pystrata.variation.SoilTypeVariation

class pystrata.variation.SoilTypeVariation(correlation, limits_mod_reduc=[0.05, 1], limits_damping=[0, 0.15], vary_bedrock=False, sample_mode='random', percentiles=None)

Base class for soil-type (modulus-reduction and damping) variation.

Parameters:

• correlation (float) – Correlation coefficient between the modulus-reduction and damping random variables.

• limits_mod_reduc (list[float], optional) – [min, max] clipping bounds for modulus reduction.

• limits_damping (list[float], optional) – [min, max] clipping bounds for damping.

• vary_bedrock (bool, optional) – Whether to include the half-space in the variation.

• sample_mode ({'random', 'fixed_percentiles'}, optional) –

  How samples are drawn when iter_varied_profiles() iterates:

  • 'random' (default) — each call draws an independent random realisation from the truncated-normal distribution.

  • 'fixed_percentiles' — each call uses a pre-specified percentile supplied via sample_index so that the same index always produces an identical realisation. The percentile is selected as percentiles[sample_index % len(percentiles)], so the list cycles when count is a multiple of its length. The caller must pass the sample_index keyword argument to __call__(), and iter_varied_profiles() does this automatically.

• percentiles (list[float] | None, optional) – Ordered sequence of quantiles in (0, 1) used in 'fixed_percentiles' mode. The sequence cycles: iteration i draws percentiles[i % len(percentiles)]. Required (and only used) when sample_mode is 'fixed_percentiles'.

__init__(correlation, limits_mod_reduc=[0.05, 1], limits_damping=[0, 0.15], vary_bedrock=False, sample_mode='random', percentiles=None)

Methods

__init__(correlation[, limits_mod_reduc, ...])
vary_profile(profile[, sample_index])	Return a profile with varied soil types.

Attributes

correlation
limits_damping
limits_mod_reduc
percentiles	Percentile list used in 'fixed_percentiles' mode, or None.
sample_mode	'random' or 'fixed_percentiles'.
vary_bedrock