Missing functionality - Replace X coordinates with Y values from another signal for calibration

[PierreRaybaut]

Missing functionality - Replace X coordinates with Y values from another signal for calibration

Problem Description

Users need to perform signal axis calibration, particularly for spectroscopy workflows where:

• One signal contains measurement data (e.g., intensity vs pixel index)
• Another signal contains calibration data (e.g., wavelength values vs pixel index)
• Goal: Plot measurement intensity vs calibrated wavelength scale

This fundamental operation was missing from Sigima/DataLab, forcing users to:

• Use external tools for calibration
• Manually export/import data
• Attempt to use X-Y mode (which incorrectly applies resampling/interpolation)

Solution Implemented

Added new replace_x_by_other_y(src1, src2) function that:

• Takes Y values from src2 and uses them as X coordinates for src1's Y values
• Directly uses Y arrays without interpolation (preserves exact calibration values)
• Requires both signals to have the same number of points
• Automatically transfers metadata (X label/unit from src2.ylabel/yunit)
• Clear error message when signal sizes don't match

Technical Details

Function signature:

def replace_x_by_other_y(src1: SignalObj, src2: SignalObj) -> SignalObj:
    """Create a new signal using Y from src1 and Y from src2 as X coordinates."""

Usage example (wavelength calibration):

# Signal 1: intensity measurements [0, 1, 2, 3, 4] -> [10, 20, 30, 40, 50]
# Signal 2: wavelength calibration [0, 1, 2, 3, 4] -> [400, 450, 500, 550, 600]
# Result: intensity vs wavelength [400, 450, 500, 550, 600] -> [10, 20, 30, 40, 50]

calibrated = sigima.proc.signal.replace_x_by_other_y(intensity, wavelength)

Rationale for Bug Fix Classification

While this adds new functionality, it's classified as a bug fix because:

1. User-reported critical gap: Real users couldn't perform a fundamental operation expected in the target domain (spectroscopy)
2. Misleading existing feature: X-Y mode appeared to solve the problem but didn't (applies unwanted interpolation)
3. Expected functionality missing: Users had reasonable expectation this should exist
4. Quality issue: Software incomplete for its intended scientific use cases

This justifies inclusion in a maintenance release (1.0.3) rather than waiting for a minor version bump.

Related

• DataLab Issue #273: Same feature added to GUI
• Clarifies distinction from X-Y mode (which resamples/interpolates signals)