TheAlgorithms/Python · error · ValueError

There's an empty parameter

Error message

There's an empty parameter

What it means

Raised by viterbi()'s non-empty validation when any of its five parameters — observations_space, states_space, initial_probabilities, transition_probabilities, emission_probabilities — is empty (falsy). The Viterbi decoder needs at least one state, one observation symbol, and complete probability tables to run, so an empty any of these makes decoding undefined.

Source

Thrown at dynamic_programming/viterbi.py:245

    >>> _validate_not_empty(["a"], ["b"], {"c":0.5}, {}, {"f": {"g": 0.7}})
    Traceback (most recent call last):
            ...
    ValueError: There's an empty parameter
    >>> _validate_not_empty(["a"], ["b"], None, {"d": {"e": 0.6}}, {"f": {"g": 0.7}})
    Traceback (most recent call last):
            ...
    ValueError: There's an empty parameter
    """
    if not all(
        [
            observations_space,
            states_space,
            initial_probabilities,
            transition_probabilities,
            emission_probabilities,
        ]
    ):
        raise ValueError("There's an empty parameter")


def _validate_lists(observations_space: Any, states_space: Any) -> None:
    """
    >>> _validate_lists(["a"], ["b"])
    >>> _validate_lists(1234, ["b"])
    Traceback (most recent call last):
            ...
    ValueError: observations_space must be a list
    >>> _validate_lists(["a"], [3])
    Traceback (most recent call last):
            ...
    ValueError: states_space must be a list of strings
    """
    _validate_list(observations_space, "observations_space")
    _validate_list(states_space, "states_space")

View on GitHub (pinned to f5988cc097)

Solutions

1. Check all five structures are non-empty before calling viterbi and log which one is empty.
2. Fix the parameter loading code to fail loudly when a config section (states, transitions, emissions) is absent.
3. If the observation sequence is legitimately empty, short-circuit in your caller instead of calling viterbi.

Example fix

# before
viterbi([], states, initial_p, trans_p, emit_p)  # ValueError

# after
if not observations:
    result = []  # nothing to decode
else:
    result = viterbi(observations, states, initial_p, trans_p, emit_p)

def valid_hmm_args(obs, states, init_p, trans_p, emit_p) -> bool:
    return all([obs, states, init_p, trans_p, emit_p])

try:
    viterbi(obs, states, init_p, trans_p, emit_p)
except ValueError as e:
    if 'empty parameter' in str(e):
        raise RuntimeError(f'HMM config incomplete: {e}') from e
    raise

When it happens

Trigger: Calling viterbi(observation, [], initial_prob, transition_prob, emission_prob) with an empty states list; passing {} for transition_probabilities; passing '' or [] for observations_space. Note: truthiness is used, so an empty dict/list/tuple all trigger it.

Common situations: Loading HMM parameters from a JSON/YAML file where a section is missing and deserializes to {}; empty observation sequence after a filter step removes all events; initializing tables programmatically and forgetting to populate them.

Related errors

• Limit for the Catalan sequence must be ≥ 0
• Negative arguments are not supported
• iterations must be defined as integers
• starting number must be and integer
• Iterations must be done more than 0 times to play FizzBuzz

AI-assisted analysis of TheAlgorithms/Python@f5988cc097 (2026-08-14). Data as JSON: /api/errors/0e122c03f981d052. Report an issue: GitHub.