spectralNMF

<code><a rd-options="" href="/link/SpectraInTime-class?package=spectralAnalysis&version=3.12.0" data-mini-rdoc="spectralAnalysis::SpectraInTime-class">SpectraInTime-class</a></code>

object

number of NMF components to be found

rank

name of the NMF method to be used. "PGNMF" (default), "HALSacc" 
and "semiNMF" are methods derived from the hNMF package. All methods from the NMF package are also available.

method

this can be a list of spectralData objects, containing 
the pure component spectra. It can also be either of the NMF factor matrices with initial values

initSpectralData

number of NMF runs. It is recommended to run the NMF analyses multiple
times when random seeding is used, to avoid a suboptimal solution

nruns

subsampling factor used during NMF analysis

subsamplingFactor

Boolean indicating whether divergence checking should be performed

checkDivergence

Perform Non-Negative Matrix factorization on spectral data


Infrared, near-infrared and Raman spectroscopic data measured during chemical reactions, provide structural fingerprints by which molecules can be identified and quantified. The application of these spectroscopic techniques as inline process analytical tools (PAT), provides the (pharma-)chemical industry with novel tools, allowing to monitor their chemical processes, resulting in a better process understanding through insight in reaction rates, mechanistics, stability, etc.
Data can be read into R via the generic spc-format, which is generally supported by spectrometer vendor software. Versatile pre-processing functions are available to perform baseline correction by linking to the 'baseline' package; noise reduction via the 'signal' package; as well as time alignment, normalization, differentiation, integration and interpolation. Implementation based on the S4 object system allows storing a pre-processing pipeline as part of a spectral data object, and easily transferring it to other datasets. Interactive plotting tools are provided based on the 'plotly' package.
Non-negative matrix factorization (NMF) has been implemented to perform multivariate analyses on individual spectral datasets or on multiple datasets at once. NMF provides a parts-based representation of the spectral data in terms of spectral signatures of the chemical compounds and their relative proportions.
The functionality to read in spc-files was adapted from the 'hyperSpec' package.

Adriaan Blommaert

spectralAnalysis

Pre-Process, Visualize and Analyse Process Analytical Data, by
Spectral Data Measurements Made During a Chemical Process

Robin Van Oirbeek

Nicolas Sauwen

Tor Maes

Jan Dijkmans

Jef Cuypers

Tatsiana Khamiakova

Michel Tiel

Claudia Beleites

spectralNMF function

<code><a rd-options='' href='SpectraInTime-class'>SpectraInTime-class</a></code>

spectralNMF: Perform Non-Negative Matrix factorization on spectral data

Description

Usage

Arguments

Value