def test_fillna_datetime_columns():
frame_data = {
"A": [-1, -2, np.nan],
"B": pd.date_range("20130101", periods=3),
"C": ["foo", "bar", None],
"D": ["foo2", "bar2", None],
}
df = pandas.DataFrame(frame_data,
index=pd.date_range("20130110", periods=3))
modin_df = pd.DataFrame(frame_data,
index=pd.date_range("20130110", periods=3))
df_equals(modin_df.fillna("?"), df.fillna("?"))
frame_data = {
"A": [-1, -2, np.nan],
"B": [
pandas.Timestamp("2013-01-01"),
pandas.Timestamp("2013-01-02"),
pandas.NaT,
],
"C": ["foo", "bar", None],
"D": ["foo2", "bar2", None],
}
df = pandas.DataFrame(frame_data,
index=pd.date_range("20130110", periods=3))
modin_df = pd.DataFrame(frame_data,
index=pd.date_range("20130110", periods=3))
df_equals(modin_df.fillna("?"), df.fillna("?"))
def test_tz_convert():
modin_idx = pd.date_range("1/1/2012",
periods=500,
freq="2D",
tz="America/Los_Angeles")
pandas_idx = pandas.date_range("1/1/2012",
periods=500,
freq="2D",
tz="America/Los_Angeles")
data = np.random.randint(0, 100, size=(len(modin_idx), 4))
modin_df = pd.DataFrame(data, index=modin_idx)
pandas_df = pandas.DataFrame(data, index=pandas_idx)
modin_result = modin_df.tz_convert("UTC", axis=0)
pandas_result = pandas_df.tz_convert("UTC", axis=0)
df_equals(modin_result, pandas_result)
modin_multi = pd.MultiIndex.from_arrays([modin_idx, range(len(modin_idx))])
pandas_multi = pandas.MultiIndex.from_arrays(
[pandas_idx, range(len(modin_idx))])
modin_series = pd.DataFrame(data, index=modin_multi)
pandas_series = pandas.DataFrame(data, index=pandas_multi)
df_equals(
modin_series.tz_convert("UTC", axis=0, level=0),
pandas_series.tz_convert("UTC", axis=0, level=0),
)
def test_asfreq():
index = pd.date_range("1/1/2000", periods=4, freq="T")
series = pd.Series([0.0, None, 2.0, 3.0], index=index)
df = pd.DataFrame({"s": series})
with pytest.warns(UserWarning):
# We are only testing that this defaults to pandas, so we will just check for
# the warning
df.asfreq(freq="30S")
def test_getitem_datetime_slice():
data = {"data": range(1000)}
index = pd.date_range("2017/1/4", periods=1000)
modin_df = pd.DataFrame(data=data, index=index)
pandas_df = pandas.DataFrame(data=data, index=index)
s = slice("2017-01-06", "2017-01-09")
df_equals(modin_df[s], pandas_df[s])
def test_first():
i = pd.date_range("2010-04-09", periods=400, freq="2D")
modin_df = pd.DataFrame({"A": list(range(400)), "B": list(range(400))}, index=i)
pandas_df = pandas.DataFrame(
{"A": list(range(400)), "B": list(range(400))}, index=i
)
df_equals(modin_df.first("3D"), pandas_df.first("3D"))
df_equals(modin_df.first("20D"), pandas_df.first("20D"))
def test_at_time():
i = pd.date_range("2008-01-01", periods=1000, freq="12H")
modin_df = pd.DataFrame({"A": list(range(1000)), "B": list(range(1000))}, index=i)
pandas_df = pandas.DataFrame(
{"A": list(range(1000)), "B": list(range(1000))}, index=i
)
df_equals(modin_df.at_time("12:00"), pandas_df.at_time("12:00"))
df_equals(modin_df.at_time("3:00"), pandas_df.at_time("3:00"))
df_equals(modin_df.T.at_time("12:00", axis=1), pandas_df.T.at_time("12:00", axis=1))
def test_tz_localize():
idx = pd.date_range("1/1/2012", periods=400, freq="2D")
data = np.random.randint(0, 100, size=(len(idx), 4))
modin_df = pd.DataFrame(data, index=idx)
pandas_df = pandas.DataFrame(data, index=idx)
df_equals(modin_df.tz_localize("UTC", axis=0), pandas_df.tz_localize("UTC", axis=0))
df_equals(
modin_df.tz_localize("America/Los_Angeles", axis=0),
pandas_df.tz_localize("America/Los_Angeles", axis=0),
)
def test_dataframe_dt_index(axis, on, closed, window):
index = pandas.date_range("31/12/2000", periods=12, freq="T")
data = {"A": range(12), "B": range(12)}
pandas_df = pandas.DataFrame(data, index=index)
modin_df = pd.DataFrame(data, index=index)
if on is not None and axis == 0 and isinstance(window, str):
pandas_df[on] = pandas.date_range("22/06/1941", periods=12, freq="T")
modin_df[on] = pd.date_range("22/06/1941", periods=12, freq="T")
else:
on = None
if axis == "columns":
pandas_df = pandas_df.T
modin_df = modin_df.T
pandas_rolled = pandas_df.rolling(window=window,
on=on,
axis=axis,
closed=closed)
modin_rolled = modin_df.rolling(window=window,
on=on,
axis=axis,
closed=closed)
if isinstance(window, int):
# This functions are very slowly for data from test_rolling
df_equals(modin_rolled.corr(modin_df, True),
pandas_rolled.corr(pandas_df, True))
df_equals(modin_rolled.corr(modin_df, False),
pandas_rolled.corr(pandas_df, False))
df_equals(modin_rolled.cov(modin_df, True),
pandas_rolled.cov(pandas_df, True))
df_equals(modin_rolled.cov(modin_df, False),
pandas_rolled.cov(pandas_df, False))
if axis == 0:
df_equals(
modin_rolled.cov(modin_df[modin_df.columns[0]], True),
pandas_rolled.cov(pandas_df[pandas_df.columns[0]], True),
)
df_equals(
modin_rolled.corr(modin_df[modin_df.columns[0]], True),
pandas_rolled.corr(pandas_df[pandas_df.columns[0]], True),
)
else:
df_equals(modin_rolled.skew(), pandas_rolled.skew())
df_equals(
modin_rolled.apply(np.sum, raw=True),
pandas_rolled.apply(np.sum, raw=True),
)
df_equals(modin_rolled.aggregate(np.sum),
pandas_rolled.aggregate(np.sum))
df_equals(modin_rolled.quantile(0.1), pandas_rolled.quantile(0.1))
# `Rolling.counts` has a buggy side-effect on other rolling functions described in:
# https://github.com/pandas-dev/pandas/issues/39554
# So the testing of `.count` should always be the last until this bug
# will be fixed in pandas, to avoid this side-effect
df_equals(modin_rolled.count(), pandas_rolled.count())
def test_dataframe_dt_index(axis, on, closed, window):
index = pandas.date_range("31/12/2000", periods=12, freq="T")
data = {"A": range(12), "B": range(12)}
pandas_df = pandas.DataFrame(data, index=index)
modin_df = pd.DataFrame(data, index=index)
if on is not None and axis == 0 and isinstance(window, str):
pandas_df[on] = pandas.date_range("22/06/1941", periods=12, freq="T")
modin_df[on] = pd.date_range("22/06/1941", periods=12, freq="T")
else:
on = None
if axis == "columns":
pandas_df = pandas_df.T
modin_df = modin_df.T
pandas_rolled = pandas_df.rolling(window=window,
on=on,
axis=axis,
closed=closed)
modin_rolled = modin_df.rolling(window=window,
on=on,
axis=axis,
closed=closed)
if isinstance(window, int):
# This functions are very slowly for data from test_rolling
df_equals(modin_rolled.corr(modin_df, True),
pandas_rolled.corr(pandas_df, True))
df_equals(modin_rolled.corr(modin_df, False),
pandas_rolled.corr(pandas_df, False))
df_equals(modin_rolled.cov(modin_df, True),
pandas_rolled.cov(pandas_df, True))
df_equals(modin_rolled.cov(modin_df, False),
pandas_rolled.cov(pandas_df, False))
if axis == 0:
df_equals(
modin_rolled.cov(modin_df[modin_df.columns[0]], True),
pandas_rolled.cov(pandas_df[pandas_df.columns[0]], True),
)
df_equals(
modin_rolled.corr(modin_df[modin_df.columns[0]], True),
pandas_rolled.corr(pandas_df[pandas_df.columns[0]], True),
)
else:
df_equals(modin_rolled.count(), pandas_rolled.count())
df_equals(modin_rolled.skew(), pandas_rolled.skew())
df_equals(
modin_rolled.apply(np.sum, raw=True),
pandas_rolled.apply(np.sum, raw=True),
)
df_equals(modin_rolled.aggregate(np.sum),
pandas_rolled.aggregate(np.sum))
df_equals(modin_rolled.quantile(0.1), pandas_rolled.quantile(0.1))
def test_reindex_like():
df1 = pd.DataFrame(
[
[24.3, 75.7, "high"],
[31, 87.8, "high"],
[22, 71.6, "medium"],
[35, 95, "medium"],
],
columns=["temp_celsius", "temp_fahrenheit", "windspeed"],
index=pd.date_range(start="2014-02-12", end="2014-02-15", freq="D"),
)
df2 = pd.DataFrame(
[[28, "low"], [30, "low"], [35.1, "medium"]],
columns=["temp_celsius", "windspeed"],
index=pd.DatetimeIndex(["2014-02-12", "2014-02-13", "2014-02-15"]),
)
with pytest.warns(UserWarning):
df2.reindex_like(df1)
def test_to_timestamp():
idx = pd.date_range("1/1/2012", periods=5, freq="M")
df = pd.DataFrame(np.random.randint(0, 100, size=(len(idx), 4)), index=idx)
with pytest.warns(UserWarning):
df.to_period().to_timestamp()
def test_tshift():
idx = pd.date_range("1/1/2012", periods=5, freq="M")
data = np.random.randint(0, 100, size=(len(idx), 4))
modin_df = pd.DataFrame(data, index=idx)
pandas_df = pandas.DataFrame(data, index=idx)
df_equals(modin_df.tshift(4), pandas_df.tshift(4))
def test_to_timestamp():
idx = pd.date_range("1/1/2012", periods=5, freq="M")
df = pd.DataFrame(np.random.randint(0, 100, size=(len(idx), 4)), index=idx)
with warns_that_defaulting_to_pandas():
df.to_period().to_timestamp()
