chemometrics.mcr.McrAR

Multivariate Curve Resolution - Alternating Regression

Multivariate curve resolution factorizes data such that the factors maximize the explained data. Additional constraints are given for the scores and loadings, such as non-negativeness, baseline restrictions, unimodality, etc. Due to its origin in the analysis of chemical information, a different nomenclature is widely used in literature and also in this implementation. Scores are interpreted as concentration matrices C_, loadings are the transposed spectral matrices ST_.

param c_regr

Instantiated regression class (or string, see Notes) for calculating the C matrix

type c_regr

str, class

param st_regr

Instantiated regression class (or string, see Notes) for calculating the S^T matrix

type st_regr

str, class

param fit_kwargs

kwargs sent to fit and fit_transform methods

type fit_kwargs

dict

param c_fit_kwargs

kwargs sent to c_regr.fit method

type c_fit_kwargs

dict

param st_fit_kwargs

kwargs sent to``st_regr.fit`` method

type st_fit_kwargs

dict

param c_constraints

List of constraints applied to calculation of C matrix

type c_constraints

list

param st_constraints

List of constraints applied to calculation of S^T matrix

type st_constraints

list

param max_iter

Maximum number of iterations. One iteration calculates both C and S^T

type max_iter

int

param err_fcn

Function to calculate error/differences after each least squares calculation (ie twice per iteration). Outputs to err attribute.

type err_fcn

function

param tol_increase

Factor increase to allow in err attribute. Set to 0 for no increase allowed. E.g., setting to 1.0 means the err can double per iteration.

type tol_increase

float

param tol_n_increase

Number of consecutive iterations for which the err attribute can increase

type tol_n_increase

int

param tol_err_change

If err changes less than tol_err_change, per iteration, break.

type tol_err_change

float

param tol_n_above_min

Number of half-iterations that can be performed without reaching a new error-minimum

type tol_n_above_min

int

chemometrics.mcr.McrAR.err

List of calculated errors (from err_fcn) after each least squares (ie twice per iteration)

Type

list

chemometrics.mcr.McrAR.C_

Most recently calculated C matrix (that did not cause a tolerance failure)

Type

ndarray [n_samples, n_targets]

chemometrics.mcr.McrAR.ST_

Most recently calculated S^T matrix (that did not cause a tolerance failure)

Type

ndarray [n_targets, n_features]

chemometrics.mcr.McrAR.components_

Synonym for ST_, providing sklearn like compatibility

Type

ndarray [n_targets, n_features]

chemometrics.mcr.McrAR.C_opt_

[Optimal] C matrix for lowest err attribute

Type

ndarray [n_samples, n_targets]

chemometrics.mcr.McrAR.ST_opt_

[Optimal] ST matrix for lowest err attribute

Type

ndarray [n_targets, n_features]

chemometrics.mcr.McrAR.n_iter

Total number of iterations performed

Type

int

chemometrics.mcr.McrAR.n_features

Total number of features, e.g. spectral frequencies.

Type

int

chemometrics.mcr.McrAR.n_samples

Total number of samples (e.g., pixels)

Type

int

chemometrics.mcr.McrAR.n_targets

Total number of targets (e.g., pure analytes)

Type

int

chemometrics.mcr.McrAR.n_iter_opt

Iteration when optimal C and ST calculated

Type

int

chemometrics.mcr.McrAR.exit_max_iter_reached

Exited iterations due to maximum number of iteration reached (max_iter parameter)

Type

bool

chemometrics.mcr.McrAR.exit_tol_increase

Exited iterations due to maximum fractional increase in error metric (via err_fcn)

Type

bool

chemometrics.mcr.McrAR.exit_tol_n_increase

Exited iterations due to maximum number of consecutive increases in error metric (via err fcn)

Type

bool

chemometrics.mcr.McrAR.exit_tol_err_change

Exited iterations due to error metric change that is smaller than tol_err_change

Type

bool

chemometrics.mcr.McrAR.exit_tol_n_above_min

Exited iterations due to maximum number of half-iterations for which the error metric increased above the minimum error

Type

bool

Notes

  • Built-in regressor classes (str can be used): OLS (ordinary least squares), NNLS (non-negatively constrained least squares). See mcr.regressors.

  • Built-in regressor methods can be given as a string to c_regr, st_regr; though instantiating an imported class gives more flexibility.

  • Setting any tolerance to None turns that check off

Module Attributes

D_

noindex:

D_opt_

components_

n_features

n_samples

n_targets