def test_row_append_combinations(self):
first_tbl = [
Tabl(),
Tabl({
'a': 2,
'b': 3
}), Tabl({
'a': "ff",
'b': "ffr"
})
]
seccond_tbl = [{
'a': 22,
'b': 13
}, [22, 33], (44, 55),
np.array([33, 33]), {
'a': "ded",
'b': "fr"
}, ["frd", "werf"], ("efds", "dew"),
np.array(["frde", "frtg"])]
for ftbl, stbl in product(first_tbl, seccond_tbl):
ftbl = deepcopy(ftbl)
fl = len(ftbl)
ftbl.row_append(stbl)
assert ftbl.valid
assert len(ftbl) == fl + 1
def test_append_combinations(self):
first_tbl = [
Tabl(),
Tabl({
'a': 2,
'b': 3
}), Tabl({
'a': "ff",
'b': "ffr"
})
]
seccond_tbl = [
Tabl({
'a': 22,
'b': 13
}),
pd.DataFrame({
'a': [22],
'b': [13]
}),
Tabl({
'a': "ff",
'b': "ffr"
}),
pd.DataFrame({
'a': ["jdw"],
'b': ["okdw"]
}),
]
for ftbl, stbl in product(first_tbl, seccond_tbl):
ftbl = deepcopy(ftbl)
fl = len(ftbl)
ftbl.append(stbl)
assert ftbl.valid
assert len(ftbl) == fl + 1
def test_inner_int_join(self):
tbl = Tabl({"a": list(range(4)), "b": ['a', 'b'] * 2})
tbl_b = Tabl({"a": list(range(2, 6)), "c": ['d', 'e'] * 2})
tbl_j = tbl.join(tbl_b, "a")
assert "c_r" in tbl_j.columns
assert tbl_j[0] == (2, 'a', 'd')
assert len(tbl_j) == 2
assert np.all(tbl[0] == (0, 'a'))
def test_inner_str_join(self):
tbl = Tabl({"a": list(map(str, range(4))), "b": ['a', 'b'] * 2})
tbl_b = Tabl({"a": list(map(str, range(4))), "c": ['d', 'e'] * 2})
tbl_j = tbl.join(tbl_b, "a", jointype="inner")
assert "c_r" in tbl_j.columns
assert tbl_j[0] == ('0', 'a', 'd')
assert len(tbl_j) == 4
assert np.all(tbl[0] == ('0', 'a'))
def test_dtype_properties(self):
tbl = Tabl({
"a": list(range(2, 6)),
"c": [u'1', u'2'] * 2,
"d": [1.1, 2.2] * 2
})
assert tbl.dtype == np.dtype([('a', '<i8'), ('c', '<U1'),
('d', '<f8')])
tbl_n = tbl.astype([np.float, np.int16, np.dtype('<U10')])
assert tbl_n.dtype == np.dtype([('a', '<f8'), ('c', '<i2'),
('d', '<U10')])
def test_transpose(self, tbls):
for tbl in tbls:
if len(tbl) == 0:
continue
lrec = [tbl[i] for i in range(len(tbl))]
tbl_n = Tabl(T(lrec))
for i, column in enumerate(tbl.columns):
assert np.all(tbl[column] == tbl_n[str(i)])
lrec = [list(tbl[i]) for i in range(len(tbl))]
tbl_n = Tabl(T(lrec))
for i, column in enumerate(tbl.columns):
assert np.all(tbl[column] == tbl_n[str(i)])
def test_outer_join(self):
tbl = Tabl({"a": list(range(4)), "b": ['a', 'b'] * 2})
tbl_b = Tabl({
"a": list(range(2, 6)),
"c": [1, 2] * 2,
"d": [1.1, 2.2] * 2
})
tbl_j = tbl.join(tbl_b, "a", jointype='outer')
naneq(tbl_j[tbl_j['a_l'] == 0, 'c_r'], Tabl.join_fill_value['integer'])
naneq(tbl_j[tbl_j['a_l'] == 0, 'd_r'], Tabl.join_fill_value['float'])
naneq(tbl_j[tbl_j['a_r'] == 4, 'b_l'], Tabl.join_fill_value['string'])
assert "c_r" in tbl_j.columns
assert len(tbl_j) == 6
def test_pandas_b_and_f(self):
data = {
'a': [12] * 6,
'b': np.arange(6),
'c': list(map(str, range(6)))
}
df = pd.DataFrame(data)
tbl = Tabl(df)
for k in data.keys():
assert np.all(tbl.dict['a'] == data['a'])
tbl = Tabl(data)
df = pd.DataFrame(tbl.dict)
for k in data.keys():
assert np.all(df['a'] == data['a'])
def test_group_by(self):
tbl = Tabl({
'a': [10, 20, 30, 40] * 3,
'b': ["100", "200"] * 6,
'c': [100, 200] * 6
})
tbl_g = tbl.group_by(['b', 'a'], [(np.sum, 'a'), (first, 'c')])
idx = np.argsort(tbl_g['a_sum'])
assert np.all(tbl_g[idx, 'a_sum'] == np.array([30, 60, 90, 120]))
assert np.all(tbl['a'] == [10, 20, 30, 40] * 3)
assert np.all(tbl_g[idx, 'c_first'] == np.array([100, 200, 100, 200]))
tbl_g = tbl.group_by(['b', 'a'], [])
idx = np.argsort(tbl_g['a'])
assert np.all(tbl_g[idx, 'a'] == [10, 20, 30, 40])
def test_del_col(self):
tbl1 = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
tbl2 = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
del tbl1['b']
assert tbl1.columns == ['a', 'c']
assert all(tbl1['a'] == tbl2['a'])
assert all(tbl1['c'] == tbl2['c'])
def test_del_row_slice(self):
tbl1 = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
tbl2 = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
del tbl1[1:3]
assert len(tbl1) == 2
assert tbl1[0] == tbl2[0]
assert tbl1[1] == tbl2[3]
def test_set_row(self):
tbl = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
tbl[0, :] = (12, "12", 12.12)
tbl[1, ['a', 'b']] = [12, "12"]
tbl[2] = (13, "13", 13.13)
tbl[3, np.array([True, False, True])] = (14, 14.14)
def test_del_row_int(self):
tbl1 = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
tbl2 = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
del tbl1[0]
assert len(tbl1) == 3
for i in range(len(tbl1)):
assert tbl1[i] == tbl2[i + 1]
del tbl1[-1]
assert len(tbl1) == 2
assert tbl1[0] == tbl2[1]
assert tbl1[1] == tbl2[2]
def test_add_column(self):
tbl = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
tbl["d"] = np.array([True] * len(tbl))
tbl["e"] = 100
tbl["f"] = "Blanc"
assert len(tbl.columns) == 6
assert np.all([len(c) == len(tbl) for c in tbl.data])
def test_sort(self):
tbl = Tabl({'aa': ['b', 'g', 'd'], 'b': list(range(3))})
tbl.sort('aa')
assert tbl[0] == ('b', 0)
assert tbl[1] == ('d', 2)
tbl.sort(['aa', 'b'])
assert tbl[1] == ('g', 1)
def test_replace_column(self):
tbl = Tabl({
"a": list(range(2, 6)),
"b": ['1', '2'] * 2,
"c": [1.1, 2.2] * 2
})
tbl[:, "a"] = np.array([1.4, 2.4] * 2)
assert tbl[0, 'a'] == 1
tbl["a"] = np.array([1.1, 2.2] * 2)
assert tbl['a'].dtype == np.float
tbl[:, "b"] = np.arange(4)
assert tbl[0, "b"] == "0"
tbl["b"] = np.arange(4)
assert tbl[0, "b"] == 0
tbl[:, "c"] = np.arange(4)
assert tbl["c"].dtype == np.float
tbl["c"] = np.arange(4)
assert tbl["c"].dtype == np.int
def tbls():
tbls = [
Tabl({
'Name': ["John", "Joe", "Jane"],
'Height': [1.82, 1.65, 2.15],
'Married': [False, False, True]
}),
Tabl([
np.array(["John", "Joe", "Jane"]),
np.array([1.82, 1.65, 2.15]),
np.array([False, False, True])
],
columns=["Name", "Height", "Married"]),
Tabl([["John", "Joe", "Jane"], [1.82, 1.65, 2.15],
[False, False, True]],
columns=["Name", "Height", "Married"]),
Tabl(),
Tabl([1, 2, 3]),
Tabl({
'a': range(10),
'b': range(10)
}),
Tabl({
'a': range(10),
'b': list(map(str, range(10)))
}),
Tabl([range(10), range(10), range(10)]),
Tabl(np.array([np.arange(10), np.arange(10, 20)])),
Tabl({
'a': 12,
'b': 13
}),
Tabl(pd.DataFrame({
'a': [12] * 6,
'b': np.arange(6),
'c': range(6)
})),
Tabl({
'a': [3, 4, 5],
'b': 1
}),
Tabl({
'a': 1,
'b': [3, 4, 5]
}),
Tabl({
'aa': [1, 1, 1],
'b': [3, 4, 5]
}),
]
return tbls