tomodrgn.analysis

Functions for analysis of particle metadata: index, pose, ctf, latent embedding, label, tomogram spatial context, etc.

Functions

cluster_gmm	Cluster latent embeddings using a K-component full covariance Gaussian mixture model.
cluster_kmeans	Cluster latent embeddings using k-means clustering.
combine_ind	Combine multiple indices selections by either intersection or union.
convert_angstroms_to_voxels	Rescale dataframe coordinates from angstroms to unitless voxels corresponding to reconstructed tomograms.
convert_original_indices	Convert selected indices relative to a filtered particle stack into indices relative to the unfiltered particle stack.
get_colors_chimerax	Sample num_colors from the ChimeraX color scheme as RGBA tuples normalized [0,1].
get_colors_matplotlib	Sample num_colors colors from the specified color map as RGBA tuples
get_ind_for_cluster	Get the indices of particles belonging to the selected clusters.
get_nearest_point	Find the closest point in data to query.
get_pc_traj	Sample latent embeddings along specified principal component dim at coordininates in PC-space specified by sampling_points.
ipy_plot_interactive	Create and display an interactive plotly scatter plot and associated ipywidgets custom widgets, allowing exploration of numeric columns of a pandas dataframe.
ipy_tomo_ptcl_viewer	An interactive tomogram and particle viewer using plotly and ipywidgets.
load_dataframe	Merge known types of numpy arrays into a single pandas dataframe for downstream analysis.
parse_all_losses	Parse MSE, KLD, and total loss at each epoch from run.log output.
parse_loss	Parse total loss at each epoch from run.log output.
plot_by_cluster	Plot all points x,y with colors per class labels, with optional annotations for each cluster center and corresponding label.
plot_by_cluster_subplot	Plot all points x,y with colors per class labels on individual subplots for each of labels_sel.
plot_euler	Plot the distribution of Euler angles as a hexbin of theta and phi, and a histogram of psi.
plot_label_count_distribution	Plot the distribution of class labels per tomogram or micrograph as a heatmap.
plot_losses	Plot the total loss, reconstruction loss, and KLD divergence per epoch.
plot_projections	Plot a stack of grayscale images.
plot_three_column_correlation	Plot two reference vectors (e.g. l-UMAP1 and l-UMAP2) for potential correlation with a third query vector (e.g. CoordinateX, DefocusU, etc.).
plot_translations	Plot the distribution of shifts in x-axis vs shifts in y-axis (units: px)
recursive_load_dataframe	Create merged dataframe containing:
run_pca	Run principal component analysis on the latent embeddings.
run_tsne	Run t-SNE dimensionality reduction on latent embeddings.
run_umap	Run UMAP dimensionality reduction on latent embeddings.
scatter_annotate	Create a scatter plot with optional annotations for each cluster center and corresponding label.
scatter_annotate_hex	Create a hexbin plot with optional annotations for each cluster center and corresponding label.
scatter_color	Create a scatter plot colored by auto-mapped values of c according to specified cmap, and plot a corresponding colorbar.