Estimation Options

The class EstimationOptions allows to set the parameters for the estimation, these are:

EstimationOptions

Member

Type (default)

Description

method

tde.TDEMethod

Specifies the time delay method to be used

use_3point_method

bool(True)

If False, use 2-point method for velocity estimation

cache

bool(True)

If True, TDE results are cached

neighbour_options

NeighbourOptions

Options for neighbour selection

cc_options

tde.CCOptions

Used if method = TDEMethod.CC

ca_options

tde.CAOptions

Used if method = TDEMethod.CA

ccf_options

tde.CCFitOptions

Used if method = TDEMethod.CCFit

As you can see, many of these options are also package classes, as NeighbourOptions or NeighbourOptions. These are further explained below. Note that the member method is used to specify which method is employed for time-delay estimation (TDE), these can be based on cross-correlation (CC), conditional averaging (CA) and cross-correlation fitting (CCf), once this is set only the options pertaining the relevant method will be used (cc_options, ca_options or ccf_options).

Neighbour Options

Used for neighbour selection

NeighbourOptions

Member

Type (default)

Description

ccf_min_lag

int(-1)

Checks that maximal correlation time at least ccf_min_lag time the sampling time

min_separation

int(1)

Minimum allowed separation between pixels

max_separation

int(1)

Maximum allowed separation

CC Options

Used if the time-delay estimation is based on cross-correlations (method=tde.TDEMethod.CC)

CCOptions

Member

Type (default)

Description

cc_window

float(none)

at Time lag window for cross-correlation; defaults to 100 samples if None

minimum_cc_value

float(0.5)

Minimum value for cross-correlation maximum; no estimate if less

running_mean

bool(True)

If True, applies running mean to cross-correlation

running_mean_window_max

int(7)

Max length for running mean; warning if exceeded

interpolate

bool(False)

If True, maximizing time lags found by interpolation

CA Options

Used if the time-delay estimation is based on conditional averaging (method=tde.TDEMethod.CA)

CAOptions

Member

Type (default)

Description

min_threshold

float(2.5)

min threshold for conditional averaged events

max_threshold

float(np.inf)

max threshold for conditional averaged events

length_of_return

float(None)

float, length of recorded events. If None, estimated as len(signal) / len(number_of_events) * time_step.

distance

float(0)

minimum distance between peaks

interpolate

bool(False)

If True the maximizing time lags are found by interpolation.

CCF Options

Used if the time-delay estimation is based on cross-correlation fitting (method=tde.TDEMethod.CCF)

CCFitOptions

Member

Type (default)

Description

fit_window

float(100)

The window employed for the fit will be centered at 0 with length 2 * fit_window + 1.

initial_guess

np.array([1, 0, 1])

Initial guess for the optimization procedure.

interpolate

bool(False)

If True the maximizing time lags are found by interpolation.