Error Messages

Pavise provides detailed error messages to help you quickly identify and fix validation issues.

Type Errors

When a column has the wrong type, Pavise shows:

Expected type and actual type
Sample invalid values (first 5 maximum)
Row numbers for each invalid value
Actual type of each invalid value

Example

from typing import Protocol
from pavise.pandas import DataFrame
from pavise.exceptions import ValidationError
import pandas as pd

class Schema(Protocol):
    age: int

df = pd.DataFrame({"age": [25, "invalid", 30, None, 35, "bad", 40]})
try:
    validated_df = DataFrame[Schema](df)
except ValidationError as e:
    print(e)

Output:

Column 'age': expected int, got object

Sample invalid values (showing first 3 of 4):
  Row 1: 'invalid' (str)
  Row 3: None (NoneType)
  Row 5: 'bad' (str)

Missing Columns

When a required column is missing:

class Schema(Protocol):
    user_id: int
    name: str

df = pd.DataFrame({"user_id": [1, 2, 3]})  # Missing 'name'
try:
    validated_df = DataFrame[Schema](df)
except ValidationError as e:
    print(e)

Output:

Column 'name': missing

Validator Errors

Each validator provides detailed error messages.

Range Validator

from typing import Annotated
from pavise.validators import Range

class Schema(Protocol):
    age: Annotated[int, Range(0, 150)]

df = pd.DataFrame({"age": [25, 200, 30, -5, 35, 300]})
try:
    validated_df = DataFrame[Schema](df)
except ValidationError as e:
    print(e)

Output:

Column 'age': values must be in range [0, 150]

Sample invalid values (showing first 3 of 4):
  Row 1: 200
  Row 3: -5
  Row 5: 300

Unique Validator

from pavise.validators import Unique

class Schema(Protocol):
    user_id: Annotated[int, Unique()]

df = pd.DataFrame({"user_id": [1, 2, 2, 3, 5, 5, 5]})
try:
    validated_df = DataFrame[Schema](df)
except ValidationError as e:
    print(e)

Output:

Column 'user_id': contains duplicate values

Sample duplicate values (showing first 2):
  Value 2 at rows: [1, 2]
  Value 5 at rows: [4, 5, 6]

In Validator

from pavise.validators import In

class Schema(Protocol):
    status: Annotated[str, In(["pending", "approved", "rejected"])]

df = pd.DataFrame({"status": ["pending", "invalid", "approved", "bad"]})
try:
    validated_df = DataFrame[Schema](df)
except ValidationError as e:
    print(e)

Output:

Column 'status': contains values not in allowed values

Sample invalid values (showing first 2 of 2):
  Row 1: 'invalid'
  Row 3: 'bad'

Regex Validator

from pavise.validators import Regex

class Schema(Protocol):
    email: Annotated[str, Regex(r'^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}$')]

df = pd.DataFrame({"email": ["alice@example.com", "invalid", "bob@test.com", "bad@"]})
try:
    validated_df = DataFrame[Schema](df)
except ValidationError as e:
    print(e)

Output:

Column 'email': contains values that don't match the pattern

Sample invalid values (showing first 2 of 2):
  Row 1: 'invalid'
  Row 3: 'bad@'

MinLen/MaxLen Validators

from pavise.validators import MinLen

class Schema(Protocol):
    username: Annotated[str, MinLen(3)]

df = pd.DataFrame({"username": ["alice", "ab", "bob", "x"]})
try:
    validated_df = DataFrame[Schema](df)
except ValidationError as e:
    print(e)

Output:

Column 'username': contains strings shorter than minimum length

Sample invalid values (showing first 2 of 2):
  Row 1: 'ab' (length: 2)
  Row 3: 'x' (length: 1)

Strict Mode Errors

When strict mode is enabled and extra columns are present:

class Schema(Protocol):
    user_id: int
    name: str

df = pd.DataFrame({
    "user_id": [1, 2, 3],
    "name": ["Alice", "Bob", "Charlie"],
    "age": [25, 30, 35],  # Extra column
    "email": ["a@test.com", "b@test.com", "c@test.com"]  # Extra column
})

try:
    validated_df = DataFrame[Schema](df, strict=True)
except ValidationError as e:
    print(e)

Output:

Strict mode: unexpected columns ['age', 'email']

Performance Notes

To avoid overwhelming output and maintain performance:

Type error checking examines only the first 100 rows
Error messages show at most 5 sample invalid values
Duplicate detection shows at most 5 duplicate value groups

For large DataFrames, consider sampling before validation during development.