Sampling
Parameter space sampling functions.
sample()
from prandtl import sample
X, Y = sample(func, bounds=[(-5, 15), (0.01, 0.1)], n=100, method="lhs", seed=42)
| Parameter |
Type |
Default |
Description |
func |
callable |
— |
Function that takes dict of params → dict of outputs |
bounds |
list[tuple] |
— |
(min, max) for each parameter |
n |
int |
— |
Number of samples |
method |
str |
"lhs" |
"lhs", "uniform", or "sobol" |
seed |
int |
None |
Random seed |
| Returns |
Type |
Description |
X |
ndarray (n, p) |
Parameter samples |
Y |
ndarray (n, q) |
Output values |
Methods
| Method |
Description |
Best for |
"lhs" |
Latin Hypercube Sampling |
General surrogate modeling |
"uniform" |
Uniform random |
Validation sets |
"sobol" |
Sobol sequence |
Deterministic, quasi-random |