Metadata-Version: 2.4
Name: cumpy
Version: 0.1.5.1
Summary: CumPy numerical library for Python
Author: Świadomy Mężczyzna
License-Expression: Apache-2.0
Project-URL: Homepage, https://brain.mosche.dev
Requires-Python: >=3.9
Requires-Dist: dpcpp-cpp-rt<2027.0,>=2026.0
Requires-Dist: intel-cmplr-lib-rt<2027.0,>=2026.0
Requires-Dist: intel-opencl-rt<2027.0,>=2026.0
Requires-Dist: mkl<2027.0,>=2026.0
Requires-Dist: onemkl-sycl-blas<2027.0,>=2026.0
Requires-Dist: onemkl-sycl-rng<2027.0,>=2026.0
Requires-Dist: onemkl-sycl-vm<2027.0,>=2026.0
Requires-Dist: tbb>=2021.11
Requires-Dist: umf<2.0,>=1.0
Description-Content-Type: text/markdown

# CumPy - Python bindings for CUM

Lightweight python binding over CUM library.

## Quickstart

### Installation

```sh
pip install cumpy
```

### Basic usage

```python
import cumpy as cp

cp.init(cp.Device.CPU)

a = cp.Tensor.from_data((2, 2), [1, 2, 3, 4])
b = cp.Tensor.ones((2, 2)) * 2.0

print(a + b)   # addition
print(a * b)   # element-wise multiplication
print(a @ b)   # matrix product
print(a.sum(), a.mean(), a.norm())

cp.shutdown()
```

## Examples

- `examples/basic_tensors.py` – creating tensors, arithmetics, unary functions, reductions, transpositions and `reshape`.
- `examples/matrix_and_vector.py` – operations on `Matrix` i `Vector`, indexing `m[r, c]` / `v[i]`, dot product, etc.
- `examples/linear_regression.py` – linear regresion trained with gradient descent using `Tensor`.
