API Reference

High-level Functions

to_healpix(ds)	Helper that loads a Dataset as a HEALPix grid (indexed by "cell_ids").
reshape_to_2d(ds)
load_tag(*, tag_root, tag_name[, ...])	Load a tag.
update_stations(*, tag, station_file_uri[, ...])	Add or replace the acoustic receiver data of a tag.
plot_tag(*, tag, tag_log[, save_html, ...])	Plot a tag.
load_model(*, uri, tag_log, time_slice, bbox)	Load and prepare a reference model.
compute_diff(*, reference_model, tag_log, ...)	Compute the difference between the reference model and the DST data of a tag.
open_diff_dataset(*, target_root, ...)	Open a diff dataset.
regrid_dataset(*, ds, refinement_level[, ...])	Regrid a dataset as a HEALPix grid, whose primary advantage is that all its cells/pixels cover the same surface area.
compute_emission_pdf(*, diff_ds, events_ds, ...)	Compute the temporal emission matrices given a dataset and tagging events.
compute_acoustic_pdf(*, emission_ds, tag, ...)	Compute a emission probability distribution from (acoustic) detection data.
combine_pdfs(*, emission_ds, acoustic_ds, chunks)	Combine and normalize 2 probability distributions (pdfs).
normalize_pdf(*, ds, chunks[, dims, plot])	Normalize a probability distributions (pdf).
optimize_pdf(*, ds, earth_radius, ...[, ...])	Optimize a temporal probability distribution.
predict_positions(*[, ds, target_root, ...])	High-level helper function for predicting fish's positions and generating the consequent trajectories.
plot_trajectories(*, target_root, ...[, ...])	Read decoded trajectories and plots an interactive visualization.
open_distributions(*, target_root, ...[, ...])	Load and merge the emission and states probability distributions into a single dataset.
plot_distributions(*, data[, bbox])	Plot an interactive visualization of dataset resulting from the merging of emission and the states distributions.
render_frames(*, ds[, time_slice])	Helper function for rendering images.
render_distributions(*, data[, time_step, ...])	Render a video of a dataset resulting from the merging of emission and the states distributions.

I/O

open_tag(root, name[, storage_options])	Open a tag
open_copernicus_catalog(cat[, chunks])	Assemble the given intake catalog into a dataset
open_copernicus_zarr([model, format, freq, ...])	Retrieve Copernicus Marine data in zarr format.
prepare_dataset(dataset[, chunks, bbox, names])	Prepares a dataset of a reference model.
save_trajectories(traj, root[, ...])
read_trajectories(names, root[, ...])	Read trajectories from disk
save_html_hvplot(plot, filepath[, ...])	Save a Holoviews plot to an HTML file either locally or on an S3 bucket.
tz_convert(df, timezones)	Convert the timezone of columns in a dataframe

Grid Manipulation

grid.center_longitude(ds, center)
healpy.geographic_to_astronomic(lat, lon, rot)	transform geographic coordinates to astronomic coordinates
healpy.astronomic_to_cartesian(theta, phi[, dim])	transform astronomic coordinates to cartesian coordinates
healpy.astronomic_to_cell_ids(nside, phi, theta)	Compute cell ids from astronomic coordinates
healpy.buffer_points(cell_ids, positions, *, ...)	select the cells within a circular buffer around the given positions

Emission Computations

acoustic.emission_probability(tag, grid, ...)	Construct emission probability maps from acoustic detections
pdf.normal(samples, mean, std, *, dims)	Compute the combined pdf of independent layers
pdf.combine_emission_pdf(raw[, exclude])
diff.diff_z(model, tag[, depth_threshold])

Tag/time Operations

tags.to_time_slice(times)
tags.adapt_model_time(slice_)
tags.reshape_by_bins(ds, *, dim, bins[, ...])
cf.bounds_to_bins(ds[, bounds_dim])
dataset_utils.broadcast_variables(ds, ...)	Broadcast variables against each other

Visualization

create_single_frame(ds, figure, **kwargs)	Default function for plotting a snapshot (i.e, timeless data) of the emission and states distributions.
render_frame(ds, *args[, figsize, frames_dir])
plot_map(arr, bbox[, x, y, rasterize, geo, ...])	Wrapper around DataArray.hvplot.quadmesh, with different defaults
plot_trajectories(trajectories, *[, subplots])
filter_by_states(ds)

Estimators

EagerEstimator(predictor_factory[, sigma])	Estimator to train and predict gaussian random walk hidden markov models
CachedEstimator(predictor_factory[, sigma, ...])	Estimator to train and predict gaussian random walk hidden markov models

Predictors

Gaussian2DCartesian(sigma, truncate, ...)
Gaussian1DHealpix(cell_ids, grid_info, ...)

Searches

GridSearch(estimator, search_grid)	Optimize estimator parameters using a search grid
EagerBoundsSearch(estimator, param_bounds, *)	Optimize estimator parameters within an interval
TargetBoundsSearch(estimator, x0, ...[, ...])	Optimize estimator parameters within an interval

Low-level Functions

Distributions

planar2d.create_covariances(cov, coord_names)
planar2d.normal_at(grid, *, pos, cov[, ...])	Multivariate normal distribution
healpix.normal_at(grid, pos, sigma)

Trajectory Generation

mean_track(X[, coords])
modal_track(X[, coords])
viterbi(emission, sigma)

HMM Filtering

score(emission, predictor, initial_probability)	Score of a single pass (forwards) of the spatial HMM filter
forward(emission, predictor, initial_probability)	Single pass (forwards) of the spatial HMM filter
backward(states, predictions, predictor[, mask])
forward_backward(emission, predictor, ...[, ...])	Double pass (forwards and backwards) of the spatial HMM filter