fix(maths): use specific exception types in collatz_sequence

[fauzan171]

Summary

Replace the generic Exception with appropriate specific exception types (TypeError and ValueError) in the collatz_sequence() input validation.

Changes

• Split the combined if not isinstance(n, int) or n < 1 check into two separate checks
• Raise TypeError when input is not an integer (e.g., float, string)
• Raise ValueError when input is an integer but not positive (e.g., 0, -1)
• Updated doctests to match the new exception types

Why

Using Exception is a Python anti-pattern. Specific exception types allow callers to handle different error cases appropriately:

• TypeError: wrong input type
• ValueError: correct type but invalid value

This follows the same pattern used by other functions in this repository (e.g., juggler_sequence(), liouville_lambda(), twin_prime()).

Checklist

• All doctests pass
• Consistent with exception handling patterns elsewhere in the repo