Skip to contents

Faster solver for per-cell optimization workflow. Performs spectral unmixing using weighted least squares.

Usage

unmix.wls.per.cell(cell.raw, spectra, weights = NULL, noise.floor = 125)

Arguments

cell.raw

Expression data from raw FCS files. A single row of cells with detectors in columns. Columns should be fluorescent data only and must match the columns in spectra.

spectra

Spectral signatures of fluorophores, normalized between 0 and 1, with fluorophores in rows and detectors in columns.

weights

The weighting values for the detectors. No checks are performed on supplied weights; unmix.wls should be used for standard cases. Default is NULL, in which case weighting is computed from this cell's own signal, floored at noise.floor.

noise.floor

Numeric, default 125. Lower clamp on this cell's signal before inversion when weights = NULL. Signal units, same convention as noise.floor elsewhere in the package. Ignored if weights is supplied.

Value

Unmixed data (one row), fluorophores in columns.