demmap#

demregpy.demmap(
dd,
ed,
rmatrix,
logt,
dlogt,
glc,
reg_tweak=1.0,
max_iter=10,
rgt_fact=1.5,
dem_norm0=None,
nmu=42,
warn=False,
l_emd=False,
)[source]#

Recover DEMs for a stack of one-dimensional observations.

Each row of dd is treated as an independent observation with the same temperature response matrix. This function is the lower-level workhorse used by demregpy.dn2dem() after the input arrays have been reshaped.

Parameters:

  • dd (array_like) – Input counts with shape (na, nf).

  • ed (array_like) – Uncertainties on dd with the same shape.

  • rmatrix (array_like) – Response matrix with shape (nt, nf).

  • logt (array_like) – Temperature-bin centres in log10(T).

  • dlogt (array_like) – Width of each temperature bin in log10(T).

  • glc (array_like) – Length-nf 0/1 mask selecting the filters used for EM loci weighting.

  • reg_tweak (float, optional) – Initial chisq target. Default is 1.0.

  • max_iter (int, optional) – Maximum number of times to attempt the gsvd before giving up, returns the last attempt if max_iter reached. Default is 10.

  • rgt_fact (float, optional) – Scale factor for the increase in chi-sqaured target for each iteration. Default is 1.5.

  • dem_norm0 (array_like, optional) – Provides a “guess” dem as a starting point, if none is supplied one is created. Default is None.

  • nmu (int, optional) – Number of reg param samples to use. Default is 42.

  • warn (bool, optional) – Print out warnings. Default is False.

  • l_emd (bool, optional) – Remove sqrt from constraint matrix (best with EMD). Default is False.

Returns:

  • dem (ndarray) – DEM values with shape (na, nt).

  • edem (ndarray) – Vertical uncertainties on dem.

  • elogt (ndarray) – Horizontal temperature resolution estimates in log10(T).

  • chisq (ndarray) – Reduced chi-squared values for each observation.

  • dn_reg (ndarray) – Reconstructed counts with shape (na, nf).

On this page