TypeError: __init__() got an unexpected keyword argument (Python) — how to fix
Traceback (most recent call last):
File "app.py", line 12, in <module>
user = User(name="Ana", emial="ana@example.com")
TypeError: User.__init__() got an unexpected keyword argument 'emial'
Quick fix: Check the real signature: import inspect; print(inspect.signature(MyClass)). Fix the spelling, or match the installed library version's API (pip show <pkg>; pin the old version or rename the argument as the changelog says). In subclasses, accept and forward extra arguments: def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs).
What it means
Python matched each name=value you passed against the parameters of the function. One name had no matching parameter and the function has no **kwargs to catch it. The message names the culprit: 'emial' above. Python 3.10+ prefixes the qualified name (User.__init__()), which tells you exactly which class is complaining — it may not be the one you expect.
See the actual signature:
import inspect
print(inspect.signature(User)) # constructor parameters
help(User)
Cause 1: typo or wrong name
User(name="Ana", emial="ana@example.com") # typo
User(name="Ana", email="ana@example.com") # fixed
Also common: camelCase vs snake_case (userId vs user_id), singular vs plural.
Cause 2: library API changed between versions
The code (or tutorial) targets a different version than you have installed. Examples:
requests/urllib3:Retry(method_whitelist=...)→allowed_methods=...(urllib3 2.0 removed the old name).- SQLAlchemy 2.0:
create_engine(..., convert_unicode=True)was removed. - pandas:
read_csv(error_bad_lines=False)→on_bad_lines="skip"(removed in 2.0). - scikit-learn:
OneHotEncoder(sparse=False)→sparse_output=False(removed in 1.4). - Pydantic v2, OpenAI v1, LangChain: many renamed arguments.
Check your version and the changelog:
pip show urllib3
Then either update the argument name or pin the version the code was written for (pip install "pandas<2"). A mismatch between two libraries (one calling the other with an argument the installed version doesn't know) is fixed by upgrading or pinning them together.
Cause 3: subclass doesn't forward arguments
class Base:
def __init__(self, name, debug=False): ...
class Child(Base):
def __init__(self, name):
super().__init__(name)
Child("x", debug=True) # TypeError: Child.__init__() got an unexpected keyword argument 'debug'
Forward what you don't handle:
class Child(Base):
def __init__(self, name, **kwargs):
super().__init__(name, **kwargs)
Cause 4: dataclasses and field definitions
Only annotated class attributes become dataclass fields:
from dataclasses import dataclass, field
@dataclass
class Config:
host: str
port = 8080 # no annotation: NOT a field, so Config(port=1) fails
tags: list = field(default_factory=list)
Write port: int = 8080. Fields declared with field(init=False) also can't be passed to the constructor.
Cause 5: decorators that hide the signature
A decorator that returns a wrapper with a fixed signature swallows keyword arguments. Use functools.wraps and pass *args, **kwargs through:
import functools
def log(fn):
@functools.wraps(fn)
def wrapper(*args, **kwargs):
return fn(*args, **kwargs)
return wrapper
Cause 6: unpacking a dict with extra keys
User(**data) fails if data (e.g. JSON from an API) contains keys the class doesn't define. Filter first:
import inspect
allowed = inspect.signature(User).parameters
user = User(**{k: v for k, v in data.items() if k in allowed})
Or use a model library that ignores extra fields (Pydantic's default behaviour).
Last updated 2026-11-09.