handling macos13 platform specific error

Author: Koookadooo

pandas/core/dtypes/dtypes.py

@@ -2274,26 +2274,31 @@ def name(self) -> str:  # type: ignore[override]
 
     @cache_readonly
     def numpy_dtype(self) -> np.dtype:
-        """Return an instance of the related numpy dtype"""
-        if pa.types.is_timestamp(self.pyarrow_dtype):
-            if self.pyarrow_dtype.tz is not None:
-                # Handle timezone-aware timestamps
-                return np.dtype(f"datetime64[{self.pyarrow_dtype.unit}]")
+    """Return an instance of the related numpy dtype."""
+    if pa.types.is_timestamp(self.pyarrow_dtype):
+        if self.pyarrow_dtype.tz is not None:
             return np.dtype(f"datetime64[{self.pyarrow_dtype.unit}]")
-        if pa.types.is_duration(self.pyarrow_dtype):
-            return np.dtype(f"timedelta64[{self.pyarrow_dtype.unit}]")
-        if pa.types.is_string(self.pyarrow_dtype) or pa.types.is_large_string(
-            self.pyarrow_dtype
-        ):
-            return np.dtype(str)
-        try:
-            np_dtype = self.pyarrow_dtype.to_pandas_dtype()
-            if isinstance(np_dtype, DatetimeTZDtype):
-                # Convert timezone-aware to naive datetime64
-                return np.dtype(f"datetime64[{np_dtype.unit}]")
-            return np.dtype(np_dtype)
-        except (NotImplementedError, TypeError):
-            return np.dtype(object)
+        return np.dtype(f"datetime64[{self.pyarrow_dtype.unit}]")
+    if pa.types.is_duration(self.pyarrow_dtype):
+        return np.dtype(f"timedelta64[{self.pyarrow_dtype.unit}]")
+    if pa.types.is_string(self.pyarrow_dtype) or pa.types.is_large_string(
+        self.pyarrow_dtype
+    ):
+        return np.dtype(str)
+    try:
+        np_dtype = self.pyarrow_dtype.to_pandas_dtype()
+        
+        if isinstance(np_dtype, object):
+            if hasattr(np_dtype, "categories") and isinstance(np_dtype.categories, pd.IntervalIndex):
+                return np.dtype(object)
+        
+        if isinstance(np_dtype, DatetimeTZDtype):
+            return np.dtype(f"datetime64[{np_dtype.unit}]")
+        
+        return np.dtype(np_dtype)
+    except (NotImplementedError, TypeError):
+        return np.dtype(object)
+
 
 
     @cache_readonly