This function is used to calculate the binned density similar to np.histogram(., density=True) not just for one dimension, but for multiple in parallel.
The input samples is assumed to have the shape: [n_samples, n_dimensions]
The output will be:
sample_densities: The density of each sample (shape: [n_samples, n_dimensions])density: The density for each dimension (shape: [n_bins + 1, n_dimensions])bins: The equally distanced bins edges, controlled byn_bins(shape [n_bins + 1, n_dimensions])