changed shape argument for ndarray from int to tuple in ./core/strings/object_array.py #44352
Conversation
|
pandas/core/strings/object_array.py:65: error: Argument "dtype" to "ndarray" has incompatible type "Union[ExtensionDtype, str, dtype[object_], dtype[Any], Type[object]]"; expected "Union[dtype[Any], None, type, _SupportsDType[dtype[Any]], str, Union[Tuple[Any, int], Tuple[Any, Union[SupportsIndex, Sequence[SupportsIndex]]], List[Any], _DTypeDict, Tuple[Any, Any]]]" [arg-type] |
pandas/core/strings/object_array.py
Outdated
| # error: Argument 1 to "ndarray" has incompatible type "int"; | ||
| # expected "Sequence[int]" | ||
| return np.ndarray(0, dtype=dtype) # type: ignore[arg-type] | ||
| return np.ndarray((0,), dtype=dtype) |
There was a problem hiding this comment.
best guess is this is intended to use np.array, not np.ndarray
There was a problem hiding this comment.
@jbrockmendel Thank you! I am new to this and appreciate the help. So you think calling np.array with similar arguments is needed instead of np.ndarray?
There was a problem hiding this comment.
@jbrockmendel also, I tried using pd.test() on my local machine and I didn't get this error message. Any advice on how to test before i push again?
There was a problem hiding this comment.
it should probably be np.array([], dtype=dtype).
I tried using pd.test() on my local machine and I didn't get this error message. Any advice on how to test before i push again?
Are you referring to the 'Argument 1 to "ndarray" ...'? That would show up if you a) removed the "type: ignore[arg-type]" and then ran 'mypy pandas'
There was a problem hiding this comment.
@jbrockmendel awesome, thank you! I will try this out.
There was a problem hiding this comment.
@jbrockmendel I am having trouble figuring out how to address the type errors for argument dtype. The suggestion you made about np.array instead of np.ndarray didn't solve the problem. Should I close this pr if I cannot figure it out?
There was a problem hiding this comment.
I am going to close this pull request for now. I am going to research the problem further, and come back to this once I have a good working solution. Thank you for helping me get started.
There was a problem hiding this comment.
what went wrong with np.array([], dtype=dtype)?
There was a problem hiding this comment.
same type error as with the np.ndarray
There was a problem hiding this comment.
looks like it is a similar error but now complaining about the dtype rather than the first arg
error: Argument "dtype" to "array" has incompatible type "Union[ExtensionDtype, str, dtype[object_], dtype[Any], Type[object]]"; expected "Union[dtype[Any], None, type, _SupportsDType[dtype[Any]], str, Union[Tuple[Any, int], Tuple[Any, Union[SupportsIndex, Sequence[SupportsIndex]]], List[Any], _DTypeDict, Tuple[Any, Any]]]" [arg-type]
this might be fixed by changing the annotaiton for 'dtype' from Dtype | None to NpDtype | None; @simonjayhawkins may know more than me about the intention here.
|
@jbrockmendel How are you today? Do you think these changes could work? |
|
lgtm cc @simonjayhawkins |
|
thanks @nickleus27 |
xref #37715