EMAN2 Aligner Manual


Last modified on Tue, 05 Apr 2022 00:12:57 CDT
This document is automatically generated. Please don't edit it.

Aligner Name Parameters Description
frm2d int maxshift: Maximum translation in pixels in any direction. If the solution yields a shift beyond this value in any direction, then the refinement is judged a failure and the original alignment is used as the solution.
FRM2D uses two rotational parameters and one translational parameter
refine emdata mask: A mask to be applied to the image being aligned prior to each similarity comparison.
int maxiter: The maximum number of iterations that can be performed by the Simplex minimizer. default=28
int maxshift: Maximum translation in pixels in any direction. If the solution yields a shift beyond this value in any direction, then the refinement is judged a failure and the original alignment is used as the solution.
int mode: Currently unused
float precision: The precision which, if achieved, can stop the iterative refinement before reaching the maximum iterations. Default is 0.04.
float stepaz: The rotational increment used to create the starting simplex. Default is 5
float stepscale: If set to any non-zero value, scale will be included in the alignment, and this will be the initial step. Images should be edgenormalized. If the scale goes beyond +-30% alignment will fail.
float stepx: The x increment used to create the starting simplex. Default is 1
float stepy: The y increment used to create the starting simplex. Default is 1
int verbose: This will cause debugging information to be printed on the screen for the iterative refinement. Larger numbers -> more info. default=0
transform xform.align2d: The Transform storing the starting guess. If unspecified the identity matrix is used
Refines a preliminary 2D alignment using a simplex algorithm. Subpixel precision.
refine_3d int maxiter: The maximum number of iterations that can be performed by the Simplex minimizer. Default is 100.
int maxshift: Maximum translation in pixels in any direction. If the solution yields a shift beyond this value in any direction, then the refinement is judged a failure and the original alignment is used as the solution.
float precision: The precision which, if achieved, can stop the iterative refinement before reaching the maximum iterations. Default is 0.01.
float spin_coeff: The multiplier appied to the spin (if it is too small or too large the simplex will not converge). Default is 10.
float stepn0: The initial simplex step size in the first quaternion vecotr component. Default is 1.
float stepn1: The initial simplex step size in the second quaternion vecotr component. Default is 1.
float stepn2: The initial simplex step size in the third quaternion vecotr component. Default is 1.
float stepx: The initial simplex step size in x. Default is 1
float stepy: The initial simplex step size in y. Default is 1
float stepz: The initial simplex step size in z. Default is 1.
transform xform.align3d: The Transform storing the starting guess. If unspecified the identity matrix is used
Refines a preliminary 3D alignment using a simplex algorithm. Subpixel precision.
refine_3d_grid float delta: The angular step size. Default is 1.
bool dotrans: Do a translational search. Default is True(1)
float range: The angular range size. Default is 10.
int search: The maximum length of the detectable translational shift - if you supply this parameter you can not supply the maxshiftx, maxshifty or maxshiftz parameters. Each approach is mutually exclusive.
int searchx: The maximum length of the detectable translational shift in the x direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3.
int searchy: The maximum length of the detectable translational shift in the y direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3.
int searchz: The maximum length of the detectable translational shift in the z direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3
bool verbose: Turn this on to have useful information printed to standard out.
transform xform.align3d: The Transform storing the starting guess. If unspecified the identity matrix is used
Refines a preliminary 3D alignment using a simplex algorithm. Subpixel precision.
refinecg emdata mask: A mask to be applied to the image being aligned prior to each similarity comparison.
int maxiter: The maximum number of iterations that can be performed by the Simplex minimizer. default=12
int maxshift: Maximum translation in pixels in any direction. If the solution yields a shift beyond this value in any direction, then the refinement is judged a failure and the original alignment is used as the solution.
int mode: Currently unused
float precision: The precision which, if achieved, can stop the iterative refinement before reaching the maximum iterations. Default is 0.02.
float step: The x increment used to create the starting simplex. Default is 0.1
float stepscale: If set to any non-zero value, scale will be included in the alignment. Images should be edgenormalized. If the scale goes beyond +-30% alignment will fail.
int verbose: This will cause debugging information to be printed on the screen for the iterative refinement. Larger numbers -> more info. default=0
transform xform.align2d: The Transform storing the starting guess. If unspecified the identity matrix is used
Refines a preliminary 2D alignment using a simplex algorithm. Subpixel precision.
rotate_flip int imask:
int rfp_mode: Either 0,1 or 2. A temporary flag for testing the rotational foot print
Performs two rotational alignments, one using the original image and one using the hand-flipped image. Decides which alignment is better using a comparitor and returns it
rotate_flip_iterative int r1: Inner ring, pixels
int r2: Outer ring, pixels
Performs two rotational alignments, iterative style, one using the original image and one using the hand-flipped image. Decides which alignment is better using a comparitor and returns it
rotate_precenter Performs rotational alignment and works accurately if the image is precentered
rotate_symmetry_3d float sym: The symmetry. Default is icos
transform transform: The transform to move to symmetry axis
bool verbose: Turn this on to have useful information printed to standard out.
3D symmetry aligner
rotate_trans_flip_scale emdata flip:
float max: Maximum scaling (default: 1.05)
int maxshift: Maximum translation in pixels
float min: Minimum scaling (default: 0.95)
int nozero: Zero translation not permitted (useful for CCD images)
int rfp_mode: Either 0,1 or 2. A temporary flag for testing the rotational foot print
float step: Scaling step (default: 0.01)
int useflcf: Use Fast Local Correlation Function rather than CCF for translational alignment
int zscore: Either 0 or 1. This option is passed directly to the rotational aligner (default=false)
Performs rotational alignment and follows this with translational and then scaling alignment.
rotate_trans_flip_scale_iter emdata flip:
float max: Maximum scaling (default: 1.05)
int maxiter: Maximum number of iterations
int maxshift: Maximum translation in pixels
float min: Minimum scaling (default: 0.95)
int r1: Inner ring, pixels
int r2: Outer ring, pixels
float step: Scaling step (default: 0.01)
Performs rotational alignment and follows this with translational alignment using the iterative method. Does this for each scale and returns the best
rotate_trans_scale_iter float max: Maximum scaling (default: 1.05)
int maxiter: Maximum number of iterations
int maxshift: Maximum translation in pixels
float min: Minimum scaling (default: 0.95)
int nozero: Zero translation not permitted (useful for CCD images)
int r1: Inner ring, pixels
int r2: Outer ring, pixels
float step: Scaling step (default: 0.01)
int useflcf: Use Fast Local Correlation Function rather than CCF for translational alignment
Performs rotational alignment and follows this with translational alignment using the iterative method. Does this for each scale and returns the best
rotate_translate int maxshift: Maximum translation in pixels
int nozero: Zero translation not permitted (useful for CCD images)
int rfp_mode: Either 0,1 or 2. A temporary flag for testing the rotational foot print
int useflcf: Use Fast Local Correlation Function rather than CCF for translational alignment
int zscore: Either 0 or 1. This option is passed directly to the rotational aligner (default=false)
Performs rotational alignment and follows this with translational alignment.
rotate_translate_2d_to_3d_tree transformarray initxform: An array of Transforms storing the starting positions.
float maxang: maximum angle from initial rotation.
float maxres: Maximum resolution to consider when full sampling is used
int maxshift: maximum shift allowed
float minres: Minimum resolution to consider when full sampling is used
string sym: The symmtery to use as the basis of the spherical sampling. Default is c1 (no symmetry)
bool verbose: Turn this on to have useful information printed to standard out.
3D rotational and translational alignment using a hierarchical approach in Fourier space. Should be very fast and not require 'refine' alignment.
rotate_translate_3d float delta: Angle the separates points on the sphere. This is exclusive of the 'n' paramater. Default is 10
bool dotrans: Do a translational search. Default is True(1)
float dphi: The angle increment in the phi direction. Default is 10
transform initxform: The Transform storing the starting position. If unspecified the identity matrix is used
int n: An alternative to the delta argument, this is the number of points you want generated on the sphere. Default is OFF
string orientgen: Advanced. The orientation generation strategy. Default is eman
float phi0: Lower bound for phi. Default it 0
float phi1: Upper bound for phi. Default it 360
int search: The maximum length of the detectable translational shift - if you supply this parameter you can not supply the maxshiftx, maxshifty or maxshiftz parameters. Each approach is mutually exclusive.
int searchx: The maximum length of the detectable translational shift in the x direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3.
int searchy: The maximum length of the detectable translational shift in the y direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3.
int searchz: The maximum length of the detectable translational shift in the z direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3
string sym: The symmtery to use as the basis of the spherical sampling. Default is c1 (asymmetry).
bool verbose: Turn this on to have useful information printed to standard out.
3D rotational and translational alignment using spherical sampling. Can reduce the search space if symmetry is supplied
rotate_translate_3d_grid float alt0: Lower bound for the altitude direction. Default it 0
float alt1: Upper bound for the altitude direction. Default it 360.0
float az0: Lower bound for the azimuth direction. Default it 0
float az1: Upper bound for the azimuth direction. Default it 180.0
float dalt: The angle increment in the altitude direction. Default is 10
float daz: The angle increment in the azimuth direction. Default is 10
bool dotrans: Do a translational search. Default is True(1)
float dphi: The angle increment in the phi direction. Default is 10
transform initxform: The Transform storing the starting position. If unspecified the identity matrix is used
float phi0: Lower bound for the phi direction. Default it 0
float phi1: Upper bound for the phi direction. Default it 360.0
int search: The maximum length of the detectable translational shift - if you supply this parameter you can not supply the maxshiftx, maxshifty or maxshiftz parameters. Each approach is mutually exclusive.
int searchx: The maximum length of the detectable translational shift in the x direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3.
int searchy: The maximum length of the detectable translational shift in the y direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3.
int searchz: The maximum length of the detectable translational shift in the z direction- if you supply this parameter you can not supply the maxshift parameters. Default is 3
bool verbose: Turn this on to have useful information printed to standard out.
3D rotational and translational alignment using specified ranges and maximum shifts
rotate_translate_3d_local_tree transformarray initxform: An array of Transforms storing the starting positions.
float maxang: maximum angle from initial rotation.
float maxres: maximum resolution to compare
int maxshift: maximum shift allowed
float minres: minimum resolution to compare
bool rand180: Ignore 180 rotation for refine search
bool randphi: Ignore phi constraint for refine search
float sigmathis: Only Fourier voxels larger than sigma times this value will be considered
float sigmato: Only Fourier voxels larger than sigma times this value will be considered
string sym: The symmtery to use as the basis of the spherical sampling. Default is c1 (no symmetry)
bool verbose: Turn this on to have useful information printed to standard out.
EXPERIMENATAL - 3D rotational and translational alignment using a hierarchical approach in Fourier space. Should be very fast and not require 'refine' alignment. This variant performs a locally normalized CCF for translational alignment and should thus work better when the reference is masked, but will be slower.
rotate_translate_3d_tree transformarray initxform: An array of Transforms storing the starting positions.
emdata mask: A mask under which to do the alignment. Mask relative to 'this'
float maxang: maximum angle from initial rotation.
float maxres: maximum resolution to compare
int maxshift: maximum shift allowed
float minres: minimum resolution to compare
bool rand180: Ignore 180 rotation for refine search
bool randphi: Ignore phi constraint for refine search
float sigmathis: Only Fourier voxels larger than sigma times this value will be considered
float sigmato: Only Fourier voxels larger than sigma times this value will be considered
string sym: The symmtery to use as the basis of the spherical sampling. Default is c1 (no symmetry)
bool verbose: Turn this on to have useful information printed to standard out.
3D rotational and translational alignment using a hierarchical approach in Fourier space. Should be very fast and not require 'refine' alignment. 'this' should be the fixed reference, aligned to symmetry axes and mask if provided. If masking, provide the mask rather than premasking the volume.
rotate_translate_bispec int harmonic: If set, uses harmonic power instead of bispectra
int maxshift: Maximum translation in pixels
int nozero: Zero translation not permitted (useful for CCD images)
int rfpn: Passed as the rfp parameter to the bispectrum calculation
int size: Passed as the size parameter to the bispectrum calculation
int useflcf: Use Fast Local Correlation Function rather than CCF for translational alignment
Performs rotational and translational alignment using bispectral invariants.
rotate_translate_flip emdata flip:
int maxshift: Maximum translation in pixels
int rfp_mode: Either 0,1 or 2. A temporary flag for testing the rotational foot print
int usebispec: Uses rotate_translate_bispec for subalignments and ignore rfp_mode.
int usedot:
int useflcf: Use Fast Local Correlation Function rather than CCF for translational alignment
int useharmonic: Uses rotate_translate_bispec in harmonic mode for alignments and ignores rfp_mode.
int zscore: Either 0 or 1. This option is passed directly to the rotational aligner (default=false)
Does two 'rotate_translate' alignments, one to accommodate for possible handedness change. Decided which alignment is better using a comparitor and returns the aligned image as the solution
rotate_translate_flip_iterative emdata flip:
int maxiter: Maximum number of iterations
int maxshift: Maximum translation in pixels
int r1: Inner ring, pixels
int r2: Outer ring, pixels
Does two 'rotate_translate.iterative' alignments, one to accommodate for possible handedness change. Decided which alignment is better using a comparitor and returns the aligned image as the solution
rotate_translate_flip_resample int r1: Inner ring, pixels
int r2: Outer ring, pixels
int tx: Maximum x translation in pixels, Default = 0
int ty: Maximum y translation in pixels, Default = 0
Performs rotational alignment, translation align, and flip by resampling to polar coordinates in real space.
rotate_translate_iterative int maxiter: Maximum number of iterations
int maxshift: Maximum translation in pixels
int nozero: Zero translation not permitted (useful for CCD images)
int r1: Inner ring, pixels
int r2: Outer ring, pixels
int useflcf: Use Fast Local Correlation Function rather than CCF for translational alignment
Performs rotational alignment and follows this with translational alignment using the iterative method.
rotate_translate_resample int r1: Inner ring, pixels
int r2: Outer ring, pixels
int tx: Maximum x translation in pixels, Default = 0
int ty: Maximum y translation in pixels, Default = 0
Performs rotational alignment and translation align by resampling to polar coordinates in real space.
rotate_translate_scale float max: Maximum scaling (default: 1.05)
int maxshift: Maximum translation in pixels
float min: Minimum scaling (default: 0.95)
int nozero: Zero translation not permitted (useful for CCD images)
int rfp_mode: Either 0,1 or 2. A temporary flag for testing the rotational foot print
float step: Scaling step (default: 0.01)
int useflcf: Use Fast Local Correlation Function rather than CCF for translational alignment
int zscore: Either 0 or 1. This option is passed directly to the rotational aligner (default=false)
Performs rotational alignment and follows this with translational and then scaling alignment.
rotate_translate_tree bool flip: Include flip in the alignment search. Default True
float maxres: Maximum resolution to consider when full sampling is used
int maxshift: Maximum acceptable translation. Used only approximately.
int verbose: Turn this on to have useful information printed to standard out.
2D rotational and translational alignment using a hierarchical approach in Fourier space. flip options specifies whether this is RT or RTF. No 'refine' alignment should be required +-1 pixel.
rotational int ambig180: Either 0 or 1. If set, will not try and resolve the 180 degree ambiguity. If not set, it will assume the particle is well centered and resolve the ambiguity that way. default=false
int maxshift: This is provided for compatibility with other aligners. It does absolutely nothing here, as there is an implicit maxshift=0.
int rfp_mode: Either 0,1 or 2. A temporary flag for testing the rotational foot print. O is the original eman1 way. 1 is just using calc_ccf without padding. 2 is using calc_mutual_correlation without padding.
int zscore: Either 0 or 1. If set, will convert per-radius CCF curves into Z-score significnace curves before averaging. In theory this should produce better results by focusing on radii with more alignment information. (default=false)
Performs rotational alignment, even on poorly centered images, but leaves a 180 degree ambiguity which requires a translational alignment to resolve. Usually called internally by rotate_translate aligner.
rotational_bispec int harmonic: If set, uses harmonic power instead of bispectra
int maxshift: This is provided for compatibility with other aligners. It does absolutely nothing here, as there is an implicit maxshift=0.
int rfpn: Passed as the rfp parameter to the bispectrum calculation
int size: Passed as the size parameter to the bispectrum calculation
Performs rotational alignment using bispectral invariants
rotational_iterative int r1: Inner ring, pixels
int r2: Outer ring, pixels
Performs rotational alignment using the SPIDER method of iterating between rotational and translational alingment in real-space
rtf_exhaustive emdata flip:
int maxshift: Maximum translation in pixels
Experimental full 2D alignment with handedness check using semi-exhaustive search (not necessarily better than RTFBest)
rtf_slow_exhaustive float angstep: The angular step (in degrees) to take in the exhaustive search for the solution angle. Typically very small i.e. 3 or smaller.
emdata flip: Optional. This is the flipped version of the images that is being aligned. If specified it will be used for the handedness check, if not a flipped copy of the image will be made
int maxshift: The maximum length of the detectable translational shift
float transtep: The translation step to take when honing the alignment, which occurs after coarse alignment
Experimental full 2D alignment with handedness check using more exhaustive search (not necessarily better than RTFBest)
scale float max: Maximum scaling (default: 1.05)
float min: Minimum scaling (default: 0.95)
float step: Scaling step (default: 0.01)
Performs real space scale alignment
symalign string avger: The sort of averager to use, Default=mean
float delta: Angle the separates points on the sphere. This is exclusive of the 'n' paramater. Default is 10
float dphi: The angle increment in the phi direction. Default is 10
float lphi: Lower bound for phi. Default it 0
string sym: The symmetry under which to do the alignment, Default=c1
float uphi: Upper bound for phi. Default it 359.9
The image is centered and rotated to the standard orientation for the specified symmetry
symalignquat int maxiter: The maximum number of iterations that can be performed by the Simplex minimizer. Default is 100.
int maxshift: Maximum translation in pixels in any direction. If the solution yields a shift beyond this value in any direction, then the refinement is judged a failure and the original alignment is used as the solution.
float precision: The precision which, if achieved, can stop the iterative refinement before reaching the maximum iterations. Default is 0.01.
float spin_coeff: The multiplier appied to the spin (if it is too small or too large the simplex will not converge). Default is 10.
float stepn0: The initial simplex step size in the first quaternion vecotr component. Default is 1.
float stepn1: The initial simplex step size in the second quaternion vecotr component. Default is 1.
float stepn2: The initial simplex step size in the third quaternion vecotr component. Default is 1.
float stepx: The initial simplex step size in x. Default is 1
float stepy: The initial simplex step size in y. Default is 1
float stepz: The initial simplex step size in z. Default is 1.
string sym: The symmettry. Default is c1
transform xform.align3d: The initial guess for to align the particel to sym axis
Finds the symmetry axis using the simplex algorithm.
translational int intonly: Integer pixel translations only
int masked: Treat zero pixels in 'this' as a mask for normalization (default false)
int maxshift: Maximum translation in pixels
int nozero: Zero translation not permitted (useful for CCD images)
int useflcf: Use Fast Local Correlation Function rather than CCF
Translational 2D and 3D alignment by cross-correlation