The geometric threshold method (GTM) identifies SRAs by comparing the strain increments to a threshold. More...

Inheritance diagram for fosanalysis.preprocessing.masking.GTM:

Public Member Functions
	__init__ (self, float delta_max=400.0, int forward_comparison_range=1, float tolerance=0.0, bool to_left=False, bool activate_reverse_sweep=True, args, *kwargs)
	Construct a GTM object.

Public Member Functions inherited from fosanalysis.preprocessing.masking.AnomalyMasker
np.array	run (self, np.array x, np.array y, np.array z, bool make_copy=True, str timespace=None, bool identify_only=False, args, *kwargs)
	Mask strain reading anomalies with `NaN`s.

Public Member Functions inherited from fosanalysis.utils.base.Task
	__init__ (self, args, *kwargs)

Public Member Functions inherited from fosanalysis.utils.base.Base
	__init__ (self, args, *kwargs)
	Construct the object and warn about unused/unknown arguments.

Public Attributes
	activate_reverse_sweep = activate_reverse_sweep
	Switch to activate the reverse sweep.

	delta_max = delta_max
	Maximum plausible absolute strain increment \(\Delta_{\mathrm{max}}\) in [µm/m].

	forward_comparison_range = forward_comparison_range
	Number of neighbors to consider in the forward neighbor comparison.

	to_left = to_left
	Switch for the direction of operation.

	tolerance = tolerance
	Tolerance ratio for the forward comparison to reaccept a candidate.

Protected Member Functions
bool	_compare_forward (self, np.array z, int index, bool to_left)
	Evaluate, if the candidate keeps its SRA flag by comparing it to its succeeding neighbors.

np.array	_run_1d (self, np.array x, np.array z, np.array SRA_array, int start_index=None, int end_index=None, bool to_left=None, bool reverse_state=False, args, *kwargs)
	Flag SRAs between the start and the end index (inclusive).

tuple	_run_2d (self, np.array x, np.array y, np.array z, np.array SRA_array, args, *kwargs)
	GTM has no true 2D operation mode.

Protected Member Functions inherited from fosanalysis.preprocessing.masking.AnomalyMasker
tuple	_map_2d (self, np.array x, np.array y, np.array z, np.array SRA_array, str timespace=None, args, *kwargs)
	Estimate, which entries are strain reading anomalies, in 2D.

Detailed Description

The geometric threshold method (GTM) identifies SRAs by comparing the strain increments to a threshold.

The implementation is improved upon the algorithm presentend in [1]. Each entry is compared to the most recent entry accepted as plausible. If the strain increment \(\Delta_{\mathrm{max}}\) is exceeded, the entry of the current index will be converted to a dropout by setting it to NaN. Dropouts creating a geometrical distance greater than \(\Delta_{\mathrm{max}}\), would result in contigous dropout fields of extended length. To avoid those dopout field, the current entry is additionally compared to its next finite neighbors. The neighbor comparison uses a variable for the range, which allows the user to set amount of neighbors to be compared. Additional fine tuning can be performed by setting a percentage tolerance, which allows a fraction of total neighbor comparisons to exceed \(\Delta_{\mathrm{max}}\).

Definition at line 102 of file masking.py.

Constructor & Destructor Documentation

◆ init()

fosanalysis.preprocessing.masking.GTM.__init__	(		self,
		float	delta_max = 400.0,
		int	forward_comparison_range = 1,
		float	tolerance = 0.0,
		bool	to_left = False,
		bool	activate_reverse_sweep = True,
		*	args,
		**	kwargs )

Construct a GTM object.

Parameters

delta_max	Maximum plausible absolute strain increment \(\Delta_{\mathrm{max}}\) in [µm/m]. For more, see delta_max.
forward_comparison_range	Number of neighbors to consider in the forward neighbor comparison. For more, see forward_comparison_range.
tolerance	Tolerance ratio for the forward comparison to reaccept a candidate. For more, see tolerance.
to_left	Switch for the direction of operation. For more, see to_left.
activate_reverse_sweep	Switch to activate the reverse sweep. For more, see activate_reverse_sweep.
*args	Additional positional arguments, will be passed to the superconstructor.
**kwargs	Additional keyword arguments, will be passed to the superconstructor.

Definition at line 117 of file masking.py.

Here is the call graph for this function:

Here is the caller graph for this function:

Member Function Documentation

◆ _compare_forward()

bool fosanalysis.preprocessing.masking.GTM._compare_forward	(		self,
		np.array	z,
		int	index,
		bool	to_left )

protected

Evaluate, if the candidate keeps its SRA flag by comparing it to its succeeding neighbors.

The candidate keeps its flag, if

\[ r_{\mathrm{tol}} \cdot n_{\mathrm{tot}} < n_{\mathrm{ex}} \]

with the tolerance ratio \(r_{\mathrm{tol}}\), the number of considered succeeding neighbors \(n_{\mathrm{tot}}\), which is the minimum of forward_comparison_range and the remaining elements to the end of the the z array, and the number of considered neighbors \(n_{\mathrm{ex}}\), whose absolute difference to the candidate exceeds delta_max.

Parameters

z	Array of strain data.
to_left	Switch for the direction of operation. For more, see to_left.
index	Current index of the flagged entry.

Returns: Returns, whether the candidate keeps ist SRA flag.

Definition at line 174 of file masking.py.

Here is the caller graph for this function:

◆ _run_1d()

np.array fosanalysis.preprocessing.masking.GTM._run_1d	(		self,
		np.array	x,
		np.array	z,
		np.array	SRA_array,
		int	start_index = None,
		int	end_index = None,
		bool	to_left = None,
		bool	reverse_state = False,
		*	args,
		**	kwargs )

protected

Flag SRAs between the start and the end index (inclusive).

Parameters

x	Array of measuring point positions in accordance to `z`.
z	Array of strain data in accordance to `x`.
SRA_array	Array of boolean values indicating SRAs by `True` and a valid entries by `False`.
start_index	Starting index of the method to filter the array. The starting index is assumed not to be an anomalous reading.
end_index	Last index to check in this sweep.
to_left	Switch for the direction of operation. For more, see to_left.
reverse_state	Switch indicating if the sweep method is currently reverse sweeping to set direction. Default value is `False`. Setting this switch to `True` supresses recursion.
*args	Additional positional arguments, ignored.
**kwargs	Additional keyword arguments, ignored.

Reimplemented from fosanalysis.preprocessing.masking.AnomalyMasker.

Definition at line 214 of file masking.py.

Here is the call graph for this function:

◆ _run_2d()

tuple fosanalysis.preprocessing.masking.GTM._run_2d	(		self,
		np.array	x,
		np.array	y,
		np.array	z,
		np.array	SRA_array,
		*	args,
		**	kwargs )

protected

GTM has no true 2D operation mode.

Set timespace to "1d_space"!

Reimplemented from fosanalysis.preprocessing.masking.AnomalyMasker.

Definition at line 277 of file masking.py.

Member Data Documentation

◆ activate_reverse_sweep

fosanalysis.preprocessing.masking.GTM.activate_reverse_sweep = activate_reverse_sweep

Switch to activate the reverse sweep.

Defaults to True A reverse sweep is trigged, whenever a candidate is accepted and the index difference to the previously accepted one is <1.

Definition at line 173 of file masking.py.

◆ delta_max

fosanalysis.preprocessing.masking.GTM.delta_max = delta_max

Maximum plausible absolute strain increment \(\Delta_{\mathrm{max}}\) in [µm/m].

Defaults to 400.0 µm/m.

Definition at line 140 of file masking.py.

◆ forward_comparison_range

fosanalysis.preprocessing.masking.GTM.forward_comparison_range = forward_comparison_range

Number of neighbors to consider in the forward neighbor comparison.

Defaults to 1. Whenever an entry is flagged by exceeding delta_max to the most recently trusted entry, a forward neighbor comparison is triggered. Setting this to 0 (zero) deactivates the neighbor comparison. The default parameter combination is equivalent to comparing the candidate to both its direct neighbors:

Parameter	Value
forward_comparison_range	1
tolerance	0.0

Definition at line 151 of file masking.py.

◆ to_left

fosanalysis.preprocessing.masking.GTM.to_left = to_left

Switch for the direction of operation.

Defaults to False (from left to right).

Definition at line 168 of file masking.py.

◆ tolerance

fosanalysis.preprocessing.masking.GTM.tolerance = tolerance

Tolerance ratio for the forward comparison to reaccept a candidate.

Defaults to 0.0. Whenever an entry is flagged by exceeding delta_max to the most recently trusted entry, a forward neighbor comparison is triggered. Then, the number of neighbors exceeding delta_max to the candidate is counted. The ratio of this number dividedy by forward_comparison_range is compared to the tolerance ratio. Setting this to 0.0 rejects the point when at least a single neighbor violates delta_max. Setting this to 1.0 would reaccept every point, rendering the whole procedure useless. The default parameter combination is equivalent to comparing the candidate to both its direct neighbors:

Parameter	Value
forward_comparison_range	1
tolerance	0.0

Definition at line 165 of file masking.py.

The documentation for this class was generated from the following file:

src/fosanalysis/preprocessing/masking.py

Public Member Functions

Public Attributes

Protected Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ __init__()

Member Function Documentation

◆ _compare_forward()

◆ _run_1d()

◆ _run_2d()

Member Data Documentation

◆ activate_reverse_sweep

◆ delta_max

◆ forward_comparison_range

◆ to_left

◆ tolerance

◆ init()