Skip to contents

For each fluorophore common to both variability.mad and mismatch.dist, computes the cosine similarity between that fluorophore's per-detector variability profile (e.g. the denoised MAD profile from assess.variability.mad()) and the per-detector magnitude of its bead-vs-cell mismatch (abs() of, e.g., bead.cell.dist()). Both profiles are non-negative by construction, so the resulting cosine similarity is bounded in [0, 1] and reflects purely where in detector space each quantity is concentrated – a value near 1 means the fluorophore's variant-to-variant variability and its bead-vs-cell mismatch peak at the same detectors, regardless of either quantity's overall magnitude; a value near 0 means they are concentrated in different, non- overlapping parts of the spectrum.

variability.mad and mismatch.dist are expected to come from independent computations and are not guaranteed to share the same detector set or column order, so the comparison is aligned explicitly by detector name rather than position. A warning is issued if the two detector sets differ.

Usage

assess.variability.alignment(variability.mad, mismatch.dist)

Arguments

variability.mad

Numeric matrix, fluorophores in rows and detectors in columns, as returned by assess.variability.mad().

mismatch.dist

Numeric matrix, fluorophores in rows and detectors in columns, as returned by bead.cell.dist(). Values are used as abs(mismatch.dist) so that alignment reflects spectral location rather than the sign of the mismatch.

Value

A one-column numeric matrix of cosine similarity values (column name "Alignment"), one row per fluorophore common to both inputs.