Inference API

Single Inferencer

class cryoPARES.inference.inferencer.SingleInferencer(particles_star_fname, checkpoint_dir, results_dir, data_halfset='allParticles', model_halfset='matchingHalf', particles_dir=None, batch_size=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, num_dataworkers=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, use_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_cpus_if_no_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, compile_model=False, top_k_poses_nnet=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, top_k_poses_localref=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, grid_distance_degs=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, reference_map=None, reference_mask=None, directional_zscore_thr=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_localrefinement=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_reconstruction=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, subset_idxs=None, n_first_particles=None, show_debug_stats=False)

Bases: object

Parameters:

• particles_star_fname (str)

• checkpoint_dir (str)

• results_dir (str)

• data_halfset (Literal['half1', 'half2', 'allParticles'])

• model_halfset (Literal['half1', 'half2', 'allCombinations', 'matchingHalf'])

• particles_dir (str | None)

• batch_size (int)

• num_dataworkers (int)

• use_cuda (bool)

• n_cpus_if_no_cuda (int)

• compile_model (bool)

• top_k_poses_nnet (int)

• top_k_poses_localref (int)

• grid_distance_degs (float)

• reference_map (str | None)

• reference_mask (str | None)

• directional_zscore_thr (float | None)

• skip_localrefinement (bool)

• skip_reconstruction (bool)

• subset_idxs (List[int] | None)

• n_first_particles (int | None)

• show_debug_stats (bool)

__init__(particles_star_fname, checkpoint_dir, results_dir, data_halfset='allParticles', model_halfset='matchingHalf', particles_dir=None, batch_size=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, num_dataworkers=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, use_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_cpus_if_no_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, compile_model=False, top_k_poses_nnet=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, top_k_poses_localref=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, grid_distance_degs=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, reference_map=None, reference_mask=None, directional_zscore_thr=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_localrefinement=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_reconstruction=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, subset_idxs=None, n_first_particles=None, show_debug_stats=False)

Initializes the SingleInferencer for running inference on a set of particles.

Parameters:

• particles_star_fname (str) – Path to the STAR file containing particle information.

• checkpoint_dir (str) – Directory where the trained model checkpoints are stored.

• results_dir (str) – Directory where the inference results will be saved.

• data_halfset (Literal['half1', 'half2', 'allParticles']) – Specifies which half-set of the data to use (“half1”, “half2”, or “allParticles”).

• model_halfset (Literal['half1', 'half2', 'allCombinations', 'matchingHalf']) – Specifies which half-set of the model to use (“half1”, “half2”, “allCombinations”, or “matchingHalf”).

• particles_dir (Optional[str]) – Directory where the particle images are located. If None, paths in the STAR file are assumed to be absolute.

• batch_size (int) – The number of particles to process in each batch.

• num_dataworkers (int) – The number of worker processes to use for data loading.

• use_cuda (bool) – Whether to use a CUDA-enabled GPU for inference.

• n_cpus_if_no_cuda (int) – The number of CPU cores to use if CUDA is not available.

• compile_model (bool) – Whether to compile the model using torch.compile for potential speed-up.

• top_k_poses_nnet (int) – The number of top predictions to predict with the nn for each particle.

• top_k_poses_localref (int) – The number of top predictions to return after local refinement.

• grid_distance_degs (float) – The size of the local refinement grid in degrees. Grid will go from -grid_distance_degs to +grid_distance_degs

• reference_map (Optional[str]) – Path to the reference map for local refinement. If not provided, it will be loaded from the checkpoint.

• reference_mask (Optional[str]) – Path to the mask of the reference map. Used only for FSC calculation.

• directional_zscore_thr (Optional[float]) – The threshold for the directional Z-score to filter particles.

• skip_localrefinement (bool) – Whether to skip local refinement of the particle poses.

• skip_reconstruction (bool) – Whether to skip 3D reconstruction from the inferred poses.

• subset_idxs (Optional[List[int]]) – A list of indices to process a subset of particles.

• n_first_particles (Optional[int]) – The number of first particles to process. Cannot be used with subset_idxs.

• show_debug_stats (bool) – Whether to print debug statistics, such as rotation errors if ground truth in the starfile.

property symmetry

The symmetry of the model, loaded from the hyperparameters file.

run()

Runs the inference process. It iterates through the specified data and model half-sets, performing inference for each combination.

Returns:

A list of tuples, where each tuple contains the particle metadata and the reconstructed volume for each inference run.

clean_reconstructer_buffers()

static resolve_halfset_lists(data_halfset, model_halfset)

Return normalized lists of data and model-half policies for iteration. - data list is concrete halves: [“half1”], [“half2”], or [“half1”,”half2”]. - model list is [“half1”], [“half2”], [None] (matchingHalf), or [“half1”,”half2”] (allCombinations).

Return type:

Tuple[List[Literal['half1', 'half2']], List[Optional[Literal['half1', 'half2']]]]

Parameters:

• data_halfset (Literal['half1', 'half2', 'allParticles'])

• model_halfset (Literal['half1', 'half2', 'allCombinations', 'matchingHalf'])

SingleInferencer Class

class cryoPARES.inference.inferencer.SingleInferencer(particles_star_fname, checkpoint_dir, results_dir, data_halfset='allParticles', model_halfset='matchingHalf', particles_dir=None, batch_size=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, num_dataworkers=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, use_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_cpus_if_no_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, compile_model=False, top_k_poses_nnet=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, top_k_poses_localref=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, grid_distance_degs=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, reference_map=None, reference_mask=None, directional_zscore_thr=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_localrefinement=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_reconstruction=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, subset_idxs=None, n_first_particles=None, show_debug_stats=False)

Bases: object

Parameters:

• particles_star_fname (str)

• checkpoint_dir (str)

• results_dir (str)

• data_halfset (Literal['half1', 'half2', 'allParticles'])

• model_halfset (Literal['half1', 'half2', 'allCombinations', 'matchingHalf'])

• particles_dir (str | None)

• batch_size (int)

• num_dataworkers (int)

• use_cuda (bool)

• n_cpus_if_no_cuda (int)

• compile_model (bool)

• top_k_poses_nnet (int)

• top_k_poses_localref (int)

• grid_distance_degs (float)

• reference_map (str | None)

• reference_mask (str | None)

• directional_zscore_thr (float | None)

• skip_localrefinement (bool)

• skip_reconstruction (bool)

• subset_idxs (List[int] | None)

• n_first_particles (int | None)

• show_debug_stats (bool)

__init__(particles_star_fname, checkpoint_dir, results_dir, data_halfset='allParticles', model_halfset='matchingHalf', particles_dir=None, batch_size=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, num_dataworkers=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, use_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_cpus_if_no_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, compile_model=False, top_k_poses_nnet=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, top_k_poses_localref=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, grid_distance_degs=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, reference_map=None, reference_mask=None, directional_zscore_thr=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_localrefinement=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_reconstruction=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, subset_idxs=None, n_first_particles=None, show_debug_stats=False)

Initializes the SingleInferencer for running inference on a set of particles.

Parameters:

• particles_star_fname (str) – Path to the STAR file containing particle information.

• checkpoint_dir (str) – Directory where the trained model checkpoints are stored.

• results_dir (str) – Directory where the inference results will be saved.

• data_halfset (Literal['half1', 'half2', 'allParticles']) – Specifies which half-set of the data to use (“half1”, “half2”, or “allParticles”).

• model_halfset (Literal['half1', 'half2', 'allCombinations', 'matchingHalf']) – Specifies which half-set of the model to use (“half1”, “half2”, “allCombinations”, or “matchingHalf”).

• particles_dir (Optional[str]) – Directory where the particle images are located. If None, paths in the STAR file are assumed to be absolute.

• batch_size (int) – The number of particles to process in each batch.

• num_dataworkers (int) – The number of worker processes to use for data loading.

• use_cuda (bool) – Whether to use a CUDA-enabled GPU for inference.

• n_cpus_if_no_cuda (int) – The number of CPU cores to use if CUDA is not available.

• compile_model (bool) – Whether to compile the model using torch.compile for potential speed-up.

• top_k_poses_nnet (int) – The number of top predictions to predict with the nn for each particle.

• top_k_poses_localref (int) – The number of top predictions to return after local refinement.

• grid_distance_degs (float) – The size of the local refinement grid in degrees. Grid will go from -grid_distance_degs to +grid_distance_degs

• reference_map (Optional[str]) – Path to the reference map for local refinement. If not provided, it will be loaded from the checkpoint.

• reference_mask (Optional[str]) – Path to the mask of the reference map. Used only for FSC calculation.

• directional_zscore_thr (Optional[float]) – The threshold for the directional Z-score to filter particles.

• skip_localrefinement (bool) – Whether to skip local refinement of the particle poses.

• skip_reconstruction (bool) – Whether to skip 3D reconstruction from the inferred poses.

• subset_idxs (Optional[List[int]]) – A list of indices to process a subset of particles.

• n_first_particles (Optional[int]) – The number of first particles to process. Cannot be used with subset_idxs.

• show_debug_stats (bool) – Whether to print debug statistics, such as rotation errors if ground truth in the starfile.

property symmetry

The symmetry of the model, loaded from the hyperparameters file.

run()

Runs the inference process. It iterates through the specified data and model half-sets, performing inference for each combination.

Returns:

A list of tuples, where each tuple contains the particle metadata and the reconstructed volume for each inference run.

clean_reconstructer_buffers()

static resolve_halfset_lists(data_halfset, model_halfset)

Return normalized lists of data and model-half policies for iteration. - data list is concrete halves: [“half1”], [“half2”], or [“half1”,”half2”]. - model list is [“half1”], [“half2”], [None] (matchingHalf), or [“half1”,”half2”] (allCombinations).

Return type:

Tuple[List[Literal['half1', 'half2']], List[Optional[Literal['half1', 'half2']]]]

Parameters:

• data_halfset (Literal['half1', 'half2', 'allParticles'])

• model_halfset (Literal['half1', 'half2', 'allCombinations', 'matchingHalf'])

Distributed Inference

cryoPARES.inference.infer.distributed_inference(particles_star_fname, checkpoint_dir, results_dir, data_halfset='allParticles', model_halfset='matchingHalf', particles_dir=None, batch_size=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_jobs=None, num_dataworkers=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, use_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_cpus_if_no_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, compile_model=False, top_k_poses_nnet=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, top_k_poses_localref=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, grid_distance_degs=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, reference_map=None, reference_mask=None, directional_zscore_thr=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_localrefinement=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_reconstruction=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, subset_idxs=None, n_first_particles=None, check_interval_secs=2.0)

Distributed inference across particles and devices, mirroring the halfset selection logic used by cryoPARES.inference.inference.SingleInferencer.

Parameters:

• particles_star_fname (str) – Path to input RELION particles .star file

• checkpoint_dir (str) – Path to training directory (or .zip file) containing half-set models with checkpoints and hyperparameters. By default they are called version_0, version_1, etc.

• results_dir (str) – Output directory for inference results including predicted poses and optional reconstructions

• data_halfset ({"half1", "half2", "allParticles"}, default "allParticles") – Which particle half-set(s) to process: “half1”, “half2”, or “allParticles”

• model_halfset ({"half1", "half2", "allCombinations", "matchingHalf"}, default "matchingHalf") – Model half-set selection policy: “half1”, “half2”, “allCombinations”, or “matchingHalf” (uses matching data/model pairs)

• particles_dir (str, optional) – Root directory for particle image paths. If provided, overrides paths in the .star file

• batch_size (int) – Number of particles per batch for inference

• n_jobs (int, optional) – Number of worker processes. Defaults to number of GPUs if CUDA enabled, otherwise 1

• num_dataworkers (int) – Number of parallel data loading workers per GPU. Each worker is a separate CPU process. Set to 0 to load data in the main thread (useful only for debugging). Try not to oversubscribe by asking more workers than CPUs

• use_cuda (bool) – Enable GPU acceleration for inference. If False, runs on CPU only

• n_cpus_if_no_cuda (int) – Maximum CPU threads per worker when CUDA is disabled

• compile_model (bool) – Compile model with torch.compile for faster inference (experimental, requires PyTorch 2.0+)

• top_k_poses_nnet (int) – Number of top pose predictions to retrieve from neural network before local refinement

• top_k_poses_localref (int) – Number of best matching poses to keep after local refinement

• grid_distance_degs (float) – Maximum angular distance in degrees for local refinement search. Grid ranges from -grid_distance_degs to +grid_distance_degs around predicted pose

• reference_map (str, optional) – Path to reference map (.mrc) for FSC computation during validation

• reference_mask (str, optional) – Path to reference mask (.mrc) for masked FSC calculation

• directional_zscore_thr (float, optional) – Confidence z-score threshold for filtering particles. Particles with scores below this are discarded as low-confidence

• skip_localrefinement (bool) – Skip local pose refinement step and use only neural network predictions

• skip_reconstruction (bool) – Skip 3D reconstruction step and output only predicted poses

• subset_idxs (list[int], optional) – List of particle indices to process (for debugging or partial processing)

• n_first_particles (int, optional) – Process only the first N particles from dataset (debug feature)

• check_interval_secs (float) – Polling interval in seconds for parent loop in distributed processing

Notes

• For n_jobs == 1, this function delegates completely to SingleInferencer, which internally handles data_halfset and model_halfset (including “matchingHalf” and “allCombinations”).

• For n_jobs > 1, particles are split by half-set and distributed across workers; model half selection follows the resolved policy per half to avoid cross-half coupling.

cryoPARES.inference.infer.worker(worker_id, output_q, *args, **kwargs)

cryoPARES.inference.infer.main()

Daemon Mode

Queue Manager

manager_server.py - Creates and manages the shared queue server. This script must be run first to establish the Manager server.

class cryoPARES.inference.daemon.queueManager.QueueManager(address=None, authkey=None, serializer='pickle', ctx=None, *, shutdown_timeout=1.0)

Bases: BaseManager

Custom Manager class to handle our shared queue.

cryoPARES.inference.daemon.queueManager.create_shared_queue(maxsize)

Factory function to create the shared queue.

cryoPARES.inference.daemon.queueManager.connect_to_queue(ip='localhost', port=50000, authkey='shared_queue_key')

Connect to the shared queue managed by the server.

cryoPARES.inference.daemon.queueManager.queue_manager_server(ip, port, authkey, queue_maxsize)

Context manager for the QueueManager server.

cryoPARES.inference.daemon.queueManager.signal_handler(signum, frame)

Handle Ctrl+C or SIGTERM gracefully.

cryoPARES.inference.daemon.queueManager.queue_connection(ip='localhost', port=50000, authkey='shared_queue_key')

Context manager for connecting to the shared queue.

cryoPARES.inference.daemon.queueManager.get_all_available_items(q)

Gets all items from a queue. If the queue is empty, it waits for the first item. Then, it retrieves all other items without blocking.

Parameters:

q (Queue)

cryoPARES.inference.daemon.queueManager.main(ip='localhost', port=50000, authkey='shared_queue_key', queue_maxsize=None)

Parameters:

• ip (str)

• port (int)

• authkey (str) – a password to use the queue

• queue_maxsize (Optional[int])

Returns:

Spooling Filler

spooler.py - Monitors a directory for new .star files and adds them to the shared queue for DaemonInferencer.

cryoPARES.inference.daemon.spoolingFiller.monitor_directory(directory, pattern='*.star', interval=10)

Monitor a directory for new files matching the pattern and yield their paths.

Parameters:

• directory (str) – Directory to monitor

• pattern (str) – File pattern to match (default: *.star)

• interval (int) – Time interval (seconds) between directory checks

cryoPARES.inference.daemon.spoolingFiller.main(directory, ip='localhost', port=50000, authkey='shared_queue_key', pattern='*.star', check_interval=10)

Main function to monitor a directory and feed new .star files to the queue.

Parameters:

• directory (str) – Directory to monitor for .star files

• ip (str) – Queue manager IP address

• port (int) – Queue manager port

• authkey (str) – Queue manager authentication key

• pattern (str) – File pattern to match

• check_interval (int) – Seconds between directory checks

Daemon Inference

class cryoPARES.inference.daemon.daemonInference.DaemonInferencer(checkpoint_dir, results_dir, net_address='localhost', net_port=50000, net_authkey='shared_queue_key', model_halfset='half1', particles_dir=None, batch_size=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, num_dataworkers=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, use_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_cpus_if_no_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, compile_model=False, top_k_poses_nnet=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, top_k_poses_localref=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, reference_map=None, reference_mask=None, directional_zscore_thr=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_localrefinement=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_reconstruction=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, show_debug_stats=False, secs_between_partial_results_written=5, resubmit_poison_pill=True)

Bases: SingleInferencer

Parameters:

• checkpoint_dir (str)

• results_dir (str)

• net_address (str)

• net_port (int)

• net_authkey (str | None)

• model_halfset (Literal['half1', 'half2'])

• particles_dir (str | None)

• batch_size (int)

• num_dataworkers (int)

• use_cuda (bool)

• n_cpus_if_no_cuda (int)

• compile_model (bool)

• top_k_poses_nnet (int)

• top_k_poses_localref (int)

• reference_map (str | None)

• reference_mask (str | None)

• directional_zscore_thr (float | None)

• skip_localrefinement (bool)

• skip_reconstruction (bool)

• show_debug_stats (bool)

• secs_between_partial_results_written (int)

• resubmit_poison_pill (bool)

__init__(checkpoint_dir, results_dir, net_address='localhost', net_port=50000, net_authkey='shared_queue_key', model_halfset='half1', particles_dir=None, batch_size=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, num_dataworkers=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, use_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, n_cpus_if_no_cuda=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, compile_model=False, top_k_poses_nnet=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, top_k_poses_localref=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, reference_map=None, reference_mask=None, directional_zscore_thr=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_localrefinement=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, skip_reconstruction=<cryoPARES.configManager.inject_defaults.CONFIG_PARAM object>, show_debug_stats=False, secs_between_partial_results_written=5, resubmit_poison_pill=True)

Initializes the DaemonInferencer for running inference on a set of particles from a queue.

Parameters:

• checkpoint_dir (str) – Directory where the trained model checkpoints are stored.

• results_dir (str) – Directory where the inference results will be saved.

• net_address (str) – Network address of the queue manager.

• net_port (int) – Network port of the queue manager.

• net_authkey (Optional[str]) – Network authentication key for the queue manager.

• model_halfset (Literal['half1', 'half2']) – Specifies which half-set of the model to use (“half1”, “half2”).

• particles_dir (Optional[str]) – Directory where the particle images are located. If None, paths in the STAR file are assumed to be absolute.

• batch_size (int) – The number of particles to process in each batch.

• num_dataworkers (int) – The number of worker processes to use for data loading.

• use_cuda (bool) – Whether to use a CUDA-enabled GPU for inference.

• n_cpus_if_no_cuda (int) – The number of CPU cores to use if CUDA is not available.

• compile_model (bool) – Whether to compile the model using torch.compile for potential speed-up.

• top_k_poses_nnet (int) – The number of top predictions to predict with the nn for each particle.

• top_k_poses_localref (int) – The number of top predictions to return after local refinement.

• reference_map (Optional[str]) – Path to the reference map for local refinement. If not provided, it will be loaded from the checkpoint.

• reference_mask (Optional[str]) – Path to the mask of the reference map. Used only for FSC calculation.

• directional_zscore_thr (Optional[float]) – The threshold for the directional Z-score to filter particles.

• skip_localrefinement (bool) – Whether to skip local refinement of the particle poses.

• skip_reconstruction (bool) – Whether to skip 3D reconstruction from the inferred poses.

• show_debug_stats (bool) – Whether to print debug statistics, such as rotation errors if ground truth in the starfile.

• secs_between_partial_results_written (int) – Partial results are saved from RAM to disk every few seconds

• resubmit_poison_pill (bool) – If True, posion pills are re-submitted, to ensure that all workers will die

run()

Runs the inference process. It iterates through the specified data and model half-sets, performing inference for each combination.

Returns:

A list of tuples, where each tuple contains the particle metadata and the reconstructed volume for each inference run.

resolve_data(starfname)

Parameters:

starfname (str | None)