API Reference#

Top-level package for SpectoPrep. SpectroPrep: A comprehensive toolkit for spectroscopic data preprocessing and modeling.

This package provides tools for preprocessing spectroscopic data, pipeline optimization, and modeling using Ridge regression.

class spectoprep.OptimizedRidgeCV(alphas=None, cv=5, scoring='neg_mean_squared_error', fit_intercept=True, normalize=False, gcv_mode=None, store_cv_values=False, groups=None)[source]#

Bases: BaseEstimator, RegressorMixin

Ridge regression with built-in cross-validation and optimization capabilities.

Parameters#

alphasarray-like, default=np.logspace(-3, 3, 10): Array of alpha values to try. A large array of values will slow down the computation.
cvint, cross-validation generator or an iterable, default=5: Determines the cross-validation splitting strategy.
scoringstr, callable, default=’neg_mean_squared_error’: A string or a scorer callable object / function with signature scorer(estimator, X, y).
fit_interceptbool, default=True: Whether to calculate the intercept for this model.
normalizebool, default=False: This parameter is ignored when fit_intercept is set to False. If True, the regressors X will be normalized before regression by subtracting the mean and dividing by the l2-norm.
gcv_mode{None, ‘auto’, ‘svd’, ‘eigen’}, default=None: Flag indicating which strategy to use when performing Generalized Cross-Validation.
store_cv_valuesbool, default=False: Flag indicating if the cross-validation values corresponding to each alpha should be stored in the cv_values_ attribute.
groupsarray-like, default=None: Group labels for the samples. Only used if cv is a group-based cross-validation splitter.

fit(X, y, sample_weight=None)[source]#

Fit Ridge regression model with cross-validation.

Parameters#

Xarray-like of shape (n_samples, n_features): Training data.
yarray-like of shape (n_samples,) or (n_samples, n_targets): Target values.
sample_weightfloat or array-like of shape (n_samples,), default=None: Individual weights for each sample.

Returns#

selfobject: Returns self.

get_cv_results()[source]#

Return cross-validation results.

Returns#

cv_resultsdict: Results from cross-validation.

predict(X)[source]#

Predict using the Ridge model.

Parameters#

Xarray-like of shape (n_samples, n_features): Samples.

Returns#

y_predarray-like of shape (n_samples,) or (n_samples, n_targets): Returns predicted values.

score(X, y, sample_weight=None)[source]#

Return the coefficient of determination R^2 of the prediction.

Parameters#

Xarray-like of shape (n_samples, n_features): Test samples.
yarray-like of shape (n_samples,) or (n_samples, n_targets): True values for X.
sample_weightarray-like of shape (n_samples,), default=None: Sample weights.

Returns#

scorefloat: R^2 of self.predict(X) wrt. y.

set_fit_request(*, sample_weight: bool | None | str = '$UNCHANGED$') → OptimizedRidgeCV#

Configure whether metadata should be requested to be passed to the fit method.

Note that this method is only relevant when this estimator is used as a sub-estimator within a meta-estimator and metadata routing is enabled with enable_metadata_routing=True (see sklearn.set_config()). Please check the User Guide on how the routing mechanism works.

The options for each parameter are:

True: metadata is requested, and passed to fit if provided. The request is ignored if metadata is not provided.
False: metadata is not requested and the meta-estimator will not pass it to fit.
None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.
str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters#

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for sample_weight parameter in fit.

Returns#

selfobject: The updated object.

set_score_request(*, sample_weight: bool | None | str = '$UNCHANGED$') → OptimizedRidgeCV#

Configure whether metadata should be requested to be passed to the score method.

The options for each parameter are:

True: metadata is requested, and passed to score if provided. The request is ignored if metadata is not provided.
False: metadata is not requested and the meta-estimator will not pass it to score.
None: metadata is not requested, and the meta-estimator will raise an error if the user provides it.
str: metadata should be passed to the meta-estimator with this given alias instead of the original name.

The default (sklearn.utils.metadata_routing.UNCHANGED) retains the existing request. This allows you to change the request for some parameters and not others.

Added in version 1.3.

Parameters#

sample_weightstr, True, False, or None, default=sklearn.utils.metadata_routing.UNCHANGED: Metadata routing for sample_weight parameter in score.

Returns#

selfobject: The updated object.

class spectoprep.PipelineOptimizer(X_train: ndarray[tuple[Any, ...], dtype[_ScalarT]], y_train: ndarray[tuple[Any, ...], dtype[_ScalarT]], preprocessing_steps: List[str] | None = None, X_test: ndarray[tuple[Any, ...], dtype[_ScalarT]] | None = None, y_test: ndarray[tuple[Any, ...], dtype[_ScalarT]] | None = None, cv_method: str = 'group_shuffle_split', n_splits: int = 3, test_size: float = 0.3, n_groups_out: int = 2, random_state: int = 42, groups: ndarray[tuple[Any, ...], dtype[_ScalarT]] | None = None, max_pipeline_length: int = 5, n_jobs: int = -1, allowed_preprocess_combinations: int | List[int] | Tuple[int, ...] | None = [1, 2], log_level: str = 'INFO')[source]#

Bases: object

A class for optimizing machine learning pipelines using Bayesian optimization. It precomputes possible pipeline configurations and then searches over both the pipeline configuration (encoded as an index) and the hyperparameters.

bayes_objective(**params) → float[source]#

Objective function for Bayesian optimization.

Args:: **params: Parameters to evaluate
Returns:: float: Negative RMSE or penalty score on error

bayesian_optimize(init_points: int = 10, n_iter: int = 50, acquisition_function: str = 'ei') → Tuple[Dict, Pipeline][source]#

Run Bayesian optimization to find the best pipeline configuration and hyperparameters.

Args:

init_points: Number of random initial points n_iter: Number of Bayesian optimization iterations acquisition_function: Acquisition function for Bayesian optimization

Returns:

Tuple containing:

Dict of best parameters
Fitted Pipeline with best configuration

export_best_pipeline(file_path: str) → None[source]#

Export the best pipeline configuration and hyperparameters to a file.

Args:: file_path: Path to save the export file
Raises:: AttributeError: If optimizer hasn’t been run yet

get_all_tested_pipelines() → List[Dict][source]#

Get details of all tested pipeline configurations.

Returns:: List of dictionaries with pipeline details

get_best_pipeline_predictions(best_pipeline: Pipeline) → Tuple[ndarray[tuple[Any, ...], dtype[_ScalarT]], float, float][source]#

Get predictions using the best pipeline.

Args:

best_pipeline: Fitted pipeline object

Returns:

Tuple containing:

Predictions array
RMSE score
R² score

print_evaluated_pipelines() → None[source]#

Print details for all evaluated pipelines from the Bayesian optimizer.

This method assumes that bayesian_optimize() has been run and that self.optimizer exists.

summarize_optimization() → Dict[source]#

Generate a summary of the optimization results.

Returns:: Dictionary containing optimization summary metrics

class spectoprep.SpectroPrepPlotter[source]#

Bases: object

A class for creating high-quality plots for spectroscopy data.

This class provides various plotting functions specifically designed for spectroscopy data and pipeline optimization results.

static plot_feature_importance(wavenumbers: ndarray, coefficients: ndarray, title: str = 'Feature Importance', xlabel: str = 'Wavenumber (cm$^{-1}$)', ylabel: str = 'Coefficient Value', figsize: Tuple[int, int] = (12, 6), color: str = 'purple', highlight_threshold: float | None = None, highlight_color: str = 'red', save_path: str | None = None)[source]#

Plot feature importance from model coefficients.

Parameters#

wavenumbersarray-like: The x-axis values (wavenumbers).
coefficientsarray-like: Model coefficients corresponding to each wavenumber.
titlestr, default=’Feature Importance’: Plot title.
xlabelstr, default=’Wavenumber (cm$^{-1}$)’: X-axis label.
ylabelstr, default=’Coefficient Value’: Y-axis label.
figsizetuple, default=(12, 6): Figure size.
colorstr, default=’purple’: Color of the line.
highlight_thresholdfloat, optional: If provided, highlights coefficients with absolute values above this threshold.
highlight_colorstr, default=’red’: Color for highlighted coefficients.
save_pathstr, optional: If provided, save the figure to this path.

Returns#

figmatplotlib.figure.Figure: The figure object.
axmatplotlib.axes.Axes: The axes object.

static plot_optimization_progress(optimizer: PipelineOptimizer, figsize: Tuple[int, int] = (12, 6), title: str = 'Optimization Progress', save_path: str | None = None)[source]#

Plot optimization progress over iterations.

Parameters#

optimizerPipelineOptimizer: The fitted pipeline optimizer.
figsizetuple, default=(12, 6): Figure size.
titlestr, default=’Optimization Progress’: Plot title.
save_pathstr, optional: If provided, save the figure to this path.

Returns#

figmatplotlib.figure.Figure: The figure object.
axmatplotlib.axes.Axes: The axes object.

static plot_optimization_results(optimizer: PipelineOptimizer, top_n: int = 5, figsize: Tuple[int, int] = (12, 8), title: str = 'Pipeline Optimization Results', save_path: str | None = None)[source]#

Plot optimization results from PipelineOptimizer.

Parameters#

optimizerPipelineOptimizer: The fitted pipeline optimizer.
top_nint, default=5: Number of top pipelines to display.
figsizetuple, default=(12, 8): Figure size.
titlestr, default=’Pipeline Optimization Results’: Plot title.
save_pathstr, optional: If provided, save the figure to this path.

Returns#

figmatplotlib.figure.Figure: The figure object.

static plot_prediction_scatter(y_true: ndarray, y_pred: ndarray, title: str = 'Prediction Performance', xlabel: str = 'Measured', ylabel: str = 'Predicted', figsize: Tuple[int, int] = (10, 8), alpha: float = 0.7, color: str = 'blue', add_metrics: bool = True, save_path: str | None = None)[source]#

Create a scatter plot of predicted vs true values.

Parameters#

y_truearray-like: True target values.
y_predarray-like: Predicted target values.
titlestr, default=’Prediction Performance’: Plot title.
xlabelstr, default=’Measured’: X-axis label.
ylabelstr, default=’Predicted’: Y-axis label.
figsizetuple, default=(10, 8): Figure size.
alphafloat, default=0.7: Transparency of the points.
colorstr, default=’blue’: Color of the scatter points.
add_metricsbool, default=True: Whether to add RMSE and R² metrics to the plot.
save_pathstr, optional: If provided, save the figure to this path.

Returns#

figmatplotlib.figure.Figure: The figure object.
axmatplotlib.axes.Axes: The axes object.

static plot_preprocessing_comparison(wavenumbers: ndarray, original_spectra: ndarray, processed_spectra: Dict[str, ndarray], sample_indices: List[int] | None = None, figsize: Tuple[int, int] = (15, 10), title: str = 'Preprocessing Comparison', color_map: str = 'tab10', save_path: str | None = None)[source]#

Plot comparison of original and processed spectra.

Parameters#

wavenumbersarray-like: The x-axis values (wavenumbers).
original_spectraarray-like: The original spectra data of shape (n_samples, n_features).
processed_spectradict: Dictionary mapping preprocessing method names to processed spectra.
sample_indiceslist of int, optional: Indices of samples to plot. If None, all samples are plotted.
figsizetuple, default=(15, 10): Figure size.
titlestr, default=’Preprocessing Comparison’: Main title for the figure.
color_mapstr, default=’tab10’: Colormap for differentiating samples.
save_pathstr, optional: If provided, save the figure to this path.

Returns#

figmatplotlib.figure.Figure: The figure object.

static plot_spectra(wavenumbers: ndarray, spectra: ndarray, labels: List[str] | None = None, title: str = 'Spectral Data', xlabel: str = 'Wavenumber (cm$^{-1}$)', ylabel: str = 'Absorbance', alpha: float = 0.7, figsize: Tuple[int, int] = (12, 6), color_map: str = 'viridis', legend_loc: str = 'best', grid: bool = True, save_path: str | None = None)[source]#

Plot spectral data.

Parameters#

wavenumbersarray-like: The x-axis values (wavenumbers).
spectraarray-like: The spectra data of shape (n_samples, n_features).
labelslist of str, optional: Labels for each spectrum. If None, spectra are numbered.
titlestr, default=’Spectral Data’: Plot title.
xlabelstr, default=’Wavenumber (cm$^{-1}$)’: X-axis label.
ylabelstr, default=’Absorbance’: Y-axis label.
alphafloat, default=0.7: Transparency of the lines.
figsizetuple, default=(12, 6): Figure size.
color_mapstr, default=’viridis’: Colormap for the spectra.
legend_locstr, default=’best’: Location of the legend.
gridbool, default=True: Whether to show grid.
save_pathstr, optional: If provided, save the figure to this path.

Returns#

figmatplotlib.figure.Figure: The figure object.
axmatplotlib.axes.Axes: The axes object.

static set_style(style='whitegrid', context='paper', font_scale=1.2)[source]#

Set the visual style for the plots.

Parameters#

stylestr, default=’whitegrid’: The seaborn style.
contextstr, default=’paper’: The seaborn context.
font_scalefloat, default=1.2: The font scale.

API Reference#

Parameters#

Parameters#

Returns#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Returns#

Parameters#

Modules#

This Page