edmkit

uv add edmkit

Every prediction method in edmkit has the same shape: learn from library pairs X → Y, then predict at query points Q.

import numpy as np

from edmkit.simplex_projection import simplex_projection
from edmkit.smap import smap

rng = np.random.default_rng(0)
X = rng.uniform(-1, 1, size=(500, 2))  # library inputs
Y = np.sin(3 * X[:, 0]) * X[:, 1]  # library targets
Q = rng.uniform(-1, 1, size=(5, 2))  # query points

print(simplex_projection(X, Y, Q))
print(smap(X, Y, Q, theta=2.0))

Inputs can be NumPy arrays or tinygrad Tensors. Results are tested against pyEDM.