metbit.preprocessing.baseline

Apply baseline correction across spectra in a DataFrame.

Parameters

Rows are spectra; columns are ppm (x-axis) values.

Explicit x-axis to use (same length as columns). If None, tries to infer from `X.columns` by casting to float, else uses index positions.

If True, also return a DataFrame of the estimated baselines. **kwargs : Any Algorithm-specific keyword arguments (e.g., lam, p, max_iter, poly_order, ...).

Returns

Baseline-corrected spectra.

Estimated baselines (returned if return_baseline=True).

Examples: >>> import numpy as np >>> import pandas as pd >>> ppm = np.linspace(0, 10, 500) >>> spectra = pd.DataFrame(np.random.rand(5, 500), columns=ppm) >>> corrected = baseline_correct(spectra, method='asls', lam=1e6) >>> corrected_df, baseline_df = baseline_correct( ... spectra, method='rubberband', return_baseline=True ... )

Backwards-compatible wrapper for ALS baseline correction.

Parameters

Rows are spectra; columns are ppm values.

Smoothing parameter for ASLS.

Max iterations for ASLS.

Examples: >>> import numpy as np >>> import pandas as pd >>> ppm = np.linspace(0, 10, 500) >>> spectra = pd.DataFrame(np.random.rand(10, 500), columns=ppm) >>> corrected = bline(spectra, lam=1e7, max_iter=30)