Giters

NVIDIA / MatX

An efficient C++17 GPU numerical computing library with Python-like syntax

Home Page:https://nvidia.github.io/MatX

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

[FEA] Configurable Print Accuracy for MatX Print Functions

tylera-nvidia opened this issue · comments

commented

The printed formats for data types are currently hardcoded per element data type. An override capability needs to be provided to allow the user to select the precision and format of each element for their use case, such as providing additional precision in the print.

example:
float types are currently limited to 4 significant digits after the decimal, however comparison to alternative implementations may require additional accuracy.