smefit.core package

Todo

docs

Submodules

smefit.core.chisquared module

Module for the computation of chi-squared values

smefit.core.chisquared.compute_chi2(config, dataset, coeffs, labels)

Compute the components for the chi2

Here we also perform the cross-validation splitting at the level of the residuals, so as to prevent singular covariances matrices.

A mask is applied to each experiment and if the dataset has only 1 datapoint it is placed in the training set

Returns the theory - exp vector and the inverse cov mat * (theory - exp)

smefit.core.chisquared.compute_total_chi2(config, datasets, coefficients, labels)

Function to compute total central chi2 for all datasets assuming no cross-correlations.

Returns the chi2 for the tr/val sets (or total chi2 if cross-validation is off)

smefit.core.compute_theory module

Module for the generation of theory predictions

smefit.core.compute_theory.make_predictions(config, dataset, coeffs, labels)

Generate the corrected theory predictions for dataset set given a set of SMEFT coefficients coeffs. Optionally a specific operator may be selected with iop

smefit.core.load_data module

Module for reading in commondata, theory and SMEFT corrections

smefit.core.load_data.CL_bounds(chi2, coeff, percent_CL)

Compute either the 68% or 95% confidence intervals

smefit.core.load_data.CoeffTuple

alias of Coefficients

class smefit.core.load_data.DataTuple(Commondata, SMTheory, CorrectionsDICT, CorrectionsVAL, HOcorrectionsDICT, HOcorrectionsVAL, ExpNames, NdataExp, Kinematics, Noise, TrainingMask, ValidationMask, CovMat)

Bases: tuple

property Commondata

Alias for field number 0

property CorrectionsDICT

Alias for field number 2

property CorrectionsVAL

Alias for field number 3

property CovMat

Alias for field number 12

property ExpNames

Alias for field number 6

property HOcorrectionsDICT

Alias for field number 4

property HOcorrectionsVAL

Alias for field number 5

property Kinematics

Alias for field number 8

property NdataExp

Alias for field number 7

property Noise

Alias for field number 9

property SMTheory

Alias for field number 1

property TrainingMask

Alias for field number 10

property ValidationMask

Alias for field number 11

smefit.core.load_data.aggregate_coefficients(config, loaded_datasets, i_rep)

Aggregate all coefficient labels and construct an array of coefficient values of suitable size. Returns a CoeffTuple of the labels and values.

smefit.core.load_data.artificial_data_check(set)

Check that the artificial noise reproduces the original covariance matrix

smefit.core.load_data.generate_CTartdata(commondata, covmat)

Generates experimental data noise for level 2 CT

smefit.core.load_data.generate_artdata(commondata, covmat, i_rep)

Generates artificial data noise for MC replica i_rep

smefit.core.load_data.generate_closure(config, dataset, coefficients, pseudocoeff, labels, level)

Generates a level 0, 1 or 2 closure test set using the coefficients table This is done by shifting the ‘Noise’ attribute of a dataset tuple accordingly We fit to pseudodata (the SM) as opposed to real data (experiment)

smefit.core.load_data.load_datasets(config, i_rep)

Loads commondata, theory and SMEFT corrections into a namedtuple

smefit.core.load_data.seed_handler(i_rep)

Allow the seed for splitting the data and artificial data replication (and artifical data for level 2 CT) to be set here