Epsilon: Numpy support

Created on 15 Feb 2020  路  3Comments  路  Source: numworks/epsilon

Problem you'd like to fix

Enhanced support for N-dimensional arrays calculation.

Describe the solution you'd like

Add support as a library to Numpy. At least in the SDK, so that developers could optionally build a sdk branch with numpy included.

Describe alternatives you've considered

At this time there is no comparable alternative to Numpy in Python.

Additional context

To operate n-dimensional arrays on the calculator, would help to build simple machine learning and neural network based algorithms (from Gradient Descend implementation of Loss and Activation functions).

enhancement

Most helpful comment

I actually have a pending pull request for Omega (which is a fork of epsilon that extends its features): Pull request

All 3 comments

Numpy is probably impossible, but maybe this could fit: https://github.com/v923z/micropython-ulab

@nsajko Amazing! I was not aware about ulab! That would be a great starting point!

I actually have a pending pull request for Omega (which is a fork of epsilon that extends its features): Pull request

Was this page helpful?
0 / 5 - 0 ratings