oscura 0.0.1__py3-none-any.whl → 0.1.0__py3-none-any.whl

oscura/analyzers/digital/clock.py

@@ -0,0 +1,805 @@
+"""Advanced clock recovery for digital signals.
+
+This module provides comprehensive clock recovery and analysis tools for digital
+signals, including frequency detection, clock reconstruction, baud rate detection,
+and jitter measurement.
+
+
+Example:
+    >>> from oscura.analyzers.digital.clock import detect_clock_frequency, recover_clock
+    >>> freq = detect_clock_frequency(data_trace, sample_rate=1e9)
+    >>> print(f"Detected clock: {freq/1e6:.2f} MHz")
+    >>> clock = recover_clock(data_trace, sample_rate=1e9, method='edge')
+    >>> metrics = measure_clock_jitter(clock, sample_rate=1e9)
+
+References:
+    Gardner, F.M.: "Phaselock Techniques" (3rd Ed), Wiley, 2005
+    Lee, E.A. & Messerschmitt, D.G.: "Digital Communication" (2nd Ed), 1994
+    IEEE 1241-2010: Standard for Terminology and Test Methods for ADCs
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+from typing import TYPE_CHECKING, Any, ClassVar, Literal
+
+import numpy as np
+from scipy import signal
+
+from oscura.core.exceptions import InsufficientDataError, ValidationError
+
+if TYPE_CHECKING:
+    from numpy.typing import NDArray
+
+
+@dataclass
+class ClockMetrics:
+    """Clock signal quality metrics.
+
+
+
+    Attributes:
+        frequency: Detected frequency in Hz.
+        period_samples: Period in samples.
+        period_seconds: Period in seconds.
+        jitter_rms: RMS jitter in seconds.
+        jitter_pp: Peak-to-peak jitter in seconds.
+        duty_cycle: Duty cycle (0.0 to 1.0).
+        stability: Stability score (0.0 to 1.0).
+        confidence: Detection confidence (0.0 to 1.0).
+    """
+
+    frequency: float
+    period_samples: float
+    period_seconds: float
+    jitter_rms: float
+    jitter_pp: float
+    duty_cycle: float
+    stability: float
+    confidence: float
+
+
+@dataclass
+class BaudRateResult:
+    """Result of baud rate detection.
+
+
+
+    Attributes:
+        baud_rate: Detected baud rate in bits per second.
+        bit_period_samples: Bit period in samples.
+        confidence: Detection confidence (0.0 to 1.0).
+        method: Method used for detection.
+    """
+
+    baud_rate: int
+    bit_period_samples: float
+    confidence: float
+    method: str
+
+
+class ClockRecovery:
+    """Recover clock signal from data.
+
+
+
+    This class provides multiple methods for clock recovery including edge-based,
+    FFT-based, and autocorrelation-based detection, as well as PLL tracking and
+    baud rate detection for asynchronous protocols.
+
+    Can be initialized with or without sample_rate:
+    - With sample_rate: ClockRecovery(sample_rate=1e9)
+    - Without: ClockRecovery() - sample_rate extracted from trace metadata
+    """
+
+    # Standard baud rates for async protocols
+    STANDARD_BAUD_RATES: ClassVar[list[int]] = [
+        300,
+        600,
+        1200,
+        2400,
+        4800,
+        9600,
+        14400,
+        19200,
+        28800,
+        38400,
+        57600,
+        115200,
+        230400,
+        460800,
+        921600,
+        1000000,
+        2000000,
+    ]
+
+    def __init__(self, sample_rate: float | None = None):
+        """Initialize with optional sample rate.
+
+        Args:
+            sample_rate: Sample rate in Hz. If None, will be extracted from trace metadata.
+
+        Raises:
+            ValidationError: If sample rate is provided and invalid.
+        """
+        if sample_rate is not None and sample_rate <= 0:
+            raise ValidationError(f"Sample rate must be positive, got {sample_rate}")
+
+        self.sample_rate: float | None = float(sample_rate) if sample_rate is not None else None
+
+    def _get_sample_rate(self, trace: Any) -> float:
+        """Extract sample rate from trace or use stored value.
+
+        Args:
+            trace: A DigitalTrace/WaveformTrace with metadata, or a numpy array.
+
+        Returns:
+            Sample rate in Hz.
+
+        Raises:
+            ValidationError: If sample rate cannot be determined.
+        """
+        if self.sample_rate is not None:
+            return self.sample_rate
+
+        # Try to extract from trace metadata
+        if hasattr(trace, "metadata") and hasattr(trace.metadata, "sample_rate"):
+            return float(trace.metadata.sample_rate)
+
+        raise ValidationError(
+            "Sample rate not set and cannot be extracted from trace. "
+            "Either provide sample_rate to constructor or use a trace with metadata."
+        )
+
+    def _get_trace_data(self, trace: Any) -> NDArray[np.float64]:
+        """Extract numpy array from trace object.
+
+        Args:
+            trace: A DigitalTrace/WaveformTrace or numpy array.
+
+        Returns:
+            Numpy array of signal data.
+        """
+        if hasattr(trace, "data"):
+            return np.asarray(trace.data, dtype=np.float64)
+        return np.asarray(trace, dtype=np.float64)
+
+    def detect_frequency(
+        self, trace: Any, method: Literal["edge", "fft", "autocorr"] = "edge"
+    ) -> float:
+        """Detect clock frequency from signal (supports DigitalTrace).
+
+
+
+        This method supports both raw numpy arrays and DigitalTrace objects.
+        Sample rate is extracted from trace metadata if not set in constructor.
+
+        Args:
+            trace: Signal trace data (DigitalTrace or numpy array).
+            method: Detection method to use.
+
+        Returns:
+            Detected frequency in Hz.
+
+        Example:
+            >>> recovery = ClockRecovery()
+            >>> freq = recovery.detect_frequency(digital_trace)
+        """
+        sample_rate = self._get_sample_rate(trace)
+        data = self._get_trace_data(trace)
+
+        # Temporarily set sample rate for internal methods
+        old_rate = self.sample_rate
+        self.sample_rate = sample_rate
+
+        try:
+            return self.detect_clock_frequency(data, method)
+        finally:
+            self.sample_rate = old_rate
+
+    def detect_clock_frequency(
+        self, trace: NDArray[np.float64], method: Literal["edge", "fft", "autocorr"] = "edge"
+    ) -> float:
+        """Detect clock frequency from signal.
+
+
+
+        Detects the dominant clock frequency using the specified method.
+        Each method has different strengths:
+        - edge: Best for clean digital signals with clear transitions
+        - fft: Best for noisy signals or periodic analog waveforms
+        - autocorr: Best for periodic patterns with timing jitter
+
+        Args:
+            trace: Signal trace data.
+            method: Detection method to use.
+
+        Returns:
+            Detected frequency in Hz.
+
+        Raises:
+            InsufficientDataError: If trace is too short.
+            ValidationError: If method is invalid or detection fails.
+        """
+        if len(trace) < 10:
+            raise InsufficientDataError("Trace must have at least 10 samples")
+
+        if self.sample_rate is None:
+            raise ValidationError(
+                "Sample rate not set. Use detect_frequency() with trace metadata."
+            )
+
+        if method == "edge":
+            return self._detect_frequency_edge(trace)
+        elif method == "fft":
+            return self._detect_frequency_fft(trace)
+        elif method == "autocorr":
+            return self._detect_frequency_autocorr(trace)
+        else:
+            raise ValidationError(f"Unknown method: {method}")
+
+    def recover_clock(
+        self, data_trace: NDArray[np.float64], method: Literal["edge", "pll", "fft"] = "edge"
+    ) -> NDArray[np.float64]:
+        """Recover clock signal from data.
+
+
+
+        Reconstructs a clock signal from the data trace. The recovered clock
+        is a square wave aligned to the detected clock transitions.
+
+        Args:
+            data_trace: Data signal trace.
+            method: Recovery method to use.
+
+        Returns:
+            Recovered clock trace (same length as input).
+
+        Raises:
+            InsufficientDataError: If trace is too short.
+            ValidationError: If method is invalid or recovery fails.
+        """
+        if len(data_trace) < 10:
+            raise InsufficientDataError("Trace must have at least 10 samples")
+
+        if self.sample_rate is None:
+            raise ValidationError("Sample rate not set")
+
+        # Detect clock frequency first
+        freq = self.detect_clock_frequency(data_trace, method=method if method != "pll" else "edge")
+
+        if freq <= 0:
+            raise ValidationError("Failed to detect valid clock frequency")
+
+        if method == "pll":
+            # Use PLL tracking for robust recovery
+            return self._pll_track(data_trace, freq)
+        else:
+            # Generate ideal square wave at detected frequency
+            _period_samples = self.sample_rate / freq
+            n_samples = len(data_trace)
+            t = np.arange(n_samples)
+
+            # Generate square wave (50% duty cycle)
+            clock_raw = signal.square(2 * np.pi * freq * t / self.sample_rate)
+
+            # Normalize to 0-1 range
+            clock = (clock_raw + 1.0) / 2.0
+
+            return np.asarray(clock, dtype=np.float64)
+
+    def detect_baud_rate(
+        self, trace: NDArray[np.float64], candidates: list[int] | None = None
+    ) -> BaudRateResult:
+        """Auto-detect baud rate for async protocols.
+
+
+
+        Detects the baud rate by analyzing bit timing. Works best with traces
+        containing start bits or transitions between different bit values.
+
+        Args:
+            trace: Signal trace data.
+            candidates: List of candidate baud rates to test. If None, uses
+                       standard rates.
+
+        Returns:
+            BaudRateResult with detected baud rate and confidence.
+
+        Raises:
+            InsufficientDataError: If trace is too short or not enough edges found.
+            ValidationError: If sample rate is not set.
+        """
+        if len(trace) < 100:
+            raise InsufficientDataError("Need at least 100 samples for baud rate detection")
+
+        if self.sample_rate is None:
+            raise ValidationError("Sample rate not set")
+
+        if candidates is None:
+            candidates = self.STANDARD_BAUD_RATES
+
+        # Detect edges to find bit transitions
+        edges = self._detect_edges_simple(trace)
+
+        if len(edges) < 3:
+            raise InsufficientDataError("Not enough edges to detect baud rate")
+
+        # Calculate inter-edge intervals
+        intervals = np.diff(edges)
+
+        # The minimum interval should be close to one bit period
+        # (assuming we have at least some single-bit pulses)
+        # Use histogram to find most common interval
+        hist, bin_edges = np.histogram(intervals, bins=50)
+        most_common_interval = bin_edges[np.argmax(hist)]
+
+        # Convert to frequency
+        detected_freq = self.sample_rate / most_common_interval
+
+        # Find closest standard baud rate
+        candidates_array = np.array(candidates)
+        errors = np.abs(candidates_array - detected_freq)
+        best_idx = np.argmin(errors)
+        best_baud = candidates_array[best_idx]
+
+        # Calculate confidence based on how close we are to standard rate
+        relative_error = errors[best_idx] / best_baud
+        confidence = max(0.0, 1.0 - relative_error * 10)
+
+        bit_period_samples = self.sample_rate / best_baud
+
+        return BaudRateResult(
+            baud_rate=int(best_baud),
+            bit_period_samples=float(bit_period_samples),
+            confidence=float(confidence),
+            method="edge_histogram",
+        )
+
+    def measure_clock_jitter(self, clock_trace: NDArray[np.float64]) -> ClockMetrics:
+        """Measure clock jitter and quality metrics.
+
+
+
+        Analyzes a clock signal to measure jitter, duty cycle, and stability.
+        Works best with recovered or measured clock signals.
+
+        Args:
+            clock_trace: Clock signal trace.
+
+        Returns:
+            ClockMetrics with comprehensive quality measurements.
+
+        Raises:
+            InsufficientDataError: If trace is too short or has too few edges.
+            ValidationError: If sample rate is not set.
+        """
+        if len(clock_trace) < 10:
+            raise InsufficientDataError("Trace must have at least 10 samples")
+
+        if self.sample_rate is None:
+            raise ValidationError("Sample rate not set")
+
+        # Detect rising and falling edges
+        rising_edges = self._detect_edges_by_type(clock_trace, "rising")
+        falling_edges = self._detect_edges_by_type(clock_trace, "falling")
+
+        if len(rising_edges) < 3:
+            raise InsufficientDataError("Need at least 3 rising edges for jitter measurement")
+
+        # Calculate periods from rising edge to rising edge
+        periods = np.diff(rising_edges)
+
+        if len(periods) == 0:
+            raise InsufficientDataError("Cannot calculate period from single edge")
+
+        # Mean period
+        mean_period_samples = np.mean(periods)
+        mean_period_seconds = mean_period_samples / self.sample_rate
+        frequency = 1.0 / mean_period_seconds
+
+        # RMS jitter (standard deviation of periods)
+        jitter_rms_samples = np.std(periods)
+        jitter_rms = jitter_rms_samples / self.sample_rate
+
+        # Peak-to-peak jitter
+        jitter_pp_samples = np.ptp(periods)
+        jitter_pp = jitter_pp_samples / self.sample_rate
+
+        # Duty cycle (high time / period)
+        if len(falling_edges) >= len(rising_edges):
+            # Can measure duty cycle
+            high_times = []
+            for _i, rise in enumerate(rising_edges):
+                # Find next falling edge
+                fall_idx = np.searchsorted(falling_edges, rise)
+                if fall_idx < len(falling_edges):
+                    high_time = falling_edges[fall_idx] - rise
+                    high_times.append(high_time)
+
+            if high_times:
+                mean_high_time = np.mean(high_times)
+                duty_cycle = mean_high_time / mean_period_samples
+            else:
+                duty_cycle = 0.5  # Assume 50% if cannot measure
+        else:
+            duty_cycle = 0.5
+
+        # Stability score (inverse of relative jitter)
+        relative_jitter = (
+            jitter_rms_samples / mean_period_samples if mean_period_samples > 0 else 1.0
+        )
+        stability = max(0.0, 1.0 - relative_jitter * 10)
+
+        # Confidence based on number of periods and stability
+        confidence = min(1.0, len(periods) / 100.0) * stability
+
+        return ClockMetrics(
+            frequency=float(frequency),
+            period_samples=float(mean_period_samples),
+            period_seconds=float(mean_period_seconds),
+            jitter_rms=float(jitter_rms),
+            jitter_pp=float(jitter_pp),
+            duty_cycle=float(np.clip(duty_cycle, 0.0, 1.0)),
+            stability=float(stability),
+            confidence=float(confidence),
+        )
+
+    def _detect_frequency_edge(self, trace: NDArray[np.float64]) -> float:
+        """Detect frequency using edge timing histogram.
+
+
+
+        Args:
+            trace: Signal trace.
+
+        Returns:
+            Detected frequency in Hz.
+
+        Raises:
+            ValidationError: If not enough edges found to detect frequency.
+        """
+        edges = self._detect_edges_simple(trace)
+
+        if len(edges) < 3:
+            raise ValidationError("Not enough edges to detect frequency")
+
+        # Calculate inter-edge intervals
+        intervals = np.diff(edges)
+
+        # Build histogram of intervals
+        # The peak should correspond to half the period (edge to edge)
+        hist, bin_edges = np.histogram(intervals, bins=50)
+        _peak_interval = bin_edges[np.argmax(hist)]
+
+        # Frequency is sample_rate / (2 * interval) for edge-to-edge
+        # But we need to check if these are half-periods or full periods
+        # Use median interval as robust estimator
+        median_interval = np.median(intervals)
+
+        # Assume median represents half-period (rising to falling or vice versa)
+        # So full period is 2x median interval
+        period_samples = 2 * median_interval
+        frequency = self.sample_rate / period_samples
+
+        return float(frequency)
+
+    def _detect_frequency_fft(self, trace: NDArray[np.float64]) -> float:
+        """Detect frequency using FFT spectral analysis.
+
+
+
+        Args:
+            trace: Signal trace.
+
+        Returns:
+            Detected frequency in Hz.
+
+        Raises:
+            ValidationError: If sample rate is not set.
+        """
+        # Remove DC component
+        trace_ac = trace - np.mean(trace)
+
+        # Apply window to reduce spectral leakage
+        window = signal.windows.hann(len(trace_ac))
+        trace_windowed = trace_ac * window
+
+        # Compute FFT
+        fft = np.fft.rfft(trace_windowed)
+        if self.sample_rate is None:
+            raise ValidationError("Sample rate not set")
+        freqs = np.fft.rfftfreq(len(trace_windowed), 1.0 / self.sample_rate)
+
+        # Find peak in magnitude spectrum
+        magnitude = np.abs(fft)
+
+        # Ignore DC and very low frequencies (below 10 Hz)
+        min_freq_hz = 10.0
+        min_freq_idx = np.searchsorted(freqs, min_freq_hz)
+        if min_freq_idx >= len(magnitude):
+            min_freq_idx = np.intp(1)
+
+        peak_idx = min_freq_idx + np.argmax(magnitude[min_freq_idx:])
+        frequency = freqs[peak_idx]
+
+        return float(frequency)
+
+    def _detect_frequency_autocorr(self, trace: NDArray[np.float64]) -> float:
+        """Detect frequency using autocorrelation.
+
+
+
+        Args:
+            trace: Signal trace.
+
+        Returns:
+            Detected frequency in Hz.
+
+        Raises:
+            ValidationError: If no periodic pattern detected or sample rate not set.
+        """
+        # Remove mean
+        trace_centered = trace - np.mean(trace)
+
+        # Compute autocorrelation
+        autocorr = signal.correlate(trace_centered, trace_centered, mode="full")
+        autocorr = autocorr[len(autocorr) // 2 :]  # Keep only positive lags
+
+        # Normalize
+        autocorr = autocorr / autocorr[0]
+
+        # Find first peak after lag 0
+        # Look for peaks in autocorrelation
+        peaks, _ = signal.find_peaks(autocorr, height=0.3)
+
+        if len(peaks) == 0:
+            raise ValidationError("No periodic pattern detected in autocorrelation")
+
+        # First peak corresponds to period
+        period_samples = peaks[0]
+        if self.sample_rate is None:
+            raise ValidationError("Sample rate not set")
+        frequency = self.sample_rate / period_samples
+
+        return float(frequency)
+
+    def _pll_track(
+        self, trace: NDArray[np.float64], initial_freq: float, bandwidth: float = 0.01
+    ) -> NDArray[np.float64]:
+        """Software PLL for phase tracking.
+
+
+
+        Implements a simple digital PLL for tracking phase and frequency
+        variations in the input signal.
+
+        Args:
+            trace: Input data trace.
+            initial_freq: Initial frequency estimate in Hz.
+            bandwidth: Loop bandwidth (0.0 to 1.0), lower = more filtering.
+
+        Returns:
+            Recovered clock signal.
+
+        Raises:
+            ValidationError: If sample rate is not set.
+        """
+        n_samples = len(trace)
+        clock = np.zeros(n_samples)
+
+        # PLL state
+        phase = 0.0
+        freq = initial_freq
+        if self.sample_rate is None:
+            raise ValidationError("Sample rate not set")
+        omega = 2 * np.pi * freq / self.sample_rate
+
+        # Loop filter gains (proportional + integral)
+        kp = 2 * bandwidth  # Proportional gain
+        ki = bandwidth**2  # Integral gain
+
+        # Detect edges for phase error calculation
+        threshold = (np.max(trace) + np.min(trace)) / 2.0
+        prev_sample = trace[0]
+
+        for i in range(n_samples):
+            # Generate clock output
+            clock[i] = 1.0 if np.cos(phase) > 0 else 0.0
+
+            # Detect phase error at edges
+            current_sample = trace[i]
+            phase_error = 0.0
+
+            # Simple phase detector: check if edge coincides with clock transition
+            if (prev_sample < threshold <= current_sample) or (
+                prev_sample > threshold >= current_sample
+            ):
+                # Edge detected
+                clock_value = np.cos(phase)
+                # Phase error is sign of clock at edge
+                phase_error = np.sign(clock_value) * 0.1
+
+            # Update frequency and phase with loop filter
+            _freq_adjust = kp * phase_error
+            omega += ki * phase_error
+
+            # Update phase
+            phase += omega
+            phase = phase % (2 * np.pi)
+
+            prev_sample = current_sample
+
+        return clock
+
+    def _detect_edges_simple(self, trace: NDArray[np.float64]) -> NDArray[np.intp]:
+        """Detect all edges in trace (both rising and falling).
+
+        Args:
+            trace: Signal trace.
+
+        Returns:
+            Array of edge indices.
+        """
+        threshold = (np.max(trace) + np.min(trace)) / 2.0
+        rising = np.where((trace[:-1] < threshold) & (trace[1:] >= threshold))[0]
+        falling = np.where((trace[:-1] > threshold) & (trace[1:] <= threshold))[0]
+
+        # Combine and sort
+        all_edges = np.concatenate([rising, falling])
+        all_edges.sort()
+
+        return all_edges
+
+    def _detect_edges_by_type(
+        self, trace: NDArray[np.float64], edge_type: Literal["rising", "falling"]
+    ) -> NDArray[np.intp]:
+        """Detect edges of specific type.
+
+        Args:
+            trace: Signal trace.
+            edge_type: Type of edge to detect.
+
+        Returns:
+            Array of edge indices.
+        """
+        threshold = (np.max(trace) + np.min(trace)) / 2.0
+
+        if edge_type == "rising":
+            edges = np.where((trace[:-1] < threshold) & (trace[1:] >= threshold))[0]
+        else:  # falling
+            edges = np.where((trace[:-1] > threshold) & (trace[1:] <= threshold))[0]
+
+        return edges + 1  # Return index after crossing
+
+
+# Convenience functions
+
+
+def detect_clock_frequency(
+    trace: NDArray[np.float64],
+    sample_rate: float,
+    method: Literal["edge", "fft", "autocorr"] = "edge",
+) -> float:
+    """Detect clock frequency from signal.
+
+
+
+    Convenience function for detecting clock frequency without creating
+    a ClockRecovery instance.
+
+    Args:
+        trace: Signal trace data.
+        sample_rate: Sample rate in Hz.
+        method: Detection method ('edge', 'fft', or 'autocorr').
+
+    Returns:
+        Detected frequency in Hz.
+
+    Example:
+        >>> freq = detect_clock_frequency(data, sample_rate=1e9, method='edge')
+        >>> print(f"Clock: {freq/1e6:.2f} MHz")
+    """
+    recovery = ClockRecovery(sample_rate)
+    return recovery.detect_clock_frequency(trace, method)
+
+
+def recover_clock(
+    data_trace: NDArray[np.float64],
+    sample_rate: float,
+    method: Literal["edge", "pll", "fft"] = "edge",
+) -> NDArray[np.float64]:
+    """Recover clock signal from data.
+
+
+
+    Convenience function for recovering clock signal without creating
+    a ClockRecovery instance.
+
+    Args:
+        data_trace: Data signal trace.
+        sample_rate: Sample rate in Hz.
+        method: Recovery method ('edge', 'pll', or 'fft').
+
+    Returns:
+        Recovered clock trace.
+
+    Example:
+        >>> clock = recover_clock(data, sample_rate=1e9, method='pll')
+    """
+    recovery = ClockRecovery(sample_rate)
+    return recovery.recover_clock(data_trace, method)
+
+
+def detect_baud_rate(
+    trace: Any, sample_rate: float | None = None, candidates: list[int] | None = None
+) -> int | BaudRateResult:
+    """Auto-detect baud rate.
+
+
+
+    Convenience function for baud rate detection. Supports both DigitalTrace
+    objects (with metadata) and raw numpy arrays (requiring sample_rate).
+
+    Args:
+        trace: Signal trace data (DigitalTrace or numpy array).
+        sample_rate: Sample rate in Hz (optional if trace has metadata).
+        candidates: List of candidate baud rates. If None, uses standard rates.
+
+    Returns:
+        Detected baud rate as int (for DigitalTrace) or BaudRateResult.
+
+    Raises:
+        ValidationError: If sample_rate is required but not provided.
+
+    Example:
+        >>> baud = detect_baud_rate(digital_trace)  # Uses metadata
+        >>> result = detect_baud_rate(data_array, sample_rate=1e6)  # Explicit rate
+    """
+    # Check if trace is a DigitalTrace with metadata
+    if hasattr(trace, "metadata") and hasattr(trace.metadata, "sample_rate"):
+        rate = trace.metadata.sample_rate
+        data = np.asarray(trace.data, dtype=np.float64)
+        recovery = ClockRecovery(rate)
+        result = recovery.detect_baud_rate(data, candidates)
+        return result.baud_rate  # Return just the baud rate for DigitalTrace
+    elif sample_rate is not None:
+        data = np.asarray(trace, dtype=np.float64)
+        recovery = ClockRecovery(sample_rate)
+        return recovery.detect_baud_rate(data, candidates)
+    else:
+        raise ValidationError("sample_rate required when trace is not a DigitalTrace with metadata")
+
+
+def measure_clock_jitter(clock_trace: NDArray[np.float64], sample_rate: float) -> ClockMetrics:
+    """Measure clock jitter.
+
+
+
+    Convenience function for jitter measurement without creating
+    a ClockRecovery instance.
+
+    Args:
+        clock_trace: Clock signal trace.
+        sample_rate: Sample rate in Hz.
+
+    Returns:
+        ClockMetrics with jitter and quality measurements.
+
+    Example:
+        >>> metrics = measure_clock_jitter(clock, sample_rate=1e9)
+        >>> print(f"RMS jitter: {metrics.jitter_rms*1e12:.2f} ps")
+    """
+    recovery = ClockRecovery(sample_rate)
+    return recovery.measure_clock_jitter(clock_trace)
+
+
+__all__ = [
+    "BaudRateResult",
+    "ClockMetrics",
+    "ClockRecovery",
+    "detect_baud_rate",
+    "detect_clock_frequency",
+    "measure_clock_jitter",
+    "recover_clock",
+]