CHROMA_PEAKS

Overview

The CHROMA_PEAKS function fits specialized peak shape models to chromatographic data using nonlinear least-squares optimization. Chromatography is an analytical technique used to separate mixtures, and the resulting output typically shows peaks corresponding to different compounds. Accurate characterization of peak shape is essential for quantitative analysis, peak deconvolution, and compound identification.

This function implements seven distinct peak models commonly used in chromatographic analysis:

  • Beta Distribution Asymmetric Peak: A flexible model using beta distribution parameters to capture asymmetric peak shapes with controllable skewness.
  • Bigaussian Asymmetric Peak: A piecewise Gaussian model with different widths (w1, w2) on each side of the peak center, ideal for simple asymmetric peaks.
  • Chesler-Cram Exponential Tail Peak: Combines a Gaussian core with an exponential tailing function to model peaks with significant trailing edges.
  • Edgeworth-Cramer Skewed Peak: Extends the Gaussian distribution using Edgeworth series expansion to incorporate skewness and kurtosis corrections.
  • Exponentially Modified Gaussian (EMG) Peak: A convolution of a Gaussian and an exponential distribution, widely used as a theoretical model for chromatographic peak shapes. See the Wikipedia article on EMG distributions for mathematical details.
  • Giddings Bessel Dispersion Peak: Based on Bessel functions to model dispersion effects in chromatographic systems, using the modified Bessel function of the first kind.
  • Weibull Amplitude Asymmetric Peak: Uses the Weibull distribution to model asymmetric peaks with flexible tail behavior.

The underlying optimization uses scipy.optimize.curve_fit from the SciPy library, which implements the Levenberg-Marquardt algorithm for unconstrained problems or trust-region reflective algorithms when parameter bounds are specified. Each model includes intelligent initial parameter guesses derived from the input data and appropriate bounds to ensure physically meaningful results. The function returns fitted parameter values along with standard errors estimated from the covariance matrix.

This example function is provided as-is without any representation of accuracy.

Excel Usage

=CHROMA_PEAKS(xdata, ydata, chroma_peaks_model)
  • xdata (list[list], required): The xdata value
  • ydata (list[list], required): The ydata value
  • chroma_peaks_model (str, required): The chroma_peaks_model value

Returns (list[list]): 2D list [param_names, fitted_values, std_errors], or error string.

Examples

Example 1: Demo case 1

Inputs:

chroma_peaks_model xdata ydata
beta_distribution_asymmetric_peak 1.2 0.86160756
1.6 1.19043718
2 1.3
2.4 1.20340707
2.8 0.95652922

Excel formula:

=CHROMA_PEAKS("beta_distribution_asymmetric_peak", {1.2;1.6;2;2.4;2.8}, {0.86160756;1.19043718;1.3;1.20340707;0.95652922})

Expected output:

"non-error"

Example 2: Demo case 2

Inputs:

chroma_peaks_model xdata ydata
bigaussian_asymmetric_peak 0.01 1.3331896701095287
2.0075 5.013805785170162
4.005 0.5789147596326245
6.0024999999999995 0.01
8 0.03795354352729467

Excel formula:

=CHROMA_PEAKS("bigaussian_asymmetric_peak", {0.01;2.0075;4.005;6.0024999999999995;8}, {1.3331896701095287;5.013805785170162;0.5789147596326245;0.01;0.03795354352729467})

Expected output:

"non-error"

Example 3: Demo case 3

Inputs:

chroma_peaks_model xdata ydata
chesler_cram_exponential_tail_peak 0.01 0.6881970247706027
2.0075 3.476308512165623
4.005 0.2972799078474512
6.0024999999999995 0.01
8 0.026603606259349277

Excel formula:

=CHROMA_PEAKS("chesler_cram_exponential_tail_peak", {0.01;2.0075;4.005;6.0024999999999995;8}, {0.6881970247706027;3.476308512165623;0.2972799078474512;0.01;0.026603606259349277})

Expected output:

"non-error"

Example 4: Demo case 4

Inputs:

chroma_peaks_model xdata ydata
edgeworth_cramer_skewed_peak 0.01 0.593946685207516
2.0075 -0.3106033211311715
4.005 -0.06975777024010682
6.0024999999999995 0.03426042007874261
8 0.006244911955196433

Excel formula:

=CHROMA_PEAKS("edgeworth_cramer_skewed_peak", {0.01;2.0075;4.005;6.0024999999999995;8}, {0.593946685207516;-0.3106033211311715;-0.06975777024010682;0.03426042007874261;0.006244911955196433})

Expected output:

"non-error"

Example 5: Demo case 5

Inputs:

chroma_peaks_model xdata ydata
exponentially_modified_gaussian_peak 0.01 1
2.0075 1
4.005 1
6.0024999999999995 1
8 1

Excel formula:

=CHROMA_PEAKS("exponentially_modified_gaussian_peak", {0.01;2.0075;4.005;6.0024999999999995;8}, {1;1;1;1;1})

Expected output:

"non-error"

Example 6: Demo case 6

Inputs:

chroma_peaks_model xdata ydata
giddings_bessel_dispersion_peak 0.01 1
2.0075 1
4.005 1
6.0024999999999995 1
8 1

Excel formula:

=CHROMA_PEAKS("giddings_bessel_dispersion_peak", {0.01;2.0075;4.005;6.0024999999999995;8}, {1;1;1;1;1})

Expected output:

"non-error"

Example 7: Demo case 7

Inputs:

chroma_peaks_model xdata ydata
weibull_amplitude_asymmetric_peak 0.01 0.08678315093623744
2.0075 1.006439786471626
4.005 0.3007064729763167
6.0024999999999995 0.03250067264273272
8 0.014649108934044

Excel formula:

=CHROMA_PEAKS("weibull_amplitude_asymmetric_peak", {0.01;2.0075;4.005;6.0024999999999995;8}, {0.08678315093623744;1.006439786471626;0.3007064729763167;0.03250067264273272;0.014649108934044})

Expected output:

"non-error"

Python Code

import numpy as np
from scipy import special as sc
from scipy.optimize import curve_fit as scipy_curve_fit
import math

def chroma_peaks(xdata, ydata, chroma_peaks_model):
    """
    Fits chroma_peaks models to data using scipy.optimize.curve_fit. See https://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.curve_fit.html for details.

    See: https://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.curve_fit.html

    This example function is provided as-is without any representation of accuracy.

    Args:
        xdata (list[list]): The xdata value
        ydata (list[list]): The ydata value
        chroma_peaks_model (str): The chroma_peaks_model value Valid options: Beta Distribution Asymmetric Peak, Bigaussian Asymmetric Peak, Chesler Cram Exponential Tail Peak, Edgeworth Cramer Skewed Peak, Exponentially Modified Gaussian Peak, Giddings Bessel Dispersion Peak, Weibull Amplitude Asymmetric Peak.

    Returns:
        list[list]: 2D list [param_names, fitted_values, std_errors], or error string.
    """
    def _validate_data(xdata, ydata):
        """Validate and convert both xdata and ydata to numpy arrays."""
        for name, arg in [("xdata", xdata), ("ydata", ydata)]:
            if not isinstance(arg, list) or len(arg) < 2:
                raise ValueError(f"{name}: must be a 2D list with at least two rows")
            vals = []
            for i, row in enumerate(arg):
                if not isinstance(row, list) or len(row) == 0:
                    raise ValueError(f"{name} row {i}: must be a non-empty list")
                try:
                    vals.append(float(row[0]))
                except Exception:
                    raise ValueError(f"{name} row {i}: non-numeric value")
            if name == "xdata":
                x_arr = np.asarray(vals, dtype=np.float64)
            else:
                y_arr = np.asarray(vals, dtype=np.float64)

        if x_arr.shape[0] != y_arr.shape[0]:
            raise ValueError("xdata and ydata must have the same number of rows")
        return x_arr, y_arr

    # Model definitions dictionary
    models = {
        'beta_distribution_asymmetric_peak': {
            'params': ['y0', 'xc', 'A', 'w1', 'w2', 'w3'],
            'model': lambda x, y0, xc, A, w1, w2, w3: y0 + A * np.power(1.0 + ((w2 + w3 - 2.0) / (w2 - 1.0)) * ((x - xc) / w1), w2 - 1.0) * np.power(1.0 - ((w2 + w3 - 2.0) / (w3 - 1.0)) * ((x - xc) / w1), w3 - 1.0),
            'guess': lambda xa, ya: (float(np.min(ya)), float(np.median(xa)), float(np.max(ya)), float(max(np.max(xa) - np.min(xa), 1e-3)), 2.0, 2.0),
            'bounds': ([-np.inf, -np.inf, 0.0, 0.0, 1.0000001, 1.0000001], np.inf),
        },
        'bigaussian_asymmetric_peak': {
            'params': ['y0', 'xc', 'H', 'w1', 'w2'],
            'model': lambda x, y0, xc, H, w1, w2: y0 + H * np.exp(-0.5 * np.square((x - xc) / np.where(x < xc, w1, w2))),
            'guess': lambda xa, ya: (float(np.min(ya)), float(np.median(xa)), float(np.ptp(ya) if np.ptp(ya) else 1.0), float(max((np.max(xa) - np.min(xa)) / 6.0, 1e-3)), float(max((np.max(xa) - np.min(xa)) / 6.0, 1e-3))),
            'bounds': ([-np.inf, -np.inf, -np.inf, 0.0, 0.0], np.inf),
        },
        'chesler_cram_exponential_tail_peak': {
            'params': ['y0', 'xc1', 'A', 'w', 'k2', 'xc2', 'B', 'k3', 'xc3'],
            'model': lambda x, y0, xc1, A, w, k2, xc2, B, k3, xc3: y0 + A * (np.exp(-np.square(x - xc1) / (2.0 * w)) + B * (1.0 - 0.5 * (1.0 - np.tanh(k2 * (x - xc2)))) * np.exp(-0.5 * k3 * (np.abs(x - xc3) + (x - xc3)))),
            'guess': lambda xa, ya: (float(np.min(ya)), float(np.median(xa)), float(np.ptp(ya) if np.ptp(ya) else 1.0), float(max(np.square(np.max(xa) - np.min(xa)) / 8.0, 1e-3)), 1.0, float(np.median(xa)), 0.5, 1.0, float(np.median(xa))),
            'bounds': ([-np.inf, -np.inf, -np.inf, 0.0, 0.0, -np.inf, -np.inf, 0.0, -np.inf], np.inf),
        },
        'edgeworth_cramer_skewed_peak': {
            'params': ['y0', 'xc', 'A', 'w', 'a3', 'a4'],
            'model': lambda x, y0, xc, A, w, a3, a4: y0 + (A / (w * np.sqrt(2.0 * np.pi))) * np.exp(-0.5 * np.square((x - xc) / w)) * (1.0 + (a3 / 6.0) * (x - xc) / w * ((np.square((x - xc) / w)) - 3.0) + (a4 / 24.0) * (np.power((x - xc) / w, 4) - 6.0 * np.square((x - xc) / w) + 3.0) + (10.0 * np.square(a3) / 720.0) * (np.power((x - xc) / w, 6) - 15.0 * np.power((x - xc) / w, 4) + 45.0 * np.square((x - xc) / w) - 15.0)),
            'guess': lambda xa, ya: (float(np.min(ya)), float(np.median(xa)), float(np.ptp(ya) if np.ptp(ya) else 1.0), float(max((np.max(xa) - np.min(xa)) / 6.0, 1e-3)), 0.0, 0.0),
            'bounds': ([-np.inf, -np.inf, 0.0, 0.0, -np.inf, -np.inf], np.inf),
        },
        'exponentially_modified_gaussian_peak': {
            'params': ['y0', 'A', 'xc', 'w', 't0'],
            'model': lambda x, y0, A, xc, w, t0: y0 + (A / t0) * np.exp(0.5 * np.square(w / t0) - (x - xc) / t0) * 0.5 * (1.0 + sc.erf(((x - xc) / w - w / (2.0 * t0)) / np.sqrt(0.5))),
            'guess': lambda xa, ya: (float(np.min(ya)), float(np.ptp(ya) if np.ptp(ya) else 1.0), float(np.median(xa)), float(max((np.max(xa) - np.min(xa)) / 6.0, 1e-3)), 1.0),
            'bounds': ([-np.inf, 0.0, -np.inf, 0.0, 0.0], np.inf),
        },
        'giddings_bessel_dispersion_peak': {
            'params': ['y0', 'xc', 'w', 'A'],
            'model': lambda x, y0, xc, w, A: y0 + A / w * np.sqrt(np.clip(xc, 1e-9, None) / np.clip(x, 1e-9, None)) * sc.iv(1, 2.0 * np.sqrt(np.clip(xc, 1e-9, None) * np.clip(x, 1e-9, None)) / w) * np.exp(-(x + xc) / w),
            'guess': lambda xa, ya: (float(np.min(ya)), float(np.median(xa)), float(max((np.max(xa) - np.min(xa)) / 4.0, 1e-3)), float(np.ptp(ya) if np.ptp(ya) else 1.0)),
            'bounds': ([-np.inf, -np.inf, 0.0, 0.0], np.inf),
        },
        'weibull_amplitude_asymmetric_peak': {
            'params': ['y0', 'xc', 'A', 'w1', 'w2'],
            'model': lambda x, y0, xc, A, w1, w2: y0 + A * np.exp(-np.exp(-(x - xc) / w2) - (x - xc) / w1),
            'guess': lambda xa, ya: (float(np.min(ya)), float(np.median(xa)), float(np.ptp(ya) if np.ptp(ya) else 1.0), float(max((np.max(xa) - np.min(xa)) / 4.0, 1e-3)), 1.0),
            'bounds': ([-np.inf, -np.inf, -np.inf, 0.0, 0.0], np.inf),
        }
    }

    # Validate model parameter
    if chroma_peaks_model not in models:
        return f"Invalid model: {str(chroma_peaks_model)}. Valid models are: {', '.join(models.keys())}"

    model_info = models[chroma_peaks_model]

    # Validate and convert input data
    try:
        x_arr, y_arr = _validate_data(xdata, ydata)
    except ValueError as e:
        return f"Invalid input: {e}"

    # Perform curve fitting
    try:
        p0 = model_info['guess'](x_arr, y_arr)
        bounds = model_info.get('bounds', (-np.inf, np.inf))
        if bounds == (-np.inf, np.inf):
            popt, pcov = scipy_curve_fit(model_info['model'], x_arr, y_arr, p0=p0, maxfev=10000)
        else:
            popt, pcov = scipy_curve_fit(model_info['model'], x_arr, y_arr, p0=p0, bounds=bounds, maxfev=10000)

        fitted_vals = [float(v) for v in popt]
        for v in fitted_vals:
            if math.isnan(v) or math.isinf(v):
                return "Fitting produced invalid numeric values (NaN or inf)."
    except ValueError as e:
        return f"Initial guess error: {e}"
    except Exception as e:
        return f"curve_fit error: {e}"

    # Calculate standard errors
    std_errors = None
    try:
        if pcov is not None and np.isfinite(pcov).all():
            std_errors = [float(v) for v in np.sqrt(np.diag(pcov))]
    except Exception:
        pass

    return [model_info['params'], fitted_vals, std_errors] if std_errors else [model_info['params'], fitted_vals]

Online Calculator