tinydot is a Python wrapper for my tiny linear algebra library writen in C. The wrapper interacts with C code via ctypes.

Everything in tinydot is based on nd Tensor structure, i.e. a matrix is a 2d tensor with shape [rows, cols].

tinydot also support basic operations on Tensors.

If you would like to play around with this framework, you can install it using pip directly from GitHub by running

pip install git+https://github.com/krylowicz/tinydot.git`.

Here's an example:

import tinydot as td
 
v1 = td.Vector(1, 3)
v2 = td.Vector(2, 4)

dot_product = td.Vector.dot(v1, v2)
v1_norm = v1.norm
v2.rotate(47) #inplace counterclockwise vector rotation by theta degrees (use -theta for clockwise)

matrix = td.Matrix([[1, 2], [3, 4], [5, 6]])
matrix.T #transposes the matrix inplace
trace = matrix.trace
matrix_norm = matrix.norm

Tinydot is written in test driven development fashion using pytest. To run the tests execute the following command:

$ python -m pytest

You should run the command above in the root directory