EMToolKit.algorithms package

EMToolKit.algorithms.multi_inst_dem_wrapper module

EMToolKit.algorithms.simple_reg_dem module

EMToolKit.algorithms.simple_reg_dem.simple_reg_dem(data, errors, exptimes, logt, tresps, kmax=100, kcon=5, steps=[0.1, 0.5], drv_con=8.0, chi2_th=1.0, tol=0.1)[source]

EMToolKit.algorithms.simple_reg_dem_parallel module

EMToolKit.algorithms.simple_reg_dem_parallel.precompute_matrices(data, errors, exptimes, logt, tresps, drv_con)[source]

Precomputes matrices and vectors required for regularized differential emission measure.

Parameters: - data: The observed data array. - errors: The uncertainties associated with the data. - exptimes: Exposure times for each observation. - logt: Logarithm of temperature bins. - tresps: Temperature response functions. - drv_con: Drive constant for regularization.

Returns: - nt_ones: Array of ones of size nt. - nx, ny: Dimensions of the data array. - Bij: Matrix used in regularization. - Rij: Matrix mapping coefficients to data. - Dij: Matrix used in regularization. - regmat: Regularization matrix. - rvec: Sum of Rij across axis=1.

EMToolKit.algorithms.simple_reg_dem_parallel.process_pixel(i, j, data, errors, Rij, regmat, rvec, nt_ones, kmax, kcon, steps, drv_con, chi2_th, tol)[source]

Processes a single pixel (i, j) in a dataset to compute its differential emission measure (DEM) and chi-squared value.

Parameters: - i, j: Indices of the pixel in the data array. - data: The observed data array. - errors: The uncertainties associated with the data. - Rij: Matrix mapping coefficients to data. - regmat: Regularization matrix. - rvec: Sum of Rij across axis=1. - nt_ones: Array of ones of size nt. - kmax: Maximum number of iterations. - kcon: Convergence criterion based on the number of iterations. - steps: Step sizes for the iterative method. - drv_con: Drive constant for regularization. - chi2_th: Chi-squared threshold for stopping criterion. - tol: Tolerance for convergence criterion.

Returns: - i, j: Indices of the processed pixel. - dem_pixel: Computed DEM value for the pixel. - chi2_pixel: Computed chi-squared value for the pixel.

Notes: - The function skips pixels with NaN values in data or errors. - If the Cholesky factorization fails, the iteration breaks early. - The function employs an iterative method to adjust the DEM based on the observed data, errors, and a regularization term.

EMToolKit.algorithms.simple_reg_dem_parallel.simple_reg_dem(*args, doParallel=True, **kwargs)[source]

Wrapper function to call the serial or parallel version of the simple regularized DEM computation.

Accepts the same parameters as simple_reg_dem_serial and simple_reg_dem_parallel.

Returns: - dems: Computed DEM values, a 3D array (nx, ny, nt). - chi2: Computed chi-squared values for each pixel, a 2D array (nx, ny).

EMToolKit.algorithms.simple_reg_dem_parallel.simple_reg_dem_parallel(data, errors, exptimes, logt, tresps, kmax=100, kcon=5, steps=[0.1, 0.5], drv_con=8.0, chi2_th=1.0, tol=0.1)[source]

Perform a parallel computation of differential emission measure (DEM) and chi-squared values for observational data against a set of response functions using multiple processes.

Parameters are the same as in simple_reg_dem_serial.

Returns: - dems: Computed DEM values, a 3D array (nx, ny, nt). - chi2: Computed chi-squared values for each pixel, a 2D array (nx, ny).

EMToolKit.algorithms.simple_reg_dem_parallel.simple_reg_dem_serial(data, errors, exptimes, logt, tresps, kmax=100, kcon=5, steps=[0.1, 0.5], drv_con=8.0, chi2_th=1.0, tol=0.1)[source]

Perform a serial computation of differential emission measure (DEM) and chi-squared values for observational data against a set of response functions.

Parameters: - data: Observed data, a 3D array (nx, ny, nd). - errors: Uncertainties in the observed data, same shape as data. - exptimes: Exposure times for each observation, 1D array of length nd. - logt: Logarithm of temperature bins, 1D array. - tresps: Temperature response functions, 2D array (nt, nd). - kmax: Maximum number of iterations for the fitting process. - kcon: Minimum number of iterations before convergence can be considered. - steps: Step sizes to use in the fitting iterations, list of two floats. - drv_con: Drive constant for the regularization term. - chi2_th: Threshold chi-squared value for stopping the iterations. - tol: Tolerance for convergence.

Returns: - dems: Computed DEM values, a 3D array (nx, ny, nt). - chi2: Computed chi-squared values for each pixel, a 2D array (nx, ny).

EMToolKit.algorithms.simple_reg_dem_wrapper module

EMToolKit.algorithms.simple_reg_dem_wrapper_parallel module

EMToolKit.algorithms.simple_reg_dem_wrapper_parallel.autoloading_simple_reg_dem_wrapper(datasequence, data_dir='../', recalc_simple=False, wrapargs={})[source]
EMToolKit.algorithms.simple_reg_dem_wrapper_parallel.simple_reg_dem_wrapper(datasequence, wrapargs=None)[source]

EMToolKit.algorithms.sparse_em module

EMToolKit.algorithms.sparse_em.simplex(zequation, constraint, m1, m2, m3, eps=None)[source]
EMToolKit.algorithms.sparse_em.sparse_em_init(trlogt_list, tresp_list, bases_sigmas=None, differential=False, bases_powers=[], normalize=False, use_lgtaxis=None)[source]
EMToolKit.algorithms.sparse_em.sparse_em_solve(image, errors, exptimes, Dict, zfac=[], eps=0.001, tolfac=1.4, relax=True, symmbuff=1.0, adaptive_tolfac=True, epsfac=1e-10)[source]

EMToolKit.algorithms.sparse_em_wrapper module

EMToolKit.algorithms.sparse_em_wrapper.autoloading_sparse_em_wrapper(datasequence, data_dir='.data/default', *, wrapargs={}, recalc=False)[source]

Wrapper function that calculates or loads a precomputed sparse regularized DEM.

This function first checks if a precomputed DEM exists in the specified directory. If not, it calculates the DEM using sparse_em_wrapper, saves the result, and returns it.

Parameters:

  • datasequence (NDCubeSequence) – A sequence of data cubes containing the observations. Each cube should contain 2D spatial data with associated uncertainties and metadata.

  • data_dir (str, optional) – The directory where the DEM result will be saved or loaded from. Default is “../.data/default”.

  • recalc (bool, optional) – If True, the DEM will be recalculated even if a precomputed result exists. Default is False.

  • wrapargs (A dictionary)

Returns:

  • emtk.dem_model – The DEM model generated from the input data.

  • tuple – The output from the sparse_em_wrapper function.

EMToolKit.algorithms.sparse_em_wrapper.sparse_em_wrapper(datasequence, wrapargs={})[source]

Wrapper function for the sparse regularized Differential Emission Measure (DEM) calculation.

This function prepares input data and passes it to the sparse_em_init and sparse_em_solve algorithms. It processes the input data to ensure that all input maps have consistent dimensions, extracts the necessary metadata, and then calls the DEM calculation.

Parameters:

  • datasequence (NDCubeSequence) – A sequence of data cubes containing the observations. Each cube should contain 2D spatial data with associated uncertainties and metadata.

  • wrapargs (dict, optional) – Additional arguments to pass to the initialization routines of the sparse_em functions.

Returns:

  • list of numpy.ndarray – The calculated DEM coefficients for each temperature bin.

  • list of numpy.ndarray – The temperature bins used in the calculation.

  • list of numpy.ndarray – The basis functions for the temperature bins.

  • WCS – World Coordinate System (WCS) information from the input data.

  • str – A string identifier for the DEM method used.

EMToolKit.algorithms.sparse_nlmap_dem_wrapper module

EMToolKit.algorithms.sparse_nlmap_dem_wrapper.autoloading_sparse_nlmap_dem_wrapper(datasequence, data_dir='.data/default/', recalc=False, wrapargs={})[source]

Wrapper function that calculates or loads a precomputed sparse non-linear map-based DEM.

This function first checks if a precomputed DEM exists in the specified directory. If not, it calculates the DEM using sparse_nlmap_dem_wrapper, saves the result, and returns it.

Parameters:

  • datasequence (NDCubeSequence) – A sequence of data cubes containing the observations. Each cube should contain 2D spatial data with associated uncertainties and metadata.

  • data_dir (str, optional) – The directory where the DEM result will be saved or loaded from. Default is “.data/default/”.

  • recalc (bool, optional) – If True, the DEM will be recalculated even if a precomputed result exists. Default is False.

  • wrapargs (dict, optional) – Additional arguments to pass to the initialization routines of the sparse_nlmap_dem_wrapper function.

Returns:

  • emtk.dem_model – The DEM model generated from the input data.

  • tuple – The output from the sparse_nlmap_dem_wrapper function.

EMToolKit.algorithms.sparse_nlmap_dem_wrapper.minmax(arg)[source]
EMToolKit.algorithms.sparse_nlmap_dem_wrapper.process_data_element(i, element, source)[source]
class EMToolKit.algorithms.sparse_nlmap_dem_wrapper.sparse_nlmap_dem_wrap_object(wrapargs, source=None)[source]

Bases: object

compute_dem(datasequence, wrapargs=None)[source]
EMToolKit.algorithms.sparse_nlmap_dem_wrapper.sparse_nlmap_dem_wrapper(datasequence, wrapargs={})[source]

Wrapper function for the sparse non-linear map-based Differential Emission Measure (DEM) calculation.

This function prepares input data and passes it to the sparse_nlmap_dem_wrapper algorithm. It processes the input data to ensure that all input maps have consistent dimensions, extracts the necessary metadata, and then calls the DEM calculation.

Parameters:

  • datasequence (NDCubeSequence) – A sequence of data cubes containing the observations. Each cube should contain 2D spatial data with associated uncertainties and metadata.

  • wrapargs (dict, optional) – Additional arguments to pass to the initialization routines of the sparse_nlmap_dem_wrapper function.

Returns:

  • list of numpy.ndarray – The calculated DEM coefficients for each temperature bin.

  • list of numpy.ndarray – The temperature bins used in the calculation.

  • list of numpy.ndarray – The basis functions for the temperature bins.

  • WCS – World Coordinate System (WCS) information from the input data.

  • str – A string identifier for the DEM method used.

EMToolKit.algorithms.sparse_nlmap_solver module

class EMToolKit.algorithms.sparse_nlmap_solver.nlmap_operator(*args, **kwargs)[source]

Bases: LinearOperator

setup(amat, regmat, map_drvvec, wgtvec, reg_map_drvvec, dtype='float32', reg_fac=1)[source]
EMToolKit.algorithms.sparse_nlmap_solver.solve(data0, errors0, amat0, guess=None, reg_fac=1, func=None, dfunc=None, ifunc=None, regmat=None, silent=False, solver=None, sqrmap=False, regfunc=None, dregfunc=None, iregfunc=None, map_reg=False, adapt_lam=True, solver_tol=0.001, niter=40, dtype='float32', steps=None, precompute_ata=False, flatguess=True, chi2_th=1.0, store_outer_Av=False, conv_chi2=1e-15)[source]

Module contents