[mypy] Add error codes (#4072)

py-polars/polars/convert.py

@@ -70,7 +70,7 @@ def from_dict(
     └─────┴─────┘
 
     """
-    return DataFrame._from_dict(data=data, columns=columns)  # type: ignore
+    return DataFrame._from_dict(data=data, columns=columns)  # type: ignore[arg-type]
 
 
 def from_dicts(

py-polars/polars/datatypes.py

@@ -12,7 +12,7 @@
 except ImportError:
     _PYARROW_AVAILABLE = False
 
-from _ctypes import _SimpleCData  # type: ignore
+from _ctypes import _SimpleCData  # type: ignore[import]
 
 try:
     from polars.polars import dtype_str_repr
@@ -27,7 +27,7 @@ class DataType:
     Base class for all Polars data types.
     """
 
-    def __new__(cls, *args, **kwargs) -> PolarsDataType:  # type: ignore
+    def __new__(cls, *args: Any, **kwargs: Any) -> PolarsDataType:  # type: ignore[misc]
         # this formulation allows for equivalent use of "pl.Type" and "pl.Type()", while
         # still respecting types that take initialisation params (eg: Duration/Datetime)
         if args or kwargs:
@@ -115,7 +115,7 @@ def __init__(self, inner: type[DataType]):
         """
         self.inner = py_type_to_dtype(inner)
 
-    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore
+    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore[override]
         # The comparison allows comparing objects to classes
         # and specific inner types to none specific.
         # if one of the arguments is not specific about its inner type
@@ -161,7 +161,7 @@ def __init__(self, time_unit: str = "us", time_zone: str | None = None):
         self.tu = time_unit
         self.tz = time_zone
 
-    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore
+    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore[override]
         # allow comparing object instances to class
         if type(other) is type and issubclass(other, Datetime):
             return True
@@ -188,7 +188,7 @@ def __init__(self, time_unit: str = "us"):
         """
         self.tu = time_unit
 
-    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore
+    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore[override]
         # allow comparing object instances to class
         if type(other) is type and issubclass(other, Duration):
             return True
@@ -228,7 +228,7 @@ def __init__(self, name: str, dtype: type[DataType]):
         self.name = name
         self.dtype = py_type_to_dtype(dtype)
 
-    def __eq__(self, other: Field) -> bool:  # type: ignore
+    def __eq__(self, other: Field) -> bool:  # type: ignore[override]
         return (self.name == other.name) & (self.dtype == other.dtype)
 
     def __repr__(self) -> str:
@@ -251,7 +251,7 @@ def __init__(self, fields: Sequence[Field]):
         """
         self.fields = fields
 
-    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore
+    def __eq__(self, other: type[DataType]) -> bool:  # type: ignore[override]
         # The comparison allows comparing objects to classes
         # and specific inner types to none specific.
         # if one of the arguments is not specific about its inner type

py-polars/polars/internals/anonymous_scan.py

@@ -47,7 +47,7 @@ def _scan_ds_impl(
     """
     if not _PYARROW_AVAILABLE:  # pragma: no cover
         raise ImportError("'pyarrow' is required for scanning from pyarrow datasets.")
-    return pl.from_arrow(ds.to_table(columns=with_columns))  # type: ignore
+    return pl.from_arrow(ds.to_table(columns=with_columns))  # type: ignore[return-value]
 
 
 def _scan_ds(ds: pa.dataset.dataset) -> pli.LazyFrame:

py-polars/polars/internals/construction.py

@@ -299,7 +299,7 @@ def _pandas_series_to_arrow(
     """
     dtype = values.dtype
     if dtype == "object" and len(values) > 0:
-        first_non_none = _get_first_non_none(values.values)  # type: ignore
+        first_non_none = _get_first_non_none(values.values)  # type: ignore[arg-type]
 
         if isinstance(first_non_none, str):
             return pa.array(values, pa.large_utf8(), from_pandas=nan_to_none)

py-polars/polars/internals/expr.py

@@ -4220,7 +4220,7 @@ def sign(self) -> Expr:
         """
         if not _NUMPY_AVAILABLE:
             raise ImportError("'numpy' is required for this functionality.")
-        return np.sign(self)  # type: ignore
+        return np.sign(self)  # type: ignore[call-overload]
 
     def sin(self) -> Expr:
         """
@@ -5173,7 +5173,7 @@ def concat(
         if not isinstance(other, list):
             other_list = [other]
         else:
-            other_list = copy.copy(other)  # type: ignore
+            other_list = copy.copy(other)  # type: ignore[arg-type]
 
         other_list.insert(0, wrap_expr(self._pyexpr))
         return pli.concat_list(other_list)

py-polars/polars/internals/frame.py

@@ -166,7 +166,7 @@ def __init__(cls, name: str, bases: tuple, clsdict: dict) -> None:
             # LazyFrame subclass by setting `cls._lazyframe_class`. We must therefore
             # dynamically create a subclass of LazyFrame with `_dataframe_class` set
             # to `cls` in order to preserve types after `.lazy().collect()` roundtrips.
-            cls._lazyframe_class = type(  # type: ignore
+            cls._lazyframe_class = type(  # type: ignore[assignment]
                 f"Lazy{name}",
                 (LazyFrame,),
                 {"_dataframe_class": cls},
@@ -289,14 +289,14 @@ class DataFrame(metaclass=DataFrameMetaClass):
 
     def __init__(
         self,
-        data: None
-        | (
+        data: (
             dict[str, Sequence[Any]]
             | Sequence[Any]
             | np.ndarray
             | pa.Table
             | pd.DataFrame
             | pli.Series
+            | None
         ) = None,
         columns: ColumnsType | None = None,
         orient: Literal["col", "row"] | None = None,
@@ -1519,10 +1519,10 @@ def to_numpy(self) -> np.ndarray:
         else:
             return out
 
-    def __getstate__(self):  # type: ignore
+    def __getstate__(self) -> list[pli.Series]:
         return self.get_columns()
 
-    def __setstate__(self, state):  # type: ignore
+    def __setstate__(self, state) -> None:  # type: ignore[no-untyped-def]
         self._df = DataFrame(state)._df
 
     def __mul__(self: DF, other: DataFrame | pli.Series | int | float | bool) -> DF:
@@ -1750,7 +1750,7 @@ def __getitem__(
 
         # df[:]
         if isinstance(item, slice):
-            return PolarsSlice(self).apply(item)  # type: ignore
+            return PolarsSlice(self).apply(item)  # type: ignore[return-value]
 
         # select rows by numpy mask or index
         # df[[1, 2, 3]]
@@ -1827,7 +1827,7 @@ def __setitem__(
             if isinstance(col_selection, str):
                 s = self.__getitem__(col_selection)
             elif isinstance(col_selection, int):
-                s = self[:, col_selection]  # type: ignore
+                s = self[:, col_selection]  # type: ignore[assignment]
             else:
                 raise ValueError(f"column selection not understood: {col_selection}")
 
@@ -2890,7 +2890,7 @@ def groupby(
             by = [by]
         return GroupBy(
             self._df,
-            by,  # type: ignore
+            by,  # type: ignore[arg-type]
             dataframe_class=self.__class__,
             maintain_order=maintain_order,
         )
@@ -4192,7 +4192,7 @@ def clone(self: DF) -> DF:
     def __copy__(self: DF) -> DF:
         return self.clone()
 
-    def __deepcopy__(self: DF, memodict={}) -> DF:  # type: ignore
+    def __deepcopy__(self: DF, memodict: dict = {}) -> DF:
         return self.clone()
 
     def get_columns(self) -> list[pli.Series]:
@@ -4904,7 +4904,7 @@ def select(
         """
         return (
             self.lazy()
-            .select(exprs)  # type: ignore
+            .select(exprs)  # type: ignore[arg-type]
             .collect(no_optimization=True, string_cache=False)
         )
 
@@ -6165,7 +6165,7 @@ def get_group(self, group_value: Any | tuple[Any]) -> DF:
 
         # should be only one match
         try:
-            groups_idx = groups[mask][0]  # type: ignore
+            groups_idx = groups[mask][0]  # type: ignore[index]
         except IndexError:
             raise ValueError(f"no group: {group_value} found")
 

py-polars/polars/internals/functions.py

@@ -140,7 +140,7 @@ def concat(
     elif isinstance(first, pli.Expr):
         out = first
         for e in items[1:]:
-            out = out.append(e)  # type: ignore
+            out = out.append(e)  # type: ignore[arg-type]
     else:
         raise ValueError(f"did not expect type: {type(first)} in 'pl.concat'.")
 

py-polars/polars/internals/lazy_frame.py

@@ -107,7 +107,7 @@ def _dataframe_class(self) -> type[DF]:
         This property is dynamically overwritten when DataFrame is sub-classed. See
         `polars.internals.frame.DataFrameMetaClass.__init__` for implementation details.
         """
-        return pli.DataFrame  # type: ignore
+        return pli.DataFrame  # type: ignore[return-value]
 
     @classmethod
     def scan_csv(
@@ -195,7 +195,7 @@ def scan_parquet(
                 scan = scan.head(n_rows)
             if row_count_name is not None:
                 scan = scan.with_row_count(row_count_name, row_count_offset)
-            return scan  # type: ignore
+            return scan  # type: ignore[return-value]
 
         self = cls.__new__(cls)
         self._ldf = PyLazyFrame.new_from_parquet(
@@ -235,7 +235,7 @@ def scan_ipc(
                 scan = scan.head(n_rows)
             if row_count_name is not None:
                 scan = scan.with_row_count(row_count_name, row_count_offset)
-            return scan  # type: ignore
+            return scan  # type: ignore[return-value]
 
         self = cls.__new__(cls)
         self._ldf = PyLazyFrame.new_from_ipc(
@@ -809,7 +809,7 @@ def clone(self: LDF) -> LDF:
     def __copy__(self: LDF) -> LDF:
         return self.clone()
 
-    def __deepcopy__(self: LDF, memodict={}) -> LDF:  # type: ignore
+    def __deepcopy__(self: LDF, memodict: dict = {}) -> LDF:
         return self.clone()
 
     def filter(self: LDF, predicate: pli.Expr | str) -> LDF:

py-polars/polars/internals/lazy_functions.py

@@ -157,7 +157,7 @@ def col(
 
     """
     if isinstance(name, pli.Series):
-        name = name.to_list()  # type: ignore
+        name = name.to_list()  # type: ignore[assignment]
 
     # note: we need the typing.cast call here twice to make mypy happy under Python 3.7
     # On Python 3.10, it is not needed. We use cast as it works across versions, ignoring
@@ -685,7 +685,7 @@ def lit(
     # numpy literals like np.float32(0)
     # have an item
     if hasattr(value, "item"):
-        value = value.item()  # type: ignore
+        value = value.item()  # type: ignore[union-attr]
     return pli.wrap_expr(pylit(value))
 
 
@@ -1697,7 +1697,7 @@ def repeat(
         if name is None:
             name = ""
         dtype = py_type_to_dtype(type(value))
-        s = pli.Series._repeat(name, value, n, dtype)  # type: ignore
+        s = pli.Series._repeat(name, value, n, dtype)  # type: ignore[arg-type]
         return s
     else:
         if isinstance(n, int):

py-polars/polars/internals/series.py

@@ -2286,7 +2286,7 @@ def to_numpy(
             kwargs will be sent to pyarrow.Array.to_numpy
         """
 
-        def convert_to_date(arr):  # type: ignore
+        def convert_to_date(arr):  # type: ignore[no-untyped-def]
             if self.dtype == Date:
                 tp = "datetime64[D]"
             elif self.dtype == Duration:
@@ -2591,7 +2591,7 @@ def sign(self) -> Series:
         """
         if not _NUMPY_AVAILABLE:
             raise ImportError("'numpy' is required for this functionality.")
-        return np.sign(self)  # type: ignore
+        return np.sign(self)  # type: ignore[return-value]
 
     def sin(self) -> Series:
         """
@@ -5423,13 +5423,13 @@ def min(self) -> date | datetime | timedelta:
         Return minimum as python DateTime
         """
         # we can ignore types because we are certain we get a logical type
-        return wrap_s(self._s).min()  # type: ignore
+        return wrap_s(self._s).min()  # type: ignore[return-value]
 
     def max(self) -> date | datetime | timedelta:
         """
         Return maximum as python DateTime
         """
-        return wrap_s(self._s).max()  # type: ignore
+        return wrap_s(self._s).max()  # type: ignore[return-value]
 
     def median(self) -> date | datetime | timedelta:
         """

py-polars/polars/io.py

@@ -1056,7 +1056,7 @@ def read_excel(
     """
 
     try:
-        import xlsx2csv  # type: ignore
+        import xlsx2csv  # type: ignore[import]
     except ImportError:
         raise ImportError(
             "xlsx2csv is not installed. Please run `pip install xlsx2csv`."

py-polars/polars/testing.py

@@ -137,8 +137,8 @@ def assert_frame_equal(
     # this does not assume a particular order
     for c in left.columns:
         _assert_series_inner(
-            left[c],  # type: ignore
-            right[c],  # type: ignore
+            left[c],  # type: ignore[arg-type, index]
+            right[c],  # type: ignore[arg-type, index]
             check_dtype,
             check_exact,
             nans_compare_equal,

py-polars/polars/utils.py

@@ -148,13 +148,13 @@ def handle_projection_columns(
     projection: list[int] | None = None
     if columns:
         if is_int_sequence(columns):
-            projection = columns  # type: ignore
+            projection = columns  # type: ignore[assignment]
             columns = None
         elif not is_str_sequence(columns):
             raise ValueError(
                 "columns arg should contain a list of all integers or all strings values."
             )
-    return projection, columns  # type: ignore
+    return projection, columns  # type: ignore[return-value]
 
 
 def _to_python_time(value: int) -> time:

py-polars/pyproject.toml

@@ -44,8 +44,7 @@ warn_redundant_casts = true
 enable_error_code = [
   "redundant-expr",
   "truthy-bool",
-  # TODO: Uncomment error below and fix mypy errors
-  # "ignore-without-code",
+  "ignore-without-code",
 ]
 
 [[tool.mypy.overrides]]

py-polars/tests/io/test_avro.py

@@ -21,7 +21,7 @@ def compressions() -> list[str]:
 def test_from_to_buffer(example_df: pl.DataFrame, compressions: list[str]) -> None:
     for compression in compressions:
         buf = io.BytesIO()
-        example_df.write_avro(buf, compression=compression)  # type: ignore
+        example_df.write_avro(buf, compression=compression)  # type: ignore[arg-type]
         buf.seek(0)
         read_df = pl.read_avro(buf)
         assert example_df.frame_equal(read_df)
@@ -33,7 +33,7 @@ def test_from_to_file(
     f = os.path.join(io_test_dir, "small.avro")
 
     for compression in compressions:
-        example_df.write_avro(f, compression=compression)  # type: ignore
+        example_df.write_avro(f, compression=compression)  # type: ignore[arg-type]
         df_read = pl.read_avro(str(f))
         assert example_df.frame_equal(df_read)