@pytest.mark.parametrize("data", test_data_values, ids=test_data_keys)
@pytest.mark.parametrize("key", indices_values, ids=indices_keys)
def test_get(data, key):
modin_df = pd.DataFrame(data)
pandas_df = pandas.DataFrame(data)
df_equals(modin_df.get(key), pandas_df.get(key))
df_equals(modin_df.get(key, default="default"),
pandas_df.get(key, default="default"))
@pytest.mark.parametrize("data", test_data_values, ids=test_data_keys)
@pytest.mark.parametrize("dummy_na",
bool_arg_values,
ids=arg_keys("dummy_na", bool_arg_keys))
@pytest.mark.parametrize("drop_first",
bool_arg_values,
ids=arg_keys("drop_first", bool_arg_keys))
def test_get_dummies(request, data, dummy_na, drop_first):
modin_df = pd.DataFrame(data)
pandas_df = pandas.DataFrame(data)
try:
pandas_result = pandas.get_dummies(pandas_df,
dummy_na=dummy_na,
drop_first=drop_first)
except Exception as e:
with pytest.raises(type(e)):
pd.get_dummies(modin_df, dummy_na=dummy_na, drop_first=drop_first)
else:
generate_multiindex,
test_data_diff_dtype,
)
from modin.config import NPartitions
NPartitions.put(4)
# Force matplotlib to not use any Xwindows backend.
matplotlib.use("Agg")
@pytest.mark.parametrize("method", ["all", "any"])
@pytest.mark.parametrize("is_transposed", [False, True])
@pytest.mark.parametrize("skipna",
bool_arg_values,
ids=arg_keys("skipna", bool_arg_keys))
@pytest.mark.parametrize("axis", axis_values, ids=axis_keys)
@pytest.mark.parametrize("data", [test_data["float_nan_data"]])
def test_all_any(data, axis, skipna, is_transposed, method):
eval_general(
*create_test_dfs(data),
lambda df: getattr((df.T if is_transposed else df), method)
(axis=axis, skipna=skipna, bool_only=None),
)
@pytest.mark.parametrize("method", ["all", "any"])
@pytest.mark.parametrize("bool_only",
bool_arg_values,
ids=arg_keys("bool_only", bool_arg_keys))
def test_all_any_specific(bool_only, method):
@pytest.mark.parametrize("data", test_data_values, ids=test_data_keys)
@pytest.mark.parametrize("key", indices_values, ids=indices_keys)
def test_get(data, key):
modin_df = pd.DataFrame(data)
pandas_df = pandas.DataFrame(data)
df_equals(modin_df.get(key), pandas_df.get(key))
df_equals(
modin_df.get(key, default="default"), pandas_df.get(key, default="default")
)
@pytest.mark.parametrize("data", test_data_values, ids=test_data_keys)
@pytest.mark.parametrize(
"dummy_na", bool_arg_values, ids=arg_keys("dummy_na", bool_arg_keys)
)
@pytest.mark.parametrize(
"drop_first", bool_arg_values, ids=arg_keys("drop_first", bool_arg_keys)
)
def test_get_dummies(request, data, dummy_na, drop_first):
modin_df = pd.DataFrame(data)
pandas_df = pandas.DataFrame(data)
try:
pandas_result = pandas.get_dummies(
pandas_df, dummy_na=dummy_na, drop_first=drop_first
)
except Exception as e:
with pytest.raises(type(e)):
pd.get_dummies(modin_df, dummy_na=dummy_na, drop_first=drop_first)
rows_number = len(next(iter(
data.values()))) # length of the first data column
level_0 = np.random.choice([0, 1, 2], rows_number)
level_1 = np.random.choice([3, 4, 5], rows_number)
index = pd.MultiIndex.from_arrays([level_0, level_1])
eval_general(
pd.DataFrame(data, index=index),
pandas.DataFrame(data, index=index),
lambda df, *args, **kwargs: df.apply(func, *args, **kwargs),
**func_kwargs,
)
@pytest.mark.parametrize("column", ["A", ["A", "C"]],
ids=arg_keys("column", ["A", ["A", "C"]]))
@pytest.mark.parametrize("ignore_index",
bool_arg_values,
ids=arg_keys("ignore_index", bool_arg_keys))
def test_explode_single_partition(column, ignore_index):
# This test data has two columns where some items are lists that
# explode() should expand. In some rows, the columns have list-like
# elements that must be expanded, and in others, they have empty lists
# or items that aren't list-like at all.
data = {
"A": [[0, 1, 2], "foo", [], [3, 4]],
"B": 1,
"C": [["a", "b", "c"], np.nan, [], ["d", "e"]],
}
eval_general(
*create_test_dfs(data),
df_equals(
modin_df.asof(modin_where.values[0], subset=subset),
pandas_df.asof(pandas_where.values[0], subset=subset),
)
@pytest.mark.parametrize("data", test_data_values, ids=test_data_keys)
def test_first_valid_index(data):
modin_df = pd.DataFrame(data)
pandas_df = pandas.DataFrame(data)
assert modin_df.first_valid_index() == (pandas_df.first_valid_index())
@pytest.mark.parametrize("data", test_data_values, ids=test_data_keys)
@pytest.mark.parametrize("n", int_arg_values, ids=arg_keys("n", int_arg_keys))
def test_head(data, n):
# Test normal dataframe head
modin_df = pd.DataFrame(data)
pandas_df = pandas.DataFrame(data)
df_equals(modin_df.head(n), pandas_df.head(n))
df_equals(modin_df.head(len(modin_df) + 1),
pandas_df.head(len(pandas_df) + 1))
# Test head when we call it from a QueryCompilerView
modin_result = modin_df.loc[:, ["col1", "col3", "col3"]].head(n)
pandas_result = pandas_df.loc[:, ["col1", "col3", "col3"]].head(n)
df_equals(modin_result, pandas_result)
@pytest.mark.skip(reason="Defaulting to Pandas")
# Named Series promoted to DF
s = pd.Series(frame_data2.get("col1"))
with pytest.raises(ValueError):
modin_df.merge(s)
s = pd.Series(frame_data2.get("col1"), name="col1")
df_equals(modin_df.merge(s), modin_df.merge(modin_df2[["col1"]]))
with pytest.raises(TypeError):
modin_df.merge("Non-valid type")
@pytest.mark.parametrize("axis", [0, 1])
@pytest.mark.parametrize("ascending",
bool_arg_values,
ids=arg_keys("ascending", bool_arg_keys))
@pytest.mark.parametrize("na_position", ["first", "last"],
ids=["first", "last"])
def test_sort_index(axis, ascending, na_position):
data = test_data["float_nan_data"]
modin_df, pandas_df = pd.DataFrame(data), pandas.DataFrame(data)
# Change index value so sorting will actually make a difference
if axis == 0:
length = len(modin_df.index)
for df in [modin_df, pandas_df]:
df.index = [(i - length / 2) % length for i in range(length)]
# Add NaNs to sorted index
for df in [modin_df, pandas_df]:
sort_index = df.axes[axis]
create_test_dfs,
generate_multiindex,
test_data_diff_dtype,
)
from modin.config import NPartitions
NPartitions.put(4)
# Force matplotlib to not use any Xwindows backend.
matplotlib.use("Agg")
@pytest.mark.parametrize("method", ["all", "any"])
@pytest.mark.parametrize("is_transposed", [False, True])
@pytest.mark.parametrize(
"skipna", bool_arg_values, ids=arg_keys("skipna", bool_arg_keys)
)
@pytest.mark.parametrize("axis", axis_values, ids=axis_keys)
@pytest.mark.parametrize("data", [test_data["float_nan_data"]])
def test_all_any(data, axis, skipna, is_transposed, method):
eval_general(
*create_test_dfs(data),
lambda df: getattr((df.T if is_transposed else df), method)(
axis=axis, skipna=skipna, bool_only=None
),
)
@pytest.mark.parametrize("method", ["all", "any"])
@pytest.mark.parametrize(
"bool_only", bool_arg_values, ids=arg_keys("bool_only", bool_arg_keys)
# Named Series promoted to DF
s = pd.Series(frame_data2.get("col1"))
with pytest.raises(ValueError):
modin_df.merge(s)
s = pd.Series(frame_data2.get("col1"), name="col1")
df_equals(modin_df.merge(s), modin_df.merge(modin_df2[["col1"]]))
with pytest.raises(TypeError):
modin_df.merge("Non-valid type")
@pytest.mark.parametrize("axis", [0, 1])
@pytest.mark.parametrize(
"ascending", bool_arg_values, ids=arg_keys("ascending", bool_arg_keys)
)
@pytest.mark.parametrize("na_position", ["first", "last"], ids=["first", "last"])
def test_sort_index(axis, ascending, na_position):
data = test_data["float_nan_data"]
modin_df, pandas_df = pd.DataFrame(data), pandas.DataFrame(data)
# Change index value so sorting will actually make a difference
if axis == 0:
length = len(modin_df.index)
for df in [modin_df, pandas_df]:
df.index = [(i - length / 2) % length for i in range(length)]
# Add NaNs to sorted index
for df in [modin_df, pandas_df]:
sort_index = df.axes[axis]
int_arg_values,
test_data,
eval_general,
create_test_dfs,
test_data_diff_dtype,
)
pd.DEFAULT_NPARTITIONS = 4
# Force matplotlib to not use any Xwindows backend.
matplotlib.use("Agg")
@pytest.mark.parametrize("axis", [0, 1])
@pytest.mark.parametrize(
"skipna", bool_arg_values, ids=arg_keys("skipna", bool_arg_keys)
)
@pytest.mark.parametrize("method", ["cumprod", "cummin", "cummax", "cumsum"])
def test_cumprod_cummin_cummax_cumsum(axis, skipna, method):
eval_general(
*create_test_dfs(test_data["float_nan_data"]),
lambda df: getattr(df, method)(axis=axis, skipna=skipna),
)
@pytest.mark.parametrize("axis", ["rows", "columns"])
@pytest.mark.parametrize("method", ["cumprod", "cummin", "cummax", "cumsum"])
def test_cumprod_cummin_cummax_cumsum_transposed(axis, method):
eval_general(
*create_test_dfs(test_data["int_data"]),
lambda df: getattr(df.T, method)(axis=axis),
eval_general,
create_test_dfs,
test_data_diff_dtype,
)
from modin.config import NPartitions
NPartitions.put(4)
# Force matplotlib to not use any Xwindows backend.
matplotlib.use("Agg")
@pytest.mark.parametrize("axis", [0, 1])
@pytest.mark.parametrize("skipna",
bool_arg_values,
ids=arg_keys("skipna", bool_arg_keys))
@pytest.mark.parametrize("method", ["cumprod", "cummin", "cummax", "cumsum"])
def test_cumprod_cummin_cummax_cumsum(axis, skipna, method):
eval_general(
*create_test_dfs(test_data["float_nan_data"]),
lambda df: getattr(df, method)(axis=axis, skipna=skipna),
)
@pytest.mark.parametrize("axis", ["rows", "columns"])
@pytest.mark.parametrize("method", ["cumprod", "cummin", "cummax", "cumsum"])
def test_cumprod_cummin_cummax_cumsum_transposed(axis, method):
eval_general(
*create_test_dfs(test_data["int_data"]),
lambda df: getattr(df.T, method)(axis=axis),
)
# Named Series promoted to DF
s = pd.Series(frame_data2.get("col1"))
with pytest.raises(ValueError):
modin_df.merge(s)
s = pd.Series(frame_data2.get("col1"), name="col1")
df_equals(modin_df.merge(s), modin_df.merge(modin_df2[["col1"]]))
with pytest.raises(TypeError):
modin_df.merge("Non-valid type")
@pytest.mark.parametrize("axis", [0, 1])
@pytest.mark.parametrize(
"ascending", bool_arg_values, ids=arg_keys("ascending", bool_arg_keys)
)
@pytest.mark.parametrize("na_position", ["first", "last"], ids=["first", "last"])
def test_sort_index(axis, ascending, na_position):
data = test_data["float_nan_data"]
modin_df, pandas_df = pd.DataFrame(data), pandas.DataFrame(data)
# Change index value so sorting will actually make a difference
if axis == 0:
length = len(modin_df.index)
for df in [modin_df, pandas_df]:
df.index = [(i - length / 2) % length for i in range(length)]
# Add NaNs to sorted index
for df in [modin_df, pandas_df]:
sort_index = df.axes[axis]
