RawOifits

class jwst.ami.oifits.RawOifits(instrument_data, method='mean')

Bases: object

Store AMI data in the format required to write out to OIFITS files.

Builds the structure needed to write out oifits files according to the schema. Populates the structure with the observables from the fringe fitter. For data arrays, the observables are populated slice-by-slice to enable fitting with FringeFitter and storage as OiFits to take place in the same loop. Angular quantities, initially in radians from fringe fitting, are converted to degrees for saving. Produces averaged and multi-integration versions, with sigma-clipped stats over integrations.

Notes

Based on ObservablesFromText from ImPlaneIA, e.g. https://github.com/anand0xff/ImPlaneIA/blob/master/nrm_analysis/misctools/implane2oifits.py#L32

Initialize the RawOifits object.

Parameters:

instrument_datajwst.ami.instrument_data.NIRISS object

Information on the mask geometry (namely # holes), instrument, wavelength obs mode.

methodstr

Method to average observables: mean or median. Default mean.

Methods Summary

average_observables(averfunc)	Average all the observables.
cov_r_theta(rr, theta, averfunc)	Calculate covariance in polar coordinates.
err_from_covmat(covmatlist)	Derive observable errors from their covariance matrices.
init_oimodel_arrays(oimodel)	Set dtypes and initialize shapes for AmiOiModel arrays.
initialize_obsarrays()	Initialize arrays of observables to empty arrays.
make_oifits()	Populate AmiOIModel.
observable_covariances(averfunc)	Calculate covariance matrices from each pair of observables.
populate_obsarray(i, nrmslc)	Populate arrays of observables with fringe fitter results.
rotate_matrix(cov_mat, theta)	Rotate a covariance matrix by an angle.

Methods Documentation

average_observables(averfunc)

Average all the observables.

Calculate covariance matrices between fringe amplitudes/fringe phases, and between triple product amps/closure phases, and closure amplitudes/quad phases. Convert r, theta (modulus, phase) to x,y. Calculate cov(x,y). Rotate resulting 2x2 matrix back to r, theta. Take sqrt of relevant covariance element to be error. This must be done with phases in radians.

Parameters:

averfuncfunction

Function for averaging, either np.mean (default) or np.median

Returns:

avg_sqvarray

Averaged squared visibilites

err_sqvarray

Standard error of the mean of averaged squared visibilities

avg_faarray

Averaged fringe (visibility) amplitudes

err_faarray

Standard error of the mean of averaged fringe (visibility) amplitudes

avg_fparray

Averaged fringe phases (rad)

err_fparray

Standard error of the mean of averaged fringe phases (rad)

avg_cparray

Averaged closure phases (rad)

err_cparray

Standard error of the mean of averaged closure phases (rad)

avg_t3amparray

Averaged triple amplitudes

err_t3amparray

Standard error of the mean of averaged triple amplitudes

avg_caarray

Averaged closure amplitudes

err_caarray

Standard error of the mean of averaged closure amplitudes

avg_q4phiarray

Averaged quad phases

err_q4phiarray

Standard error of the mean of averaged quad phases

avg_pistarray

Averaged segment pistons

err_pistarray

Standard error of the mean of averaged segment pistons

cov_r_theta(rr, theta, averfunc)

Calculate covariance in polar coordinates.

Calculate covariance in x, y coordinates, then rotate covariance matrix by average phase (over integrations) to get matrix in (r,theta).

Parameters:

rrarray

Complex number modulus

thetaarray

Complex number phase

averfuncfunction

Function for averaging, either np.mean (default) or np.median

Returns:

cov_mat_r_thetaarray (2,2)

Covariance matrix in r, theta coordinates

err_from_covmat(covmatlist)

Derive observable errors from their covariance matrices.

Return sqrt of [0,0] and [1,1] elements of each of a list of covariance matrices, divided by sqrt(N_ints), for use as observable errors (standard error of the mean). If using median, error calculation is questionable because this is NOT the standard error of the median.

Parameters:

covmatlistarray

Array of covariance matrices for each baseline/triple/quad shape e.g. (21,2,2) or (35,2,2)

Returns:

err_00array

Standard errors of the mean of the first observable. shape e.g. (21)

err_11array

Standard errors of the mean of the second observable. shape e.g. (21)

init_oimodel_arrays(oimodel)

Set dtypes and initialize shapes for AmiOiModel arrays.

Supports averaged or multi-integration versions of oimodel.

Parameters:

oimodelAmiOIModel object

Empty model

initialize_obsarrays()

Initialize arrays of observables to empty arrays.

make_oifits()

Populate AmiOIModel.

Perform final manipulations of observable arrays, calculate uncertainties, and populate AmiOIModel.

Returns:

mAmiOIModel

Fully populated datamodel

observable_covariances(averfunc)

Calculate covariance matrices from each pair of observables.

For each baseline/triple/quad, calculate covariance between each fringe amplitude/phase quantity.

Parameters:

averfuncfunction

Function for averaging, either np.mean (default) or np.median

Returns:

cov_mat_fringesarray

Array of 21 covariance matrices for fringes (amplitudes, phases)

cov_mat_triplesarray

Array of 35 covariance matrices for triples (t3 amplitudes, closure phases)

cov_mat_quadsarray

Array of 35 covariance matrices for quads (closure amplitudes, quad phases)

populate_obsarray(i, nrmslc)

Populate arrays of observables with fringe fitter results.

Parameters:

iint

Index of the integration

nrmslcobject

Object containing the results of the fringe fitting for this integration

rotate_matrix(cov_mat, theta)

Rotate a covariance matrix by an angle.

Parameters:

cov_matarray

The matrix to be rotated. 2x2

thetafloat

Angle by which to rotate the matrix (radians)

Returns:

cv_rotatedarray

The rotated matrix