chore(ci): fix numpy type errors and revert #22610

setup.cfg

@@ -17,9 +17,9 @@
 [metadata]
 name = Superset
 summary = a data exploration platform
-description-file = README.md
+description_file = README.md
 author = Apache Superset Dev
-author-email = [email protected]
+author_email = [email protected]
 license = Apache License, Version 2.0
 
 [files]

superset/db_engine_specs/hive.py

@@ -206,7 +206,7 @@ def df_to_sql(
             with cls.get_engine(database) as engine:
                 engine.execute(f"DROP TABLE IF EXISTS {str(table)}")
 
-        def _get_hive_type(dtype: np.dtype) -> str:
+        def _get_hive_type(dtype: np.dtype[Any]) -> str:
             hive_type_by_dtype = {
                 np.dtype("bool"): "BOOLEAN",
                 np.dtype("float64"): "DOUBLE",

superset/reports/commands/alert.py

@@ -18,12 +18,13 @@
 import logging
 from operator import eq, ge, gt, le, lt, ne
 from timeit import default_timer
-from typing import Optional
+from typing import Any, Optional
 
 import numpy as np
 import pandas as pd
 from celery.exceptions import SoftTimeLimitExceeded
 from flask_babel import lazy_gettext as _
+from numpy.typing import NDArray
 
 from superset import app, jinja_context, security_manager
 from superset.commands.base import BaseCommand
@@ -84,12 +85,12 @@ def run(self) -> bool:
         except (KeyError, json.JSONDecodeError) as ex:
             raise AlertValidatorConfigError() from ex
 
-    def _validate_not_null(self, rows: np.recarray) -> None:
+    def _validate_not_null(self, rows: NDArray[Any]) -> None:
         self._validate_result(rows)
         self._result = rows[0][1]
 
     @staticmethod
-    def _validate_result(rows: np.recarray) -> None:
+    def _validate_result(rows: NDArray[Any]) -> None:
         # check if query return more than one row
         if len(rows) > 1:
             raise AlertQueryMultipleRowsError(
@@ -108,7 +109,7 @@ def _validate_result(rows: np.recarray) -> None:
                 )
             )
 
-    def _validate_operator(self, rows: np.recarray) -> None:
+    def _validate_operator(self, rows: NDArray[Any]) -> None:
         self._validate_result(rows)
         if rows[0][1] in (0, None, np.nan):
             self._result = 0.0

superset/result_set.py

@@ -24,6 +24,7 @@
 import numpy as np
 import pandas as pd
 import pyarrow as pa
+from numpy.typing import NDArray
 
 from superset.db_engine_specs import BaseEngineSpec
 from superset.superset_typing import DbapiDescription, DbapiResult, ResultSetColumnType
@@ -62,16 +63,16 @@ def stringify(obj: Any) -> str:
     return json.dumps(obj, default=utils.json_iso_dttm_ser)
 
 
-def stringify_values(array: np.ndarray) -> np.ndarray:
+def stringify_values(array: NDArray[Any]) -> NDArray[Any]:
     result = np.copy(array)
 
-    with np.nditer(result, flags=["refs_ok"], op_flags=["readwrite"]) as it:
+    with np.nditer(result, flags=["refs_ok"], op_flags=[["readwrite"]]) as it:
         for obj in it:
-            if pd.isna(obj):
+            if na_obj := pd.isna(obj):
                 # pandas <NA> type cannot be converted to string
-                obj[pd.isna(obj)] = None
+                obj[na_obj] = None  # type: ignore
             else:
-                obj[...] = stringify(obj)
+                obj[...] = stringify(obj)  # type: ignore
 
     return result
 
@@ -106,7 +107,7 @@ def __init__(  # pylint: disable=too-many-locals
         pa_data: List[pa.Array] = []
         deduped_cursor_desc: List[Tuple[Any, ...]] = []
         numpy_dtype: List[Tuple[str, ...]] = []
-        stringified_arr: np.ndarray
+        stringified_arr: NDArray[Any]
 
         if cursor_description:
             # get deduped list of column names
@@ -208,7 +209,7 @@ def convert_table_to_df(table: pa.Table) -> pd.DataFrame:
             return table.to_pandas(integer_object_nulls=True, timestamp_as_object=True)
 
     @staticmethod
-    def first_nonempty(items: List[Any]) -> Any:
+    def first_nonempty(items: NDArray[Any]) -> Any:
         return next((i for i in items if i), None)
 
     def is_temporal(self, db_type_str: Optional[str]) -> bool:

superset/utils/pandas_postprocessing/boxplot.py

@@ -57,10 +57,10 @@ def boxplot(
     """
 
     def quartile1(series: Series) -> float:
-        return np.nanpercentile(series, 25, interpolation="midpoint")
+        return np.nanpercentile(series, 25, interpolation="midpoint")  # type: ignore
 
     def quartile3(series: Series) -> float:
-        return np.nanpercentile(series, 75, interpolation="midpoint")
+        return np.nanpercentile(series, 75, interpolation="midpoint")  # type: ignore
 
     if whisker_type == PostProcessingBoxplotWhiskerType.TUKEY:
 
@@ -99,8 +99,8 @@ def whisker_low(series: Series) -> float:
             return np.nanpercentile(series, low)
 
     else:
-        whisker_high = np.max
-        whisker_low = np.min
+        whisker_high = np.max  # type: ignore
+        whisker_low = np.min  # type: ignore
 
     def outliers(series: Series) -> Set[float]:
         above = series[series > whisker_high(series)]
@@ -126,7 +126,7 @@ def outliers(series: Series) -> Set[float]:
     # nanpercentile needs numeric values, otherwise the isnan function
     # that's used in the underlying function will fail
     for column in metrics:
-        if df.dtypes[column] == np.object:
+        if df.dtypes[column] == np.object_:
             df[column] = to_numeric(df[column], errors="coerce")
 
     return aggregate(df, groupby=groupby, aggregates=aggregates)

superset/utils/pandas_postprocessing/flatten.py

@@ -85,7 +85,7 @@ def flatten(
         _columns = []
         for series in df.columns.to_flat_index():
             _cells = []
-            for cell in series if is_sequence(series) else [series]:
+            for cell in series if is_sequence(series) else [series]:  # type: ignore
                 if pd.notnull(cell):
                     # every cell should be converted to string and escape comma
                     _cells.append(escape_separator(str(cell)))

superset/utils/pandas_postprocessing/utils.py

@@ -24,7 +24,7 @@
 
 from superset.exceptions import InvalidPostProcessingError
 
-NUMPY_FUNCTIONS = {
+NUMPY_FUNCTIONS: Dict[str, Callable[..., Any]] = {
     "average": np.average,
     "argmin": np.argmin,
     "argmax": np.argmax,