def test_merge_3():
table_a = Table()
table_b = Table()
table_a.keys = ['a', 'b']
table_a.data = [np.array([1, 2, 3]), np.array([1, 2, 3])]
table_a.index = np.array(np.array([[1, 1, 1], [1, 1, 1]], dtype=np.uint8))
table_b.keys = ['a', 'c']
table_b.data = [np.array([1, 2, 3]), np.array([1, 2, 3])]
table_b.index = np.array(np.array([[1, 1, 1], [1, 1, 1]], dtype=np.uint8))
table_a.merge(table_b, 'a')
assert np.all(table_a.a.index == np.array([1, 1, 1, 1, 1, 1]))
assert np.all(table_a.b.index == np.array([1, 0, 1, 0, 1, 0]))
assert np.all(table_a.c.index == np.array([0, 1, 0, 1, 0, 1]))
def test_fill_table():
t = Table()
t.keys = ['a', 'b']
t.data = [np.arange(10), np.array([1, 2])]
t.index = np.array(
[[1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [0, 0, 1, 0, 0, 0, 0, 1, 0, 0]],
dtype=np.uint8)
t.fill_column('b', fillvalue=3)
assert np.all(t.b.values == np.array([3, 3, 1, 3, 3, 3, 3, 2, 3, 3]))
def test_fill_empty_column():
t = Table()
t.keys = ['a', 'b']
t.data = [np.arange(10), np.array([])]
t.index = np.array(
[[1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]],
dtype=np.uint8)
b = t.b
b.fill(fillvalue=3)
assert np.all(b.values == np.array([3, 3, 3, 3, 3, 3, 3, 3, 3, 3]))
def test_fill_full_column():
t = Table()
t.keys = ['a', 'b']
t.data = [np.arange(10), np.arange(10)]
t.index = np.array(
[[1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]],
dtype=np.uint8)
b = t.b
b.fill(fillvalue=3)
assert np.all(b.values == np.array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]))
def test_fillna_reverse_fillvalue():
t = Table()
t.keys = ['a', 'b']
t.data = [np.arange(10), np.array([1, 2])]
t.index = np.array(
[[1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [0, 0, 1, 0, 0, 0, 0, 1, 0, 0]],
dtype=np.uint8)
b = t.b
b.fillna(reverse=True, fillvalue=-1)
assert np.all(b.values == np.array([1, 1, 1, 2, 2, 2, 2, 2, -1, -1]))
assert np.all(b.index == np.array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1]))
def full_outer_join(table_left, table_right, column, check_sorted=True):
"""
Inner join. If columns are repeated, the left table has preference.
:param table_left:
:param table_right:
:param column:
:param check_sorted: If True may increase performance, but breaks if the
column used as index is not sorted.
:return:
"""
if column not in table_left.keys:
raise ValueError('{} not in left table'.format(column))
if column not in table_right.keys:
raise ValueError('{} not in right table'.format(column))
all_columns = set(chain(table_left.keys, table_right.keys))
joined_columns = all_columns - {column}
common_left = table_left.get(column)
common_right = table_right.get(column)
if check_sorted:
if not np.all(common_left.values == np.sort(common_left.values)):
raise ValueError('Trying to join with a non sorted column')
if not np.all(common_right.values == np.sort(common_right.values)):
raise ValueError('Trying to join with a non sorted column')
common_rec = union_sorted(common_left.values, common_right.values)
data_joined, global_left, global_right = join_low_level(
common_left.values, common_left.index,
common_right.values, common_right.index, common_rec)
data = list()
index = list()
keys = list()
data.append(data_joined)
index.append(np.ones(len(data_joined), dtype=np.uint8))
keys.append(column)
for i_column in joined_columns:
if (i_column in table_left) and (i_column in table_right):
cl = table_left.get(i_column)
cr = table_right.get(i_column)
c_values, c_index = reindex_join_columns(
cl, cr, global_left, global_right)
keys.append(i_column)
data.append(c_values)
index.append(c_index)
elif i_column in table_left:
c = table_left.get(i_column)
c = c.reindex(global_left)
keys.append(i_column)
data.append(c.values)
index.append(c.index)
elif i_column in table_right:
c = table_right.get(i_column)
c = c.reindex(global_right)
keys.append(i_column)
data.append(c.values)
index.append(c.index)
res = Table()
res.data = data
res.index = np.vstack(index)
res.keys = keys
return res
def inner_join(table_left, table_right, column):
"""
Inner join. If columns are repeated, the left table has preference.
:param table_left:
:param table_right:
:param column:
:return:
"""
if column not in table_left.keys:
raise ValueError('{} not in left table'.format(column))
if column not in table_right.keys:
raise ValueError('{} not in right table'.format(column))
all_columns = set(chain(table_left.keys, table_right.keys))
joined_columns = all_columns - {column}
common_left = table_left.get(column)
common_right = table_right.get(column)
if not np.all(common_left.values == np.sort(common_left.values)):
raise ValueError('Trying to join with a non sorted column')
if not np.all(common_right.values == np.sort(common_right.values)):
raise ValueError('Trying to join with a non sorted column')
common_rec = intersection_sorted(common_left.values, common_right.values)
data_joined, global_left, global_right = join_low_level(
common_left.values, common_left.index,
common_right.values, common_right.index, common_rec)
data = list()
index = list()
keys = list()
data.append(data_joined)
index.append(np.ones(len(data_joined), dtype=np.uint8))
keys.append(column)
for i_column in joined_columns:
if i_column in table_left:
c = table_left.get(i_column)
c = c.reindex(global_left)
keys.append(i_column)
data.append(c.values)
index.append(c.index)
elif i_column in table_right:
c = table_right.get(i_column)
c = c.reindex(global_right)
keys.append(i_column)
data.append(c.values)
index.append(c.index)
res = Table()
res.data = data
res.index = np.vstack(index)
res.keys = keys
return res