What is the typing.NamedTuple vs dataclass?

Question

Accepted Answer

**NamedTuple** is immutable, **dataclass** is mutable by default.

```python
from typing import NamedTuple
from dataclasses import dataclass

# NamedTuple — immutable
class Point(NamedTuple):
    x: float
    y: float

p = Point(1.0, 2.0)
print(p.x, p.y)  # 1.0 2.0
# p.x = 3.0  # AttributeError!

# Dataclass — mutable
@dataclass
class PointDC:
    x: float
    y: float

p = PointDC(1.0, 2.0)
p.x = 3.0  # OK

# Frozen dataclass — immutable
@dataclass(frozen=True)
class FrozenPoint:
    x: float
    y: float

# Comparison
Point(1, 2) == Point(1, 2)    # True (tuple comparison)
PointDC(1, 2) == PointDC(1, 2) # True (field comparison)

# Hashable
hash(Point(1, 2))  # OK — immutable
# hash(PointDC(1, 2))  # TypeError — unhashable
```

**When to use:**
- `NamedTuple`: immutable records, dict keys, sets.
- `dataclass`: mutable objects, complex logic.
- `frozen dataclass`: immutable with methods.

Python: What is the typing.NamedTuple vs dataclass?