Tools for Handling Extraction of Features from Time Series

[Up] [Top]

Documentation for package ‘theft’ version 0.4.1.1

Help Pages

calculate_features Compute features on an input time series dataset
check_vector_quality Check data quality of a vector
compute_top_features Return an object containing results from top-performing features on a classification task
demo_multi_outputs Computed values for multi-feature classification results for use in vignette
demo_outputs Computed values for top features results for use in vignette
feature_list All features available in theft in tidy format
fit_multi_feature_classifier Fit a classifier to feature matrix using all features or all features by set
fit_single_feature_classifier Fit a classifier to feature matrix to extract top performers
init_theft Communicate to R the correct Python version containing the relevant libraries for calculating features
minmax_scaler This function rescales a vector of numerical values into the unit interval [0,1]
normalise_feature_frame Scale each feature vector into a user-specified range for visualisation and modelling
normalise_feature_vector Scale each value into a user-specified range for visualisation and analysis
normalize_feature_frame Scale each feature vector into a user-specified range for visualisation and modelling
normalize_feature_vector Scale each value into a user-specified range for visualisation and analysis
plot_all_features Produce a heatmap matrix of the calculated feature value vectors and each unique time series with automatic hierarchical clustering.
plot_feature_correlations Produce a correlation matrix plot showing pairwise correlations of feature vectors by unique id with automatic hierarchical clustering.
plot_feature_matrix Produce a heatmap matrix of the calculated feature value vectors and each unique time series with automatic hierarchical clustering.
plot_low_dimension Produce a principal components analysis (PCA) on normalised feature values and render a bivariate plot to visualise it
plot_quality_matrix Produce a matrix visualisation of data types computed by feature calculation function.
plot_ts_correlations Produce a correlation matrix plot showing pairwise correlations of time series with automatic hierarchical clustering
process_hctsa_file Load in hctsa formatted MATLAB files of time series data into a tidy format ready for feature extraction
robustsigmoid_scaler This function rescales a vector of numerical values with an outlier-robust Sigmoidal transformation
sigmoid_scaler This function rescales a vector of numerical values with a Sigmoidal transformation
simData Sample of randomly-generated time series to produce function tests and vignettes
theft Tools for Handling Extraction of Features from Time-series
zscore_scaler This function rescales a vector of numerical values into z-scores