src.toolbox.steps.custom.find_profiles_beta

Attributes

DEFAULT_RESAMPLE_CADENCE
DEFAULT_GRADIENT_THRESHOLDS
DEFAULT_HORIZ_GRADIENT_THRESHOLD
DEFAULT_FILTER_WINDOW_SIZES
DEFAULT_EDGE_SQUEEZE
DEFAULT_DIVE_SCALE
DEFAULT_MAX_DEPTH_GAP
DEFAULT_MIN_HORIZONTAL_DURATION
DEFAULT_MIN_HORIZONTAL_DEPTH
FIXED_SAVGOL_WINDOW_VERT
FIXED_SAVGOL_WINDOW_HORIZ
FIXED_SAVGOL_POLY
FIXED_MIN_VALID_DEPTH
FIXED_MIN_POINTS_VERT
FIXED_MIN_POINTS_HORIZ
COLOUR_UP
COLOUR_DOWN
COLOUR_HORIZONTAL
COLOUR_TURNING
COLOUR_VELOCITY
COLOUR_RAW
COLOUR_SMOOTH
MARKER_SIZE
LINE_WIDTH

Classes

FindProfilesBetaStep

Functions

find_profiles_beta(df_sorted, cadence, ...)

Identifies and classifies vertical and horizontal profiles from depth-time data.

Module Contents

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_RESAMPLE_CADENCE = '30s'

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_GRADIENT_THRESHOLDS

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_HORIZ_GRADIENT_THRESHOLD = 0.01

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_FILTER_WINDOW_SIZES = [1, 2]

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_EDGE_SQUEEZE = 0

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_DIVE_SCALE = 15

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_MAX_DEPTH_GAP = 60.0

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_MIN_HORIZONTAL_DURATION = '20min'

src.toolbox.steps.custom.find_profiles_beta.DEFAULT_MIN_HORIZONTAL_DEPTH = 1.0

src.toolbox.steps.custom.find_profiles_beta.FIXED_SAVGOL_WINDOW_VERT = 5

src.toolbox.steps.custom.find_profiles_beta.FIXED_SAVGOL_WINDOW_HORIZ = 3

src.toolbox.steps.custom.find_profiles_beta.FIXED_SAVGOL_POLY = 2

src.toolbox.steps.custom.find_profiles_beta.FIXED_MIN_VALID_DEPTH = -0.5

src.toolbox.steps.custom.find_profiles_beta.FIXED_MIN_POINTS_VERT = 5

src.toolbox.steps.custom.find_profiles_beta.FIXED_MIN_POINTS_HORIZ = 20

src.toolbox.steps.custom.find_profiles_beta.COLOUR_UP = 'tab:blue'

src.toolbox.steps.custom.find_profiles_beta.COLOUR_DOWN = 'tab:green'

src.toolbox.steps.custom.find_profiles_beta.COLOUR_HORIZONTAL = 'tab:purple'

src.toolbox.steps.custom.find_profiles_beta.COLOUR_TURNING = 'tab:orange'

src.toolbox.steps.custom.find_profiles_beta.COLOUR_VELOCITY = 'tab:red'

src.toolbox.steps.custom.find_profiles_beta.COLOUR_RAW = 'tab:gray'

src.toolbox.steps.custom.find_profiles_beta.COLOUR_SMOOTH = 'black'

src.toolbox.steps.custom.find_profiles_beta.MARKER_SIZE = 2

src.toolbox.steps.custom.find_profiles_beta.LINE_WIDTH = 1.5

src.toolbox.steps.custom.find_profiles_beta.find_profiles_beta(df_sorted, cadence, filter_win_sizes, gradient_thresholds, horiz_grad_thresh, edge_squeeze, dive_scale, max_depth_gap, min_horizontal_duration, min_horizontal_depth, depth_col)

Identifies and classifies vertical and horizontal profiles from depth-time data.

This function applies compound smoothing and velocity thresholds to classify vehicle movements into descending (+1), ascending (-1), horizontal (0), or turning (NaN) segments. Vertical profiles are identified first based on steep vertical velocities and depth spans. Unassigned data is then evaluated for horizontal profiles based on flatter velocities, minimum depth, and duration. Final profile IDs are assigned chronologically.

Parameters:

• df_sorted (pandas.DataFrame) – Time-indexed dataframe containing the depth measurements.

• cadence (str) – Resampling frequency string (e.g., ’30s’) used to regularise the time series.

• filter_win_sizes (list) – Two-element list defining the rolling median and mean window sizes.

• gradient_thresholds (list) – [positive_threshold, negative_threshold] for vertical velocity bounds (m/s).

• horiz_grad_thresh (float) – Maximum absolute velocity (m/s) to be considered a horizontal profile.

• edge_squeeze (int) – Number of points to iteratively erode from the edges of turning regions.

• dive_scale (float) – Minimum total vertical distance (m) required for a valid vertical profile.

• max_depth_gap (float) – Maximum allowable depth gap (m) between consecutive points in a profile.

• min_horizontal_duration (str) – Minimum time duration string (e.g., ‘20min’) for a horizontal profile.

• min_horizontal_depth (float) – Minimum depth (m) required to evaluate a horizontal profile.

• depth_col (str) – Name of the column containing depth or pressure data.

Returns:

• df_out (pandas.DataFrame) – Dataframe aligned to original indices with added ‘PROFILE_ID’, ‘DIRECTION’, and ‘GRADIENT’.

• df (pandas.DataFrame) – The resampled and smoothed diagnostic dataframe.

class src.toolbox.steps.custom.find_profiles_beta.FindProfilesBetaStep

Bases: toolbox.steps.base_step.BaseStep, toolbox.utils.qc_handling.QCHandlingMixin

step_name = 'Find Profiles Beta'

required_variables = ['TIME']

provided_variables = ['PROFILE_NUMBER']

run()

generate_diagnostics()