def test_from_pandas_datetime():
df = pd.DataFrame({
"datetime": ["2021-01-01", "2021-01-02"],
"foo": [1, 2]
})
df["datetime"] = pd.to_datetime(df["datetime"])
pl.from_pandas(df)
def test_from_pandas_dataframe() -> None:
pd_df = pd.DataFrame([[1, 2, 3], [4, 5, 6]], columns=["a", "b", "c"])
df = pl.from_pandas(pd_df)
assert df.shape == (2, 3)
# if not a pandas dataframe, raise a ValueError
with pytest.raises(ValueError):
_ = pl.from_pandas([1, 2]) # type: ignore
def test_from_pandas_nan_to_none() -> None:
from pyarrow import ArrowInvalid
df = pd.Series([2, np.nan, None], name="pd")
out_true = pl.from_pandas(df)
out_false = pl.from_pandas(df, nan_to_none=False)
df.loc[2] = pd.NA
assert [val is None for val in out_true]
assert [np.isnan(val) for val in out_false[1:]]
with pytest.raises(ArrowInvalid, match="Could not convert"):
pl.from_pandas(df, nan_to_none=False)
def test_from_optional_not_available() -> None:
with patch("polars.convert._NUMPY_AVAILABLE", False):
with pytest.raises(ImportError):
pl.from_numpy(np.array([[1, 2], [3, 4]]), columns=["a", "b"])
with patch("polars.convert._PYARROW_AVAILABLE", False):
with pytest.raises(ImportError):
pl.from_arrow(pa.table({"a": [1, 2, 3], "b": [4, 5, 6]}))
with pytest.raises(ImportError):
pl.from_pandas(pd.Series([1, 2, 3]))
with patch("polars.convert._PANDAS_AVAILABLE", False):
with pytest.raises(ImportError):
pl.from_pandas(pd.Series([1, 2, 3]))
def test_from_pandas_nan_to_none() -> None:
from pyarrow import ArrowInvalid
df = pd.DataFrame({
"bools_nulls": [None, True, False],
"int_nulls": [1, None, 3],
"floats_nulls": [1.0, None, 3.0],
"strings_nulls": ["foo", None, "ham"],
"nulls": [None, np.nan, np.nan],
})
out_true = pl.from_pandas(df)
out_false = pl.from_pandas(df, nan_to_none=False)
df.loc[2, "nulls"] = pd.NA
assert all(val is None for val in out_true["nulls"])
assert all(np.isnan(val) for val in out_false["nulls"][1:])
with pytest.raises(ArrowInvalid, match="Could not convert"):
pl.from_pandas(df, nan_to_none=False)
df = pd.Series([2, np.nan, None], name="pd") # type: ignore
out_true = pl.from_pandas(df)
out_false = pl.from_pandas(df, nan_to_none=False)
df.loc[2] = pd.NA
assert [val is None for val in out_true]
assert [np.isnan(val) for val in out_false[1:]]
with pytest.raises(ArrowInvalid, match="Could not convert"):
pl.from_pandas(df, nan_to_none=False)
def test_from_pandas_nested_list() -> None:
# this panicked in https://github.com/pola-rs/polars/issues/1615
pddf = pd.DataFrame(
{"a": [1, 2, 3, 4], "b": [["x", "y"], ["x", "y", "z"], ["x"], ["x", "y"]]}
)
pldf = pl.from_pandas(pddf)
assert pldf.shape == (4, 2)
def test_from_pandas_datetime():
ts = datetime.datetime(2021, 1, 1, 20, 20, 20, 20)
s = pd.Series([ts, ts])
s = pl.from_pandas(s.to_frame("a"))["a"]
assert s.hour()[0] == 20
assert s.minute()[0] == 20
assert s.second()[0] == 20
def build_gene_annotation_df(pset_dict):
"""
Build a table mapping each gene in a dataset to its gene annotations.
@param pset_dict: [`dict`] A nested dictionary containing all tables in the PSet
@return: [`DataFrame`] A table of all gene annotations, mapped to genes
"""
# Extract the all molecular data types for the pSet
df_list = [
pl.from_pandas(pset_dict['molecularProfiles'][mDataType]['rowData'])
for mDataType in pset_dict['molecularProfiles']
]
# Get columns of interest, add columns needed later
for i in range(len(df_list)):
df_list[i] = df_list[i].select(['.features'])
empty_column = [None for _ in range(len(df_list[i]['.features']))]
df_list[i]['symbol'] = pl.Series('symbol', empty_column, dtype=pl.Utf8)
df_list[i]['gene_seq_start'] = pl.Series('gene_seq_start',
empty_column,
dtype=pl.Int64)
df_list[i]['gene_seq_end'] = pl.Series('gene_seq_end',
empty_column,
dtype=pl.Int64)
# Merge to a single DataFrame
gene_annotation_df = pl.concat(df_list) \
.rename({'.features': 'gene_id'})
# Remove Ensembl gene version
gene_annotation_df['gene_id'] = gene_annotation_df['gene_id'] \
.apply(lambda x: re.sub(r'\..*$', '', x))
gene_annotation_df = gene_annotation_df \
.drop_duplicates() \
.to_pandas()
return gene_annotation_df
def test_struct_to_pandas() -> None:
df = pd.DataFrame([{"a": {"b": {"c": 2}}}])
pl_df = pl.from_pandas(df)
assert isinstance(pl_df.dtypes[0], pl.datatypes.Struct)
assert pl_df.to_pandas().equals(df)
def test_from_empty_pandas() -> None:
pandas_df = pd.DataFrame({
"A": [],
"fruits": [],
})
polars_df = pl.from_pandas(pandas_df)
assert polars_df.columns == ["A", "fruits"]
assert polars_df.dtypes == [pl.Float64, pl.Float64]
def test_from_pandas_ns_resolution() -> None:
df = pd.DataFrame(
[
pd.Timestamp(
year=2021, month=1, day=1, hour=1, second=1, nanosecond=1)
],
columns=["date"],
)
assert pl.from_pandas(df)[0, 0] == datetime(2021, 1, 1, 1, 0, 1)
def test_from_pandas_datetime() -> None:
ts = datetime(2021, 1, 1, 20, 20, 20, 20)
pl_s = pd.Series([ts, ts])
tmp = pl.from_pandas(pl_s.to_frame("a"))
s = tmp["a"]
assert s.dt.hour()[0] == 20
assert s.dt.minute()[0] == 20
assert s.dt.second()[0] == 20
date_times = pd.date_range(
"2021-06-24 00:00:00", "2021-06-24 10:00:00", freq="1H", closed="left"
)
s = pl.from_pandas(date_times)
assert s[0] == datetime(2021, 6, 24, 0, 0)
assert s[-1] == datetime(2021, 6, 24, 9, 0)
df = pd.DataFrame({"datetime": ["2021-01-01", "2021-01-02"], "foo": [1, 2]})
df["datetime"] = pd.to_datetime(df["datetime"])
pl.from_pandas(df)
def test_cast_inner() -> None:
a = pl.Series([[1, 2]])
for t in [bool, pl.Boolean]:
b = a.cast(pl.List(t))
assert b.dtype == pl.List(pl.Boolean)
assert b.to_list() == [[True, True]]
# this creates an inner null type
df = pl.from_pandas(pd.DataFrame(data=[[[]], [[]]], columns=["A"]))
assert df["A"].cast(pl.List(int)).dtype.inner == pl.Int64 # type: ignore[arg-type, attr-defined]
def test_from_pandas_datetime():
ts = datetime.datetime(2021, 1, 1, 20, 20, 20, 20)
s = pd.Series([ts, ts])
s = pl.from_pandas(s.to_frame("a"))["a"]
assert s.dt.hour()[0] == 20
assert s.dt.minute()[0] == 20
assert s.dt.second()[0] == 20
date_times = pd.date_range("2021-06-24 00:00:00",
"2021-06-24 10:00:00",
freq="1H",
closed="left")
s = pl.from_pandas(date_times)
assert s[0] == 1624492800000
assert s[-1] == 1624525200000
# checks dispatch
s.dt.round("hour", 2)
s.dt.round("day", 5)
# checks lazy dispatch
pl.DataFrame([s.rename("foo")])[pl.col("foo").dt.round("hour", 2)]
def test_join_dates():
date_times = pd.date_range("2021-06-24 00:00:00",
"2021-06-24 10:00:00",
freq="1H",
closed="left")
dts = (pl.from_pandas(date_times).apply(
lambda x: x + np.random.randint(1_000 * 60, 60_000 * 60)).cast(
pl.Date64))
# some df with sensor id, (randomish) datetime and some value
df = pl.DataFrame({
"sensor": ["a"] * 5 + ["b"] * 5,
"datetime": dts,
"value": [2, 3, 4, 1, 2, 3, 5, 1, 2, 3],
})
df.join(df, on="datetime")
def test_from_pandas():
df = pd.DataFrame({
"bools": [False, True, False],
"bools_nulls": [None, True, False],
"int": [1, 2, 3],
"int_nulls": [1, None, 3],
"floats": [1.0, 2.0, 3.0],
"floats_nulls": [1.0, None, 3.0],
"strings": ["foo", "bar", "ham"],
"strings_nulls": ["foo", None, "ham"],
"strings-cat": ["foo", "bar", "ham"],
})
df["strings-cat"] = df["strings-cat"].astype("category")
out = pl.from_pandas(df)
assert out.shape == (3, 9)
def test_struct_logical_types_to_pandas() -> None:
timestamp = datetime(2022, 1, 1)
df = pd.DataFrame([{"struct": {"timestamp": timestamp}}])
assert pl.from_pandas(df).dtypes == [pl.Struct]
def test_from_null_column() -> None:
assert pl.from_pandas(pd.DataFrame(data=[pd.NA, pd.NA])).shape == (2, 1)
def test_from_empty_pandas_strings() -> None:
df = pd.DataFrame(columns=["a", "b"])
df["a"] = df["a"].astype(str)
df["b"] = df["b"].astype(float)
df_pl = pl.from_pandas(df)
assert df_pl.dtypes == [pl.Utf8, pl.Float64]
def test_from_pandas_series() -> None:
pd_series = pd.Series([1, 2, 3], name="pd")
df = pl.from_pandas(pd_series)
assert df.shape == (3, )
def test_from_pandas_categorical_none() -> None:
s = pd.Series(["a", "b", "c", pd.NA], dtype="category")
out = pl.from_pandas(s)
assert out.dtype == pl.Categorical
assert out.to_list() == ["a", "b", "c", None]