import pytest import raccoon as rc from collections import OrderedDict from copy import deepcopy from raccoon.utils import assert_frame_equal from blist import blist def test_use_blist(): def check_blist(): assert isinstance(df.index, blist) assert isinstance(df.columns, blist) assert isinstance(df.data, blist) assert all([isinstance(df.data[x], blist) for x in range(len(df.columns))]) df = rc.DataFrame(use_blist=True) assert isinstance(df, rc.DataFrame) assert df.data == [] assert df.columns == [] assert df.index == [] assert df.sort is True check_blist() # add a new row and col df.set_cell(1, 'a', 1) check_blist() # add a new row df.set_cell(2, 'a', 2) check_blist() # add a new col df.set_cell(1, 'b', 3) check_blist() # add a complete new row df.set_row(3, {'a': 4, 'b': 5}) check_blist() # add a complete new col df.set_column([2, 3], 'c', [6, 7]) check_blist() def test_default_list(): def check_list(): assert isinstance(df.index, list) assert isinstance(df.columns, list) assert isinstance(df.data, list) assert all([isinstance(df.data[x], list) for x in range(len(df.columns))]) df = rc.DataFrame() assert isinstance(df, rc.DataFrame) assert df.data == [] assert df.columns == [] assert df.index == [] assert df.sort is True check_list() # add a new row and col df.set_cell(1, 'a', 1) check_list() # add a new row df.set_cell(2, 'a', 2) check_list() # add a new col df.set_cell(1, 'b', 3) check_list() # add a complete new row df.set_row(3, {'a': 4, 'b': 5}) check_list() # add a complete new col df.set_column([2, 3], 'c', [6, 7]) check_list() def test_to_dict(): df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, index=['a', 'b', 'c'], columns=['b', 'a']) # with index actual = df.to_dict(index=True) assert actual == {'index': ['a', 'b', 'c'], 'a': [1, 2, 3], 'b': [4, 5, 6]} # without index actual = df.to_dict(index=False) assert actual == {'a': [1, 2, 3], 'b': [4, 5, 6]} # ordered act_order = df.to_dict(ordered=True) expected = OrderedDict([('index', ['a', 'b', 'c']), ('b', [4, 5, 6]), ('a', [1, 2, 3])]) assert act_order == expected def test_to_list(): df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'c': [7, 8, 9]}, index=[4, 5, 6], columns=['b', 'a', 'c']) assert df['b'].to_list() == [4, 5, 6] assert df[[4, 6], 'a'].to_list() == [1, 3] assert df[4:5, 'c'].to_list() == [7, 8] assert df[[6], 'c'].to_list() == [9] # cannot to_list on a multi-column DataFrame with pytest.raises(TypeError): df.to_list() def test_json(): df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'c': [7, 8, 9]}, index=[4, 5, 6], columns=['b', 'a', 'c']) str = df.to_json() actual = rc.DataFrame.from_json(str) assert_frame_equal(df, actual) df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'c': [7, 8, 9]}, use_blist=True, sort=False) str = df.to_json() actual = rc.DataFrame.from_json(str) assert_frame_equal(df, actual) # empty DataFrame df = rc.DataFrame({'a': [], 'b': [], 'c': []}) str = df.to_json() actual = rc.DataFrame.from_json(str) assert_frame_equal(df, actual) df = rc.DataFrame() str = df.to_json() actual = rc.DataFrame.from_json(str) assert_frame_equal(df, actual) def test_json_objects(): # test with a compound object returning a representation df = rc.DataFrame({'a': [1, 2], 'b': [4, blist([5, 6])]}) str = df.to_json() actual = rc.DataFrame.from_json(str) # the DataFrames are not equal because the blist() was converted to a representation with pytest.raises(AssertionError): assert_frame_equal(df, actual) assert actual[1, 'b'] != blist([5, 6]) assert actual[1, 'b'] == 'blist([5, 6])' def test_json_multi_index(): df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'c': [7, 8, 9]}, index=[('a', 4), ('b', 5), ('c', 6)]) str = df.to_json() actual = rc.DataFrame.from_json(str) assert_frame_equal(df, actual) df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'c': [7, 8, 9]}, index=[('a', 4), ('b', 5), ('c', 6)], index_name=('first', 'second')) str = df.to_json() actual = rc.DataFrame.from_json(str) assert_frame_equal(df, actual) def test_append(): # duplicate indexes df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b'], index=[0, 1, 2]) df2 = rc.DataFrame({'a': [11, 12, 13], 'b': [14, 15, 16]}, columns=['a', 'b'], index=[2, 3, 4]) with pytest.raises(ValueError): df.append(df2) # all same columns df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b'], index=[0, 1, 2]) df2 = rc.DataFrame({'a': [11, 12, 13], 'b': [14, 15, 16]}, columns=['a', 'b'], index=[3, 4, 5]) df.append(df2) assert_frame_equal(df, rc.DataFrame({'a': [1, 2, 3, 11, 12, 13], 'b': [4, 5, 6, 14, 15, 16]}, columns=['a', 'b'], index=[0, 1, 2, 3, 4, 5])) # all new columns df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b'], index=[0, 1, 2]) df2 = rc.DataFrame({'x': [11, 12, 13], 'y': [14, 15, 16]}, columns=['x', 'y'], index=[3, 4, 5]) df.append(df2) assert_frame_equal(df, rc.DataFrame({'a': [1, 2, 3, None, None, None], 'b': [4, 5, 6, None, None, None], 'x': [None, None, None, 11, 12, 13], 'y': [None, None, None, 14, 15, 16]}, columns=['a', 'b', 'x', 'y'], index=[0, 1, 2, 3, 4, 5])) # some same, some new columns df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b'], index=[0, 1, 2]) df2 = rc.DataFrame({'b': [11, 12, 13], 'y': [14, 15, 16]}, columns=['b', 'y'], index=[3, 4, 5]) df.append(df2) assert_frame_equal(df, rc.DataFrame({'a': [1, 2, 3, None, None, None], 'b': [4, 5, 6, 11, 12, 13], 'y': [None, None, None, 14, 15, 16]}, columns=['a', 'b', 'y'], index=[0, 1, 2, 3, 4, 5])) # append empty DataFrame df2 = rc.DataFrame(columns=['a', 'b', 'y']) df.append(df2) assert_frame_equal(df, rc.DataFrame({'a': [1, 2, 3, None, None, None], 'b': [4, 5, 6, 11, 12, 13], 'y': [None, None, None, 14, 15, 16]}, columns=['a', 'b', 'y'], index=[0, 1, 2, 3, 4, 5])) def test_rename_columns(): df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, index=['a', 'b', 'c'], columns=['b', 'a']) df.rename_columns({'b': 'b_new'}) assert df.columns == ['b_new', 'a'] df.rename_columns({'b_new': 'b2', 'a': 'a2'}) assert df.columns == ['b2', 'a2'] with pytest.raises(ValueError): df.rename_columns({'a2': 'a', 'bad': 'nogo'}) def test_show(): df = rc.DataFrame({'a': [1, 2, 3], 'b': [1.0, 2.55, 3.1], 'c': ['first', 'second', None]}, columns=['b', 'c', 'a'], index=['row1', 'row2', 'row3'], use_blist=True) # __repr__ produces a simple representation expected = "object id: %s\ncolumns:\nblist(['b', 'c', 'a'])\ndata:\nblist([blist([1.0, 2.55, 3.1]), blist([" \ "'first', 'second', None]), blist([1, 2, 3])])\nindex:\nblist(['row1', 'row2', 'row3'])\n" % id(df) actual = df.__repr__() assert actual == expected # __str__ produces the standard table expected = 'index b c a\n------- ---- ------ ---\nrow1 1 first 1\n' \ 'row2 2.55 second 2\nrow3 3.1 3' actual = df.__str__() assert actual == expected # show() method will pass along any argument for the tabulate.tabulate function df.show() def test_input_data_mutability(): input_data = {'a': [1, 2, 3], 'b': [4, 5, 6]} # without defining column order df = rc.DataFrame(input_data) orig_data = deepcopy(df.data) # change input_data input_data['c'] = [6, 7, 8] assert df.to_dict(index=False) != input_data assert df.data == orig_data # change an inner index of input data input_data['a'].append(99) assert df.data == orig_data # Now make an inner element a mutable item, confirm that mutability remains input_data = {'a': [[1], [2], [3]], 'b': [4, 5, 6]} df = rc.DataFrame(input_data) orig_data = deepcopy(df.data) # changing the input data changes the inner data in DataFrame input_data['a'][0].append(11) assert df.data != orig_data assert df.get(0, 'a') == [1, 11] # using set to change the DataFrame data does not effect the input data df[1, 'a'] = [2, 22] assert input_data['a'] == [[1, 11], [2], [3]] df.set(columns='b', values=[44, 55, 66]) assert input_data['b'] == [4, 5, 6] def test_get_data_mutability(): # the .data method only returns a shallow copy, and changes to the return values will corrupt the DataFrame df = rc.DataFrame({'a': [1, 2, 3], 'b': [1.0, 2.55, 3.1], 'c': ['first', 'second', None]}, columns=['a', 'b', 'c']) orig_data = deepcopy(df.data) data = df.data data[0].append(99) assert df.data != orig_data assert df.data[0] == [1, 2, 3, 99] # using the get commands returns a shallow copy df = rc.DataFrame({'a': [1, 2, 3], 'b': [[1], [2], [3]]}, columns=['a', 'b']) orig_data = deepcopy(df.data) new_df = df['a'] new_df[3, 'a'] = 100 assert df.data == orig_data # get a slice new_df = df['b'] # mutate inner value new_df[1, 'b'].append(22) # changes the new_df assert new_df.data == [[[1], [2, 22], [3]]] # changes original df assert new_df.data[0] == df.data[1] def test_len(): df = rc.DataFrame() assert len(df) == 0 df = rc.DataFrame({'a': [1, 2, 3], 'b': [1.0, 2.55, 3.1]}, columns=['a', 'b'], sort=False) assert len(df) == 3 df['a', 3] = 99 assert len(df) == 4 def test_equality(): df = rc.DataFrame({'z': [1, 2, 1, 2, 1, 1]}) assert df.sort is True assert df.equality('z', value=1) == [True, False, True, False, True, True] assert df.equality('z', [1, 2, 3], 2) == [True, False, True] assert df.equality('z', [False, False, False, True, True, True], 1) == [False, True, True] # change all 1 to 3 df.set(indexes=df.equality('z', value=1), columns='z', values=3) assert df.data == [[3, 2, 3, 2, 3, 3]] df = rc.DataFrame({'z': [1, 2, 1, 2, 1, 1]}, sort=False) assert df.sort is False assert df.equality('z', value=1) == [True, False, True, False, True, True] assert df.equality('z', [1, 2, 3], 2) == [True, False, True] assert df.equality('z', [False, False, False, True, True, True], 1) == [False, True, True] # not enough booleans to match index len with pytest.raises(ValueError): df.equality('z', [True, True], 2) def test_math(): df = rc.DataFrame({'a': [1, 2, 3, 4], 'b': [5, 6, 7, 8]}, columns=['a', 'b']) res = df.add('a', 'b') assert res == [6, 8, 10, 12] df['c'] = res assert df.data == [[1, 2, 3, 4], [5, 6, 7, 8], [6, 8, 10, 12]] assert df.columns == ['a', 'b', 'c'] res = df.add('a', 'b', [1, 3]) assert res == [8, 12] res = df.subtract('c', 'b') assert res == [1, 2, 3, 4] res = df.multiply('a', 'c', [True, True, False, False]) assert res == [6, 16] res = df.divide('b', 'c', [0, 2]) assert res == [5/6, 7/10] with pytest.raises(ValueError): df.multiply('a', 'c', [True, True]) def test_select_index(): # simple index, not sort df = rc.DataFrame({'a': [1, 2, 3, 4, 5, 6]}, index=['a', 'b', 'c', 'd', 'e', 'f']) actual = df.select_index('c', 'value') assert actual == ['c'] actual = df.select_index('d', 'boolean') assert actual == [False, False, False, True, False, False] # simple index, sort df = rc.DataFrame({'a': [1, 2, 3, 4, 5, 6]}, index=['a', 'b', 'c', 'd', 'e', 'f'], sort=True) actual = df.select_index('c', 'value') assert actual == ['c'] actual = df.select_index('d', 'boolean') assert actual == [False, False, False, True, False, False] with pytest.raises(ValueError): df.select_index('a', 'BAD') # simple index, not sort, blist df = rc.DataFrame({'a': [1, 2, 3, 4, 5, 6]}, index=['a', 'b', 'c', 'd', 'e', 'f'], use_blist=True) actual = df.select_index('c', 'value') assert actual == ['c'] actual = df.select_index('d', 'boolean') assert actual == [False, False, False, True, False, False] # tuple index tuples = [('a', 1, 3), ('a', 1, 4), ('a', 2, 3), ('b', 1, 4), ('b', 2, 1), ('b', 3, 3)] df = rc.DataFrame({'a': [1, 2, 3, 4, 5, 6]}, index=tuples) compare = ('a', None, None) assert df.select_index(compare) == [True, True, True, False, False, False] compare = ('a', None, 3) assert df.select_index(compare, 'boolean') == [True, False, True, False, False, False] compare = (None, 2, None) assert df.select_index(compare, 'value') == [('a', 2, 3), ('b', 2, 1)] compare = (None, 3, 3) assert df.select_index(compare) == [False, False, False, False, False, True] compare = (None, None, 3) assert df.select_index(compare, 'value') == [('a', 1, 3), ('a', 2, 3), ('b', 3, 3)] compare = ('a', 1, 4) assert df.select_index(compare, 'value') == [('a', 1, 4)] compare = ('a', 100, 99) assert df.select_index(compare, 'value') == [] compare = (None, None, None) assert df.select_index(compare) == [True] * 6 df = rc.DataFrame({'a': [1, 2, 3, 4, 5, 6]}) assert df.select_index(3) == [False, False, False, True, False, False] assert df.select_index(3, 'value') == [3] def test_isin(): df = rc.DataFrame({'first': [1, 2, 3, 4, 5], 'second': ['a', 2, 'b', None, 5]}) assert df.isin('first', [2, 3, 4]) == [False, True, True, True, False] assert df.isin('first', [3]) == [False, False, True, False, False] assert df.isin('first', [6, 7]) == [False, False, False, False, False] assert df.isin('second', ['a', 2]) == [True, True, False, False, False] assert df.isin('second', ['a', 'b']) == [True, False, True, False, False] assert df.isin('second', ['a', 'b', None]) == [True, False, True, True, False] def test_reset_index(): # no index defined df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b']) df.reset_index() expected = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'index_0': [0, 1, 2]}, columns=['a', 'b', 'index_0']) assert_frame_equal(df, expected) # with index and index name defined df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b'], index=['x', 'y', 'z'], index_name='jelo') df.reset_index() expected = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'jelo': ['x', 'y', 'z']}, columns=['a', 'b', 'jelo'], sort=False) assert_frame_equal(df, expected) # with a tuple multi-index df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b'], index=[('a', 10, 'x'), ('b', 11, 'y'), ('c', 12, 'z')], index_name=('melo', 'helo', 'gelo')) df.reset_index() expected = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], 'melo': ['a', 'b', 'c'], 'helo': [10, 11, 12], 'gelo': ['x', 'y', 'z']}, columns=['a', 'b', 'melo', 'helo', 'gelo'], sort=False) assert_frame_equal(df, expected) # drop df = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6]}, columns=['a', 'b'], index=['x', 'y', 'z'], index_name='jelo') df.reset_index(drop=True) expected = rc.DataFrame({'a': [1, 2, 3], 'b': [4, 5, 6], }, columns=['a', 'b'], sort=False) assert_frame_equal(df, expected)