The total variance retained in the 2D PCA dataset is 16.18%. This means that the first two principal components together capture approximately 16.18% of the total variance in the original dataset.

The total variance retained in the 3D PCA dataset is 24.14%. This means that the first three principal components together capture approximately 24.14% of the total variance in the original dataset.

To retain at least 95% of the total variance in the dataset, 13 principal components are required. This indicates that the dataset has high dimensionality, and reducing it to fewer components (e.g., 2 or 3) results in significant information loss. Using 13 components ensures that the majority of the original dataset's structure and relationships are preserved while still achieving dimensionality reduction.