bucky.util.scoring#

Provides various scoring metrics for probabilistic forecasts.

Module Contents#

Functions#

IS(x, lower, upper, alp)

Interval score.

WIS(x, q, x_q, norm=False, log=False, smooth=False)

Weighted Interval Score.

logistic(x, x0=0.0, k=1.0, L=1.0)

Logistic function.

smooth_IS(x, lower, upper, alp)

Approx Interval Score with smooth derivatives.

bucky.util.scoring.IS(x, lower, upper, alp)[source]#

Interval score.

Parameters:
  • x

  • lower

  • upper

  • alp

Return type:

numpy.ndarray or cupy.ndarray if using CuPy

bucky.util.scoring.WIS(x, q, x_q, norm=False, log=False, smooth=False)[source]#

Weighted Interval Score.

Parameters:
  • x

  • q

  • x_q

  • norm (bool, optional) –

  • log (bool, optional) –

  • smooth (bool, optional) –

Return type:

numpy.ndarray or cupy.ndarray if using CuPy

bucky.util.scoring.logistic(x, x0=0.0, k=1.0, L=1.0)[source]#

Logistic function.

Parameters:
  • x

  • x0

  • k

  • L

Return type:

numpy.ndarray or cupy.ndarray if using CuPy

bucky.util.scoring.smooth_IS(x, lower, upper, alp)[source]#

Approx Interval Score with smooth derivatives.

Parameters:
  • x

  • lower

  • upper

  • alp

Return type:

numpy.ndarray or cupy.ndarray if using CuPy