PreProcessor

class patato.PreProcessor(time_factor: int = 3, detector_factor: int = 2, irf: bool = True, hilbert: bool = True, lp_filter: float | None = None, hp_filter: float | None = None, filter_window_size: int = 512, window: str = 'hann', absolute: str | None = None, universal_backprojection=False)

Bases: TimeSeriesProcessingAlgorithm

Preprocesses MSOT time series data. Uses JAX in the background.

__init__(time_factor: int = 3, detector_factor: int = 2, irf: bool = True, hilbert: bool = True, lp_filter: float | None = None, hp_filter: float | None = None, filter_window_size: int = 512, window: str = 'hann', absolute: str | None = None, universal_backprojection=False)

Methods

__init__([time_factor, detector_factor, ...])
add_child(child)
get_algorithm_name()	Get the name of the algorithm.
get_hdf5_group_name()	Return the name of the group in the HDF5 file
pre_compute_filter(n_samples, fs[, irf])	Precompute the filter to be applied.
run(time_series[, pa_data, irf, detectors])	Run the preprocessing step on a given time series and detectors.

static get_algorithm_name() → str | None

Get the name of the algorithm.

Return type:

str or None

static get_hdf5_group_name() → str | None

Return the name of the group in the HDF5 file

Return type:

str or None

pre_compute_filter(n_samples: int, fs: float, irf: ndarray[Any, dtype[_ScalarType_co]] | None = None)

Precompute the filter to be applied.

Parameters:

• n_samples (int) –

• fs (float) –

• irf (Array) –

run(time_series, pa_data=None, irf=None, detectors=None, **kwargs) → Tuple[PATimeSeries, Dict, list | None]

Run the preprocessing step on a given time series and detectors. This allows batch processing, e.g. if the data doesn’t fit into memory.

Parameters:

• time_series –

• pa_data –

• irf –

• detectors –

• kwargs –

Return type:

tuple of PATimeSeries, dict, list