 'number_sum'))
self.assertSequenceEqual(results.rows[0], ('table1', 'a', 2, 4))
self.assertSequenceEqual(results.rows[1], ('table1', 'b', 1, 2))
self.assertSequenceEqual(results.rows[2], ('table2', 'b', 1, 0))
self.assertSequenceEqual(results.rows[3], ('table2', 'a', 1, 2))
self.assertSequenceEqual(results.rows[4], ('table2', 'c', 1, 5))
self.assertSequenceEqual(results.rows[5], ('table3', 'a', 2, 3))
self.assertSequenceEqual(results.rows[6], ('table3', 'c', 1, 3))
def test_proxy_table_invalid(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
with self.assertRaises(AttributeError) as cm:
tableset.print_table()
self.assertEquals(str(cm.exception), 'Table method "print_table" cannot be used as a TableSet method.')
def setUp(self):
self.table1 = (('a', 1, 4), ('b', 3, 7), ('c', 2, 2))
self.table2 = (('a', 0, 3), ('b', 2, 3), ('c', 5, 3))
self.table3 = (('a', 1, 10), ('b', 2, 1), ('c', 3, None))
self.text_type = Text()
self.number_type = Number()
self.column_names = ['one', 'two', 'three']
self.column_types = [
self.text_type, self.number_type, self.number_type
]
self.tables = OrderedDict([
('table1', Table(self.table1, self.column_names,
self.column_types)),
('table2', Table(self.table2, self.column_names,
self.column_types)),
('table3', Table(self.table3, self.column_names,
self.column_types))
])
self.tablesets = TableSet(self.tables.values(), self.tables.keys())
def test_compute(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.compute([
('new_column', Formula(self.text_type, lambda r: '%(letter)s-%(number)i' % r))
])
new_table = new_tableset['table1']
self.assertColumnNames(new_table, ('letter', 'number', 'new_column',))
self.assertColumnTypes(new_table, (Text, Number, Text))
self.assertRows(new_table, [
('a', 1, 'a-1'),
('a', 3, 'a-3'),
('b', 2, 'b-2')
])
new_table = new_tableset['table2']
self.assertRows(new_table, [
('b', 0, 'b-0'),
('a', 2, 'a-2'),
('c', 5, 'c-5')
])
new_table = new_tableset['table3']
self.assertSequenceEqual(new_table.rows[0], ('a', 1, 'a-1'))
self.assertSequenceEqual(new_table.rows[1], ('a', 2, 'a-2'))
self.assertSequenceEqual(new_table.rows[2], ('c', 3, 'c-3'))
def test_compute(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.compute([
(Formula(self.text_type, lambda r: '%(letter)s-%(number)i' % r), 'new_column')
])
new_table = new_tableset['table1']
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 3)
self.assertSequenceEqual(new_table._column_types, (self.text_type, self.number_type, self.text_type,))
self.assertSequenceEqual(new_table._column_names, ('letter', 'number', 'new_column',))
self.assertSequenceEqual(new_table.rows[0], ('a', 1, 'a-1'))
self.assertSequenceEqual(new_table.rows[1], ('a', 3, 'a-3'))
self.assertSequenceEqual(new_table.rows[2], ('b', 2, 'b-2'))
new_table = new_tableset['table2']
self.assertSequenceEqual(new_table.rows[0], ('b', 0, 'b-0'))
self.assertSequenceEqual(new_table.rows[1], ('a', 2, 'a-2'))
self.assertSequenceEqual(new_table.rows[2], ('c', 5, 'c-5'))
new_table = new_tableset['table3']
self.assertSequenceEqual(new_table.rows[0], ('a', 1, 'a-1'))
self.assertSequenceEqual(new_table.rows[1], ('a', 2, 'a-2'))
self.assertSequenceEqual(new_table.rows[2], ('c', 3, 'c-3'))
def test_proxy_table_invalid(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
with self.assertRaises(AttributeError) as cm:
tableset.print_table()
self.assertEquals(str(cm.exception), 'Table method "print_table" cannot be used as a TableSet method.')
def test_compute(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.compute([
('new_column', Formula(self.text_type, lambda r: '%(letter)s-%(number)i' % r))
])
new_table = new_tableset['table1']
self.assertColumnNames(new_table, ('letter', 'number', 'new_column',))
self.assertColumnTypes(new_table, (Text, Number, Text))
self.assertRows(new_table, [
('a', 1, 'a-1'),
('a', 3, 'a-3'),
('b', 2, 'b-2')
])
new_table = new_tableset['table2']
self.assertRows(new_table, [
('b', 0, 'b-0'),
('a', 2, 'a-2'),
('c', 5, 'c-5')
])
new_table = new_tableset['table3']
self.assertSequenceEqual(new_table.rows[0], ('a', 1, 'a-1'))
self.assertSequenceEqual(new_table.rows[1], ('a', 2, 'a-2'))
self.assertSequenceEqual(new_table.rows[2], ('c', 3, 'c-3'))
def test_compute(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.compute([
(Formula(self.text_type, lambda r: '%(letter)s-%(number)i' % r), 'new_column')
])
new_table = new_tableset['table1']
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 3)
self.assertSequenceEqual(new_table._column_types, (self.text_type, self.number_type, self.text_type,))
self.assertSequenceEqual(new_table._column_names, ('letter', 'number', 'new_column',))
self.assertSequenceEqual(new_table.rows[0], ('a', 1, 'a-1'))
self.assertSequenceEqual(new_table.rows[1], ('a', 3, 'a-3'))
self.assertSequenceEqual(new_table.rows[2], ('b', 2, 'b-2'))
new_table = new_tableset['table2']
self.assertSequenceEqual(new_table.rows[0], ('b', 0, 'b-0'))
self.assertSequenceEqual(new_table.rows[1], ('a', 2, 'a-2'))
self.assertSequenceEqual(new_table.rows[2], ('c', 5, 'c-5'))
new_table = new_tableset['table3']
self.assertSequenceEqual(new_table.rows[0], ('a', 1, 'a-1'))
self.assertSequenceEqual(new_table.rows[1], ('a', 2, 'a-2'))
self.assertSequenceEqual(new_table.rows[2], ('c', 3, 'c-3'))
def test_compute(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.compute(
[(Formula(self.text_type, lambda r: "%(letter)s-%(number)i" % r), "new_column")]
)
new_table = new_tableset["table1"]
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 3)
self.assertSequenceEqual(new_table._column_types, (self.text_type, self.number_type, self.text_type))
self.assertSequenceEqual(new_table._column_names, ("letter", "number", "new_column"))
self.assertSequenceEqual(new_table.rows[0], ("a", 1, "a-1"))
self.assertSequenceEqual(new_table.rows[1], ("a", 3, "a-3"))
self.assertSequenceEqual(new_table.rows[2], ("b", 2, "b-2"))
new_table = new_tableset["table2"]
self.assertSequenceEqual(new_table.rows[0], ("b", 0, "b-0"))
self.assertSequenceEqual(new_table.rows[1], ("a", 2, "a-2"))
self.assertSequenceEqual(new_table.rows[2], ("c", 5, "c-5"))
new_table = new_tableset["table3"]
self.assertSequenceEqual(new_table.rows[0], ("a", 1, "a-1"))
self.assertSequenceEqual(new_table.rows[1], ("a", 2, "a-2"))
self.assertSequenceEqual(new_table.rows[2], ("c", 3, "c-3"))
def test_aggregate_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name="test")
new_table = tableset.aggregate([("count", Length())])
self.assertIsInstance(new_table, Table)
self.assertColumnNames(new_table, ("test", "count"))
self.assertColumnTypes(new_table, [Text, Number])
def test_aggregeate_bad_column(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
with self.assertRaises(KeyError):
tableset.aggregate([("one", Sum(), "one_sum")])
with self.assertRaises(KeyError):
tableset.aggregate([("bad", Sum(), "bad_sum")])
def test_merge_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='foo')
table = tableset.merge()
self.assertSequenceEqual(table.column_names, ['foo', 'letter', 'number'])
self.assertIsInstance(table.column_types[0], Text)
self.assertSequenceEqual(table.column_types[1:], [self.text_type, self.number_type])
def test_aggregeate_bad_column(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
with self.assertRaises(KeyError):
tableset.aggregate([('one', Sum(), 'one_sum')])
with self.assertRaises(KeyError):
tableset.aggregate([('bad', Sum(), 'bad_sum')])
def test_print_structure(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
output = StringIO()
tableset.print_structure(output=output)
lines = output.getvalue().strip().split('\n')
self.assertEqual(len(lines), 5)
def test_select(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.select(['number'])
for name, new_table in new_tableset.items():
self.assertColumnNames(new_table, ('number',))
self.assertColumnTypes(new_table, (Number,))
def test_select(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.select(['number'])
for name, new_table in new_tableset.items():
self.assertColumnNames(new_table, ('number',))
self.assertColumnTypes(new_table, (Number,))
def test_print_structure(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
output = StringIO()
tableset.print_structure(output=output)
lines = output.getvalue().strip().split('\n')
self.assertEqual(len(lines), 7)
def test_aggregate_row_names(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='test')
new_table = tableset.aggregate([
('count', Count())
])
self.assertRowNames(new_table, ['table1', 'table2', 'table3'])
def test_merge_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='foo')
table = tableset.merge()
self.assertSequenceEqual(table.column_names, ['foo', 'letter', 'number'])
self.assertIsInstance(table.column_types[0], Text)
self.assertSequenceEqual(table.column_types[1:], [self.text_type, self.number_type])
def test_aggregeate_bad_column(self):
tableset = TableSet(self.tables)
with self.assertRaises(ColumnDoesNotExistError):
tableset.aggregate([('one', Sum(), 'one_sum')])
with self.assertRaises(ColumnDoesNotExistError):
tableset.aggregate([('bad', Sum(), 'bad_sum')])
def test_aggregate_row_names(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name="test")
new_table = tableset.aggregate([("number", Length(), "count")])
self.assertSequenceEqual(new_table.row_names, ["table1", "table2", "table3"])
self.assertSequenceEqual(new_table.rows["table1"], ["table1", 3])
self.assertSequenceEqual(new_table.rows["table2"], ["table2", 3])
self.assertSequenceEqual(new_table.rows["table3"], ["table3", 3])
def test_aggregate_sum(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_table = tableset.aggregate([("count", Length()), ("number_sum", Sum("number"))])
self.assertIsInstance(new_table, Table)
self.assertColumnNames(new_table, ("group", "count", "number_sum"))
self.assertColumnTypes(new_table, [Text, Number, Number])
self.assertRows(new_table, [("table1", 3, 6), ("table2", 3, 7), ("table3", 3, 6)])
def test_aggregate_max_length(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_table = tableset.aggregate([("count", Length()), ("letter_max_length", MaxLength("letter"))])
self.assertIsInstance(new_table, Table)
self.assertColumnNames(new_table, ("group", "count", "letter_max_length"))
self.assertColumnTypes(new_table, [Text, Number, Number])
self.assertRows(new_table, [("table1", 3, 1), ("table2", 3, 1), ("table3", 3, 1)])
def test_select(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.select(['number'])
for name, new_table in new_tableset.items():
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 1)
self.assertSequenceEqual(new_table._column_types, (self.number_type,))
self.assertSequenceEqual(new_table._column_names, ('number',))
def test_aggregate_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='test')
new_table = tableset.aggregate([
('count', Count())
])
self.assertIsInstance(new_table, Table)
self.assertColumnNames(new_table, ('test', 'count'))
self.assertColumnTypes(new_table, [Text, Number])
def test_select(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_tableset = tableset.select(['number'])
for name, new_table in new_tableset.items():
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 1)
self.assertSequenceEqual(new_table._column_types, (self.number_type,))
self.assertSequenceEqual(new_table._column_names, ('number',))
def test_merge(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
table = tableset.merge()
self.assertColumnNames(table, ['group', 'letter', 'number'])
self.assertColumnTypes(table, [Text, Text, Number])
self.assertEqual(len(table.rows), 9)
self.assertSequenceEqual(table.rows[0], ['table1', 'a', 1])
self.assertSequenceEqual(table.rows[8], ['table3', 'c', 3])
def test_print_structure_row_limit(self):
tables = self.tables
for i in range(25):
tables[str(i)] = self.tables['table1']
tableset = TableSet(tables.values(), tables.keys())
output = StringIO()
tableset.print_structure(output=output)
lines = output.getvalue().strip().split('\n')
self.assertEqual(len(lines), 22)
def test_print_structure_row_limit(self):
tables = self.tables
for i in range(25):
tables[str(i)] = self.tables['table1']
tableset = TableSet(tables.values(), tables.keys())
output = StringIO()
tableset.print_structure(output=output)
lines = output.getvalue().strip().split('\n')
self.assertEqual(len(lines), 24)
def test_aggregate_row_names(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='test')
new_table = tableset.aggregate([
('number', Length(), 'count')
])
self.assertSequenceEqual(new_table.row_names, ['table1', 'table2', 'table3'])
self.assertSequenceEqual(new_table.rows['table1'], ['table1', 3])
self.assertSequenceEqual(new_table.rows['table2'], ['table2', 3])
self.assertSequenceEqual(new_table.rows['table3'], ['table3', 3])
def test_merge(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
table = tableset.merge()
self.assertColumnNames(table, ["group", "letter", "number"])
self.assertColumnTypes(table, [Text, Text, Number])
self.assertEqual(len(table.rows), 9)
self.assertSequenceEqual(table.rows[0], ["table1", "a", 1])
self.assertSequenceEqual(table.rows[8], ["table3", "c", 3])
def test_aggregate_min(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_table = tableset.aggregate([('count', Length()),
('number_min', Min('number'))])
self.assertIsInstance(new_table, Table)
self.assertColumnNames(new_table, ('group', 'count', 'number_min'))
self.assertColumnTypes(new_table, [Text, Number, Number])
self.assertRows(new_table, [('table1', 3, 1), ('table2', 3, 0),
('table3', 3, 1)])
def test_merge_groups(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='foo')
table = tableset.merge(groups=['red', 'blue', 'green'], group_name='color_code')
self.assertColumnNames(table, ['color_code', 'letter', 'number'])
self.assertColumnTypes(table, [Text, Text, Number])
self.assertEqual(len(table.rows), 9)
self.assertSequenceEqual(table.rows[0], ['red', 'a', 1])
self.assertSequenceEqual(table.rows[8], ['green', 'c', 3])
def test_merge_groups(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name="foo")
table = tableset.merge(groups=["red", "blue", "green"], group_name="color_code")
self.assertColumnNames(table, ["color_code", "letter", "number"])
self.assertColumnTypes(table, [Text, Text, Number])
self.assertEqual(len(table.rows), 9)
self.assertSequenceEqual(table.rows[0], ["red", "a", 1])
self.assertSequenceEqual(table.rows[8], ["green", "c", 3])
def test_aggregate_sum(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_table = tableset.aggregate([('count', Count()),
('number_sum', Sum('number'))])
self.assertIsInstance(new_table, Table)
self.assertColumnNames(new_table, ('group', 'count', 'number_sum'))
self.assertColumnTypes(new_table, [Text, Number, Number])
self.assertRows(new_table, [('table1', 3, 6), ('table2', 3, 7),
('table3', 3, 6)])
def test_merge(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
table = tableset.merge()
self.assertColumnNames(table, ['group', 'letter', 'number'])
self.assertColumnTypes(table, [Text, Text, Number])
self.assertEqual(len(table.rows), 9)
self.assertSequenceEqual(table.rows[0], ['table1', 'a', 1])
self.assertSequenceEqual(table.rows[8], ['table3', 'c', 3])
def test_aggregate_row_names(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='test')
new_table = tableset.aggregate([
('number', Length(), 'count')
])
self.assertSequenceEqual(new_table.row_names, ['table1', 'table2', 'table3'])
self.assertSequenceEqual(new_table.rows['table1'], ['table1', 3])
self.assertSequenceEqual(new_table.rows['table2'], ['table2', 3])
self.assertSequenceEqual(new_table.rows['table3'], ['table3', 3])
def test_having_simple(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='test')
new_tableset = tableset.having([
('count', Count())
], lambda t: t['count'] < 3)
self.assertIsInstance(new_tableset, TableSet)
self.assertSequenceEqual(new_tableset.keys(), ['table3'])
self.assertIs(new_tableset.values()[0], tableset['table3'])
self.assertEqual(new_tableset.key_name, 'test')
def test_aggregate_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name="test")
new_table = tableset.aggregate([("number", Length(), "count")])
self.assertIsInstance(new_table, Table)
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 2)
self.assertSequenceEqual(new_table._column_names, ("test", "count"))
self.assertIsInstance(new_table._column_types[0], Text)
self.assertIsInstance(new_table._column_types[1], Number)
def test_merge(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
table = tableset.merge()
self.assertSequenceEqual(table.column_names, ['group', 'letter', 'number'])
self.assertIsInstance(table.column_types[0], Text)
self.assertSequenceEqual(table.column_types[1:], [self.text_type, self.number_type])
self.assertEqual(len(table.rows), 9)
self.assertSequenceEqual(table.rows[0], ['table1', 'a', 1])
self.assertSequenceEqual(table.rows[8], ['table3', 'c', 3])
def test_having_simple(self):
tableset = TableSet(self.tables.values(),
self.tables.keys(),
key_name='test')
new_tableset = tableset.having([('count', Count())],
lambda t: t['count'] < 3)
self.assertIsInstance(new_tableset, TableSet)
self.assertSequenceEqual(new_tableset.keys(), ['table3'])
self.assertIs(new_tableset.values()[0], tableset['table3'])
self.assertEqual(new_tableset.key_name, 'test')
def test_having_complex(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='test')
new_tableset = tableset.having([
('count', Count()),
('number_sum', Sum('number'))
], lambda t: t['count'] >= 3 and t['number_sum'] > 6)
self.assertIsInstance(new_tableset, TableSet)
self.assertSequenceEqual(new_tableset.keys(), ['table2'])
self.assertIs(new_tableset.values()[0], tableset['table2'])
self.assertEqual(new_tableset.key_name, 'test')
def test_proxy_maintains_key(self):
number_type = Number()
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='foo', key_type=number_type)
self.assertEqual(tableset.key_name, 'foo')
self.assertEqual(tableset.key_type, number_type)
new_tableset = tableset.select(['number'])
self.assertEqual(new_tableset.key_name, 'foo')
self.assertEqual(new_tableset.key_type, number_type)
def test_proxy_maintains_key(self):
number_type = Number()
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='foo', key_type=number_type)
self.assertEqual(tableset.key_name, 'foo')
self.assertEqual(tableset.key_type, number_type)
new_tableset = tableset.select(['number'])
self.assertEqual(new_tableset.key_name, 'foo')
self.assertEqual(new_tableset.key_type, number_type)
def test_aggregate_sum(self):
tableset = TableSet(self.tables)
new_table = tableset.aggregate([('number', Sum(), 'number_sum')])
self.assertIsInstance(new_table, Table)
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 3)
self.assertSequenceEqual(new_table._column_names,
('group', 'count', 'number_sum'))
self.assertSequenceEqual(new_table.rows[0], ('table1', 3, 6))
self.assertSequenceEqual(new_table.rows[1], ('table2', 3, 7))
self.assertSequenceEqual(new_table.rows[2], ('table3', 3, 6))
def test_aggregate_key_name(self):
tableset = TableSet(self.tables.values(), self.tables.keys(), key_name='test')
new_table = tableset.aggregate([
('number', Length(), 'count')
])
self.assertIsInstance(new_table, Table)
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 2)
self.assertSequenceEqual(new_table._column_names, ('test', 'count'))
self.assertIsInstance(new_table._column_types[0], Text)
self.assertIsInstance(new_table._column_types[1], Number)
def test_having_complex(self):
tableset = TableSet(self.tables.values(),
self.tables.keys(),
key_name='test')
new_tableset = tableset.having(
[('count', Count()), ('number_sum', Sum('number'))],
lambda t: t['count'] >= 3 and t['number_sum'] > 6)
self.assertIsInstance(new_tableset, TableSet)
self.assertSequenceEqual(new_tableset.keys(), ['table2'])
self.assertIs(new_tableset.values()[0], tableset['table2'])
self.assertEqual(new_tableset.key_name, 'test')
def test_aggregate_min(self):
tableset = TableSet(self.tables)
new_table = tableset.aggregate([('number', Min(), 'number_min')])
self.assertIsInstance(new_table, Table)
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 3)
self.assertSequenceEqual(new_table._column_names,
('group', 'count', 'number_min'))
self.assertIsInstance(new_table.columns['number_min'], NumberColumn)
self.assertSequenceEqual(new_table.rows[0], ('table1', 3, 1))
self.assertSequenceEqual(new_table.rows[1], ('table2', 3, 0))
self.assertSequenceEqual(new_table.rows[2], ('table3', 3, 1))
def test_aggregate_max_length(self):
tableset = TableSet(self.tables)
new_table = tableset.aggregate([('letter', MaxLength(),
'letter_max_length')])
self.assertIsInstance(new_table, Table)
self.assertEqual(len(new_table.rows), 3)
self.assertEqual(len(new_table.columns), 3)
self.assertSequenceEqual(new_table._column_names,
('group', 'count', 'letter_max_length'))
self.assertSequenceEqual(new_table.rows[0], ('table1', 3, 1))
self.assertSequenceEqual(new_table.rows[1], ('table2', 3, 1))
self.assertSequenceEqual(new_table.rows[2], ('table3', 3, 1))
def test_aggregate_two_ops(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
new_table = tableset.aggregate([('count', Length()),
('number_sum', Sum('number')),
('number_mean', Mean('number'))])
self.assertIsInstance(new_table, Table)
self.assertColumnNames(new_table,
('group', 'count', 'number_sum', 'number_mean'))
self.assertColumnTypes(new_table, [Text, Number, Number, Number])
self.assertRows(new_table, [('table1', 3, 6, 2),
('table2', 3, 7, Decimal(7) / 3),
('table3', 3, 6, 2)])
def test_to_json(self):
tableset = TableSet(self.tables.values(), self.tables.keys())
tableset.to_json('.test-tableset')
for name in ['table1', 'table2', 'table3']:
with open('.test-tableset/%s.json' % name) as f:
contents1 = json.load(f)
with open('examples/tableset/%s.json' % name) as f:
contents2 = json.load(f)
self.assertEqual(contents1, contents2)
shutil.rmtree('.test-tableset')
def test_to_csv(self):
tableset = TableSet(self.tables)
tableset.to_csv('.test-tableset')
for name in ['table1', 'table2', 'table3']:
with open('.test-tableset/%s.csv' % name) as f:
contents1 = f.read()
with open('examples/tableset/%s.csv' % name) as f:
contents2 = f.read()
self.assertEqual(contents1, contents2)
shutil.rmtree('.test-tableset')
