Parameters¶

This page provides a comprehensive overview of all configuration parameters available in zea. These parameters are used in the YAML config files to control data loading, preprocessing, model settings, scan parameters, and more.

Note

You can use these configs to, for instance, initialize zea.Models or zea.Pipeline.

Configs are written in YAML format and can be loaded, edited, and saved using the zea API.

How to Load and Save a Config¶

Here is a minimal example of how to load and save a config file using zea:

from zea import Config
from zea.config import check_config

# Load a config from file
config = Config.from_yaml("configs/config_picmus_rf.yaml")
# or some predefined from Hugging Face Hub
config = Config.from_hf("zeahub/configs", "config_picmus_rf.yaml", repo_type="dataset")

# We can check if the config has valid parameters (zea compliance)
config = check_config(config)

# Access or change parameters
print(config.scan.sampling_frequency)
config.scan.sampling_frequency = 8e6

# Save the config back to file
config.save_to_yaml("configs/config_picmus_rf_modified.yaml")

Parameter List¶

Below is a hierarchical list of all configuration parameters, grouped by section. Descriptions are shown for each parameter.

Parameters Reference¶

Parameter	Description
`data`	The data section contains the parameters for the data.
`data.apodization`	The receive apodization to use.
`data.dataset_folder`	The path of the folder to load data files from (relative to the user data root as set in users.yaml)
`data.dtype`	The form of data to load (raw_data, rf_data, iq_data, beamformed_data, envelope_data, image, image_sc)
`data.dynamic_range`	The dynamic range for showing data in db [min, max]
`data.file_path`	The path of the file to load when running the UI (either an absolute path or one relative to the dataset folder)
`data.frame_no`	The frame number to load when running the UI (null, int, ‘all’)
`data.input_range`	The range of the input data in db (null, [min, max])
`data.local`	true: use local data on this device, false: use data from NAS
`data.output_range`	The output range to which the data should be mapped (e.g. [0, 1]).
`data.resolution`	The spatial resolution of the data in meters per pixel (float, optional).
`data.to_dtype`	The type of data to convert to (raw_data, aligned_data, beamformed_data, envelope_data, image, image_sc)
`data.user`	The user to use when loading data (null, dict)
`device`	The device to run on (‘cpu’, ‘gpu:0’, ‘gpu:1’, …)
`git`	The git commit hash or branch for reproducibility (string, optional).
`hide_devices`	List of device indices to hide from selection (list of int, optional).
`pipeline`	This section contains the necessary parameters for building the pipeline.
`pipeline.jit_kwargs`	Additional keyword arguments for the JIT compiler. Defaults to None.
`pipeline.jit_options`	The JIT options to use. Must be ‘pipeline’, ‘ops’, or None. ‘pipeline’ compiles the entire pipeline as a single function. ‘ops’ compiles each operation separately. None disables JIT compilation. Defaults to ‘ops’.
`pipeline.name`	The name of the pipeline. Defaults to ‘pipeline’.
`pipeline.operations`	The operations to perform on the data. This is a list of dictionaries, where each dictionary contains the parameters for a single operation.
`pipeline.validate`	Whether to validate the pipeline. Defaults to True.
`pipeline.with_batch_dim`	Whether operations should expect a batch dimension in the input. Defaults to True.
`plot`	Settings pertaining to plotting when running the UI (zea –config <path-to-config.yaml>)
`plot.fliplr`	Set to true to flip the image left to right
`plot.fps`	Frames per second for video output.
`plot.headless`	Set to true to run the UI in headless mode
`plot.image_extension`	The file extension to use when saving the image (png, jpg)
`plot.plot_lib`	The plotting library to use (opencv, matplotlib)
`plot.save`	Set to true to save the plots to disk, false to only display them in the UI
`plot.selector`	Type of selector to use for ROI selection in the UI (‘rectangle’, ‘lasso’, or None).
`plot.selector_metric`	Metric to use for evaluating selected regions (e.g., ‘gcnr’).
`plot.tag`	The name for the plot
`plot.video_extension`	The file extension to use when saving the video (mp4, gif)
`scan`	The scan section contains the parameters pertaining to the reconstruction.
`scan.grid_size_x`	The number of pixels in the beamforming grid in the x-direction
`scan.grid_size_z`	The number of pixels in the beamforming grid in the z-direction
`scan.apply_lens_correction`	Set to true to apply lens correction in the time-of-flight calculation
`scan.center_frequency`	The center frequency of the transducer in Hz
`scan.demodulation_frequency`	The demodulation frequency of the data in Hz. This is the assumed center frequency of the transmit waveform used to demodulate the rf data to iq data.
`scan.f_number`	The receive f-number for apodization. Set to zero to disable masking. The f-number is the ratio between the distance from the transducer and the size of the aperture.
`scan.fill_value`	Value to fill the image with outside the defined region (float, default 0.0).
`scan.lens_sound_speed`	The speed of sound in the lens in m/s. Usually around 1000 m/s
`scan.lens_thickness`	The thickness of the lens in meters
`scan.n_ax`	The number of samples in a receive recording per channel.
`scan.n_ch`	The number of channels in the raw data (1 for rf data, 2 for iq data)
`scan.phi_range`	The range of phi values in radians for 3D scan conversion (null, [min, max]).
`scan.resolution`	The resolution for scan conversion in meters per pixel (float, optional).
`scan.rho_range`	The range of rho values in meters for scan conversion (null, [min, max]).
`scan.sampling_frequency`	The sampling frequency of the data in Hz
`scan.selected_transmits`	The number of transmits in a frame. Can be ‘all’ for all transmits, an integer for a specific number of transmits selected evenly from the transmits in the frame, or a list of integers for specific transmits to select from the frame.
`scan.theta_range`	The range of theta values in radians for scan conversion (null, [min, max]).
`scan.xlims`	The limits of the x-axis in the scan in meters (null, [min, max])
`scan.ylims`	The limits of the y-axis in the scan in meters (null, [min, max])
`scan.zlims`	The limits of the z-axis in the scan in meters (null, [min, max])