aquakin.ProfileResult#

class aquakin.ProfileResult(profiled, grid, loss, delta_loss, mle, ci, fits, delta)[source]#

Bases: object

Result of profile_likelihood().

Variables:

  • profiled (str) – Name of the profiled quantity (a parameter or a species).

  • grid (np.ndarray) – The fixed values, shape (n_grid,).

  • loss (np.ndarray) – Best attainable objective at each grid value (nan where the inner fit failed).

  • delta_loss (np.ndarray) – loss - min(loss): the profile relative to its minimum, which is what the likelihood-ratio threshold applies to.

  • mle (float) – Grid value at the profile minimum (the maximum-likelihood estimate, to grid resolution).

  • ci (tuple[float | None, float | None]) – (lo, hi) confidence bounds, the interpolated points where delta_loss crosses delta either side of the minimum. None on a side means the profile never crosses the threshold there (the bound is open / the parameter is not identified on that side).

  • fits (list[CalibrationResult | None]) – The re-optimised fit at each grid point, for extracting RMSEs or the re-optimised parameters. None where the inner fit failed.

  • delta (float) – The likelihood-ratio threshold used.

Parameters:

  • profiled (str)

  • grid (ndarray)

  • loss (ndarray)

  • delta_loss (ndarray)

  • mle (float)

  • ci (tuple[float | None, float | None])

  • fits (list)

  • delta (float)

__init__(profiled, grid, loss, delta_loss, mle, ci, fits, delta)#
Parameters:

  • profiled (str)

  • grid (ndarray)

  • loss (ndarray)

  • delta_loss (ndarray)

  • mle (float)

  • ci (tuple[float | None, float | None])

  • fits (list)

  • delta (float)

Return type:

None

Methods

__init__(profiled, grid, loss, delta_loss, ...)

Attributes

profiled

grid

loss

delta_loss

mle

ci

fits

delta