THR_DIR_BOTH = "both"
THR_DIR_UP = "upper"
THR_DIR_LO = "lower"
THR_DIR_AUTO = "auto"
HIGH_THRESHOLD_ZSCORE_MULTIPLIER = 1.2
CRITICAL_THRESHOLD_ZSCORE_MULTIPLIER = 1.4

NO_RECOMMEND_NOT_ENOUGH_DATA = "NO_RECOMMEND_NOT_ENOUGH_DATA"

def cascade_thresholds(threshold):
    """
    Simple cascading thresholds for the initial release
    """
    medium = threshold
    high = medium * HIGH_THRESHOLD_ZSCORE_MULTIPLIER
    critical = medium * CRITICAL_THRESHOLD_ZSCORE_MULTIPLIER

    return (critical, high, medium)


def output_thresholds_dict(
    threshold, threshold_rounding, threshold_direction, static_params=None
):
    def transform(value):
        if static_params:
            return round(value * stdev + mean, threshold_rounding)
        else:
            return round(value, threshold_rounding)

    critical, high, medium = cascade_thresholds(threshold)
    if static_params:
        mean = static_params[0]
        stdev = static_params[1]

    if threshold_direction == THR_DIR_BOTH:
        critical_final = transform(critical)
        high_final = [transform(high), transform(-critical)]
        medium_final = [transform(medium), transform(-high)]
        normal_final = transform(-1 * medium)
        return f"{{'critical': {critical_final}, 'high': {high_final}, 'medium': {medium_final}, 'normal': {normal_final}}}"
    elif threshold_direction == THR_DIR_UP:
        critical_final = transform(critical)
        high_final = transform(high)
        medium_final = transform(medium)
        return f"{{'critical': {critical_final}, 'high': {high_final}, 'medium': {medium_final}}}"
    else:
        critical_final = transform(-critical)
        high_final = transform(-high)
        medium_final = transform(-medium)
        return f"{{'high': {critical_final}, 'medium': {high_final}, 'normal': {medium_final}}}"


def confidence_description(score):
    if score > 0.6:
        return "High"
    elif score > 0.4:
        return "Medium"
    elif score > 0.1:
        return "Low"
    else:
        return "No Pattern"