Improve float representation accuracy by increasing precision from 8 to 9 decimal digits## Description

[ZhichuCen]

Using 8 digits to represent a float may lead to accuracy loss during decimal-to-binary conversion.

Problem

When floats are represented with only 8 decimal digits, the precision can be insufficient, potentially causing inaccuracies during decimal-to-binary conversions.

Solution

Increase the precision to 9 decimal digits, ensuring more accurate float representation and minimizing the risk of conversion errors.

Impact

This change improves numerical accuracy in scenarios requiring high precision, without introducing significant computational overhead.

[christophe0606]

I'll merge it as soon as the issues with the github actions have been corrected (issues not related to this PR but preventing this PR from being checked)