scikit-diveMove is a Python interface to R package diveMove for scientific data analysis, with a focus on diving behaviour analysis. It has utilities to represent, visualize, filter, analyse, and summarize time-depth recorder (TDR) data. Miscellaneous functions for handling position and 3D kinematics data are also provided. scikit-diveMove communicates with a single R instance for access to low-level tools of package diveMove.

The table below shows which features of diveMove are accessible from scikit-diveMove:

diveMove		scikit-diveMove	Notes
Functionality	Functions/Methods
Movement	austFilter rmsDistFilter grpSpeedFilter distSpeed readLocs		Under consideration.
Bout analysis	boutfreqs boutinit bouts2.nlsFUN bouts2.nls bouts3.nlsFUN bouts3.nls bouts2.mleFUN bouts2.ll bouts2.LL bouts.mle labelBouts plotBouts plotBouts2.cdf bec2 bec3	BoutsNLS BoutsMLE	Fully implemented in Python.
Dive analysis	readTDR createTDR	TDR.__init__ TDRSource.__init__	Fully implemented. Single TDR class for data with or without speed measurements.
	calibrateDepth	TDR.calibrate TDR.zoc TDR.detect_wet TDR.detect_dives TDR.detect_dive_phases	Fully implemented
	calibrateSpeed rqPlot	TDR.calibrate_speed	New implementation of the algorithm entirely in Python. The procedure generates the plot concurrently.
	diveStats stampDive timeBudget	TDR.dive_stats TDR.time_budget TDR.stamp_dives	Fully implemented
	plotTDR plotDiveModel plotZOC	TDR.plot TDR.plot_zoc_filters TDR.plot_phases TDR.plot_dive_model	Fully implemented. Interactivity is the default, as standard matplotlib.
	getTDR getDepth getSpeed getTime getCCData getDtime getFileName	TDR.tdr TDR.get_depth TDR.get_speed TDR.tdr.index TDR.src_file TDR.dtime	Fully implemented. getCCData deemed redundant, as the columns can be accessed directly from the TDR.tdr attribute.
	getDAct getDPhaseLab getDiveDeriv getDiveModel getGAct	TDR.get_wet_activity TDR.get_dives_details TDR.get_dive_deriv	Fully implemented
	extractDive		Fully implemented

scikit-diveMove also provides useful tools for processing signals from tri-axial Inertial Measurement Units (IMU), such as thermal calibration, corrections for shifts in coordinate frames, as well as computation of orientation using a variety of current methods. Analyses are fully tractable by encouraging the use of xarray data structures that can be read from and written to NetCDF file format. Using these data structures, meta-data attributes can be easily appended at all layers as analyses progress.

Installation

Type the following at a terminal command line:

pip install scikit-diveMove

Or install from source tree by typing the following at the command line:

python setup.py install

The documentation can also be installed as described in Documentation.

Once installed, skdiveMove can be easily imported as:

import skdiveMove as skdive

Dependencies

skdiveMove depends primarily on R package diveMove, which must be installed and available to the user running Python. If needed, install diveMove at the R prompt:

install.packages("diveMove")

Documentation

Available at: https://spluque.github.io/scikit-diveMove

Alternatively, installing the package as follows:

pip install -e .["dev"]

allows the documentation to be built locally (choosing the desired target {"html", "pdf", etc.}):

make -C docs/ html