[BUG] Unstable distribution in get_gauss_dist_value

[Sinitca-Aleksandr]

Sometimes the resulting distribution is not normal.

=================================== FAILURES ===================================
_______________ TestPybind11Bindings.test_gaussian_distribution ________________

self = <test_pybind11_bindings.TestPybind11Bindings object at 0x10a326f90>
sample_data = {'C0_input': -1.0, 'U': array([0.5, 0.6, 0.7, 0.8, 0.9]), 'input_vector': array([0.93853618, 6.02024286, 2.63349139, 1...9, 2.49825269, 2.86078551, 1.64777499, 2.94777993,
       1.36162946, 1.47935768, 1.11562839, 0.05149555, 0.22833487])}

    def test_gaussian_distribution(self, sample_data):
        """Test Gaussian distribution value generation"""
        if not NEW_API_AVAILABLE:
            pytest.skip("New API not available")
    
        # Generate 50 values and test statistical properties
        values = [StatTools_bindings.get_gauss_dist_value() for _ in range(500)]
    
        # All values should be floats
        assert all(isinstance(x, float) for x in values)
    
        # Test for normality using Shapiro-Wilk test
        # Null hypothesis: the sample comes from a normal distribution
        # We want to fail to reject the null hypothesis (p-value > 0.05)
        from scipy.stats import shapiro
    
        stat, p_value = shapiro(values)
>       assert (
            p_value > 0.05
        ), f"Values do not appear to be normally distributed (p-value: {p_value:.4f})"
E       AssertionError: Values do not appear to be normally distributed (p-value: 0.0131)
E       assert np.float64(0.013082312439144999) > 0.05

tests/test_pybind11_bindings.py:83: AssertionError

To Reproduce
Steps to reproduce the behavior:

1. Do automatic test