Usage

Ember provides a Tensor class and various mathematical operations.

Creating Tensors

You can create tensors from Python lists or NumPy arrays.

import ember as em
import numpy as np

from ember import Tensor

# From list
a = Tensor([1, 2, 3])

# From NumPy
np_arr = np.array([1.0, 2.0, 3.0], dtype=np.float32)
b = Tensor(np_arr) # Automatically converts from buffer

Operations

Ember supports element-wise operations and broadcasting with scalars.

Arithmetic

c = a + b
d = a * 2.0
e = em.max(a, b)

Trigonometric & Hyperbolic

# Standard trig
s = em.sin(a)
c = em.cos(a)

# Hyperbolic
sh = em.sinh(a)
th = em.tanh(a)

Matrix Multiplication

# Create 2x2 matrices (flattened logic for now, or reshape support pending)
# Note: Current implementation is basic.
pass

Data Exchange

Convert back to standard Python/NumPy types easily.

# To List
data_list = c.to_list()

# To NumPy
data_np = c.to_np()