The package can already tell the variance and loadings of PC but it doesn't output the actual data yet. Users can refer to sklearn.decomposition to run the reduction but it seem like a far transfer in my opinion.pca_transform would include the trio in a single library for users which makes it easy and reusable exactly what this whole package is designed for.
Proposed implementation:
pca_transform(df, decimals = None, n_components=None, standardise=True) -> DataFrame.
Make note of polars implementation too, I would recommend using numpy to follow the implementations of pca_variance and loadings.
I would recommend it to be added to stats.py
The package can already tell the variance and loadings of PC but it doesn't output the actual data yet. Users can refer to
sklearn.decompositionto run the reduction but it seem like a far transfer in my opinion.pca_transform would include the trio in a single library for users which makes it easy and reusable exactly what this whole package is designed for.Proposed implementation:
pca_transform(df, decimals = None, n_components=None, standardise=True) -> DataFrame.Make note of polars implementation too, I would recommend using numpy to follow the implementations of
pca_varianceandloadings.I would recommend it to be added to stats.py