allow test/query data to be used with Transform() API call

[Stevod]

Currently, it is not obvious how to apply a fitted PaCMAP model to a test dataset. Although a transform() call is available, it is not obvious how that is used, nor the applicable syntax, so some documentation on that would be appreciated.

[hyhuang00]

Currently the documentation website is under construction, but the docstrings are already available within the source code. I will copy paste the documentation for the transform() method here as a reference:

Projects a high dimensional dataset into existing embedding space and return the embedding.

    Parameters
    ---------
    X: numpy.ndarray
        The new high-dimensional dataset that is being projected. 
        An embedding will get created based on parameters of the PaCMAP instance.

    basis: numpy.ndarray
        The original dataset that have already been applied during the `fit` or `fit_transform` process.
        If `save_tree == False`, then the basis is required to reconstruct the ANNOY tree instance.
        If `save_tree == True`, then it's unnecessary to provide the original dataset again.

    init: str, optional
        One of ['pca', 'random']. Initialization of the embedding, default='pca'.
        If 'pca', then the low dimensional embedding is initialized to the PCA mapped dataset. 
        The PCA instance will be the same one that was applied to the original dataset during the `fit` or `fit_transform` process. 
        If 'random', then the low dimensional embedding is initialized with a Gaussian distribution.

    save_pairs: bool, optional
        Whether to save the pairs that are sampled from the dataset. Useful for reproducing results.