analysis.dimensionality_reduction.pca_dim_reduction

Methods for dimensionality reduction using PCA.

Functions

plot_pca_feature_scatter(data, features, pca)

Plot scatter of PCA components colored by the given features.

plot_pca_inverse_transform(pca, pca_results)

Plot inverse transform of PCA.

run_pca(data)

Run Principal Component Analysis (PCA) on simulation data.

save_pca_results(pca_results, save_location, ...)

Save PCA results data.

save_pca_trajectories(pca_results, ...)

Save PCA trajectories data.

save_pca_transforms(pca, points, ...)

Save PCA transform data.
run_pca(data: DataFrame) → tuple[DataFrame, PCA][source]

Run Principal Component Analysis (PCA) on simulation data.

Parameters:

data – Simulated fiber data.

Returns:

Dataframe with PCA components appended and the PCA object.

save_pca_results(pca_results: DataFrame, save_location: str, save_key: str, resample: bool = True) → None[source]

Save PCA results data.

Parameters:

  • pca_results – PCA trajectory data.

  • save_location – Location for output file (local path or S3 bucket).

  • save_key – Name key for output file.

  • resample – True if data should be resampled before saving, False otherwise.

save_pca_trajectories(pca_results: DataFrame, save_location: str, save_key: str) → None[source]

Save PCA trajectories data.

Parameters:

  • pca_results – PCA trajectory data.

  • save_location – Location for output file (local path or S3 bucket).

  • save_key – Name key for output file.

save_pca_transforms(pca: PCA, points: list[list[float]], save_location: str, save_key: str) → None[source]

Save PCA transform data.

Parameters:

  • pca – PCA object.

  • points – List of inverse transform points.

  • save_location – Location for output file (local path or S3 bucket).

  • save_key – Name key for output file.

plot_pca_feature_scatter(data: DataFrame, features: dict, pca: PCA, save_location: str | None = None, save_key: str = 'pca_feature_scatter.png') → None[source]

Plot scatter of PCA components colored by the given features.

Parameters:

  • data – PCA results data.

  • features – Map of feature name to coloring.

  • pca – PCA object.

  • save_location – Location for output file (local path or S3 bucket).

  • save_key – Name key for output file.

plot_pca_inverse_transform(pca: PCA, pca_results: DataFrame, save_location: str | None = None, save_key: str = 'pca_inverse_transform.png') → None[source]

Plot inverse transform of PCA.

Parameters:

  • pca – PCA object.

  • pca_results – PCA results data.

  • save_location – Location for output file (local path or S3 bucket).

  • save_key – Name key for output file.