aiPCAClassification

aiPCAClassification classifies a set of particles using the PCA method.

aiPCAClassification needs a configuration file provided by the -u argument. Some options can be overwritten using command line arguments.

Options are:

CudaDeviceIDs

The deviceIDs of the GPUs to use.
Argument for command line: -d or –CudaDeviceIDs
Type: List<int>, a list of integer values
Option is mandatory: true

LP

Low pass filter value.
Argument for command line: -lp or –LP
Type: float
Option is mandatory: true

LPS

Low pass filter sigma value.
Argument for command line: -lps or –LPS
Type: float
Option is mandatory: true

HP

High pass filter value.
Argument for command line: -hp or –HP
Type: float
Option is mandatory: false
Default value if not set: 0

HPS

High pass filter sigma value.
Argument for command line: -hps or –HPS
Type: float
Option is mandatory: false
Default value if not set: 0

StartWith

For a large number of particle, PCA classification can be a long process. To facilitate usage, the process can be split into multiple steps. This indicates the step to start processing with.
Type: one of [COVAR_MATRIX, EIGEN_IMAGES, EIGEN_VALUES, KMEANS, SUM_CLASSES, SUM_PARTICLES, WEIGHT_MATRIX]
Option is mandatory: true
Possible notations:

  • COVAR_MATRIX: COVARMATRIX, COVAR-MATRIX, Covar-Matrix, covar-matrix, CovarMatrix, covarmatrix, COVAR_MATRIX
  • EIGEN_IMAGES: EIGENIMAGES, EIGEN-IMAGES, Eigen-Images, eigen-images, EigenImages, eigenimages, EIGEN_IMAGES
  • EIGEN_VALUES: EIGENVALUES, EIGEN-VALUES, Eigen-Values, eigen-values, EigenValues, eigenvalues, EIGEN_VALUES
  • KMEANS: KMEANS, Kmeans, KMeans, kmeans
  • SUM_CLASSES: SUMCLASSES, SUM-CLASSES, Sum-Classes, sum-classes, SumClasses, sumclasses, SUM_CLASSES
  • SUM_PARTICLES: SUMPARTICLES, SUM-PARTICLES, Sum-Particles, sum-particles, SumParticles, sumparticles, SUM_PARTICLES
  • WEIGHT_MATRIX: WEIGHTMATRIX, WEIGHT-MATRIX, Weight-Matrix, weight-matrix, WeightMatrix, weightmatrix, WEIGHT_MATRIX

EndWith

For a large number of particle, PCA classification can be a long process. To facilitate usage, the process can be split into multiple steps. This indicates the step to end processing with (inclusive).
Type: one of [COVAR_MATRIX, EIGEN_IMAGES, EIGEN_VALUES, KMEANS, SUM_CLASSES, SUM_PARTICLES, WEIGHT_MATRIX]
Option is mandatory: true
Possible notations:

  • COVAR_MATRIX: COVARMATRIX, COVAR-MATRIX, Covar-Matrix, covar-matrix, CovarMatrix, covarmatrix, COVAR_MATRIX
  • EIGEN_IMAGES: EIGENIMAGES, EIGEN-IMAGES, Eigen-Images, eigen-images, EigenImages, eigenimages, EIGEN_IMAGES
  • EIGEN_VALUES: EIGENVALUES, EIGEN-VALUES, Eigen-Values, eigen-values, EigenValues, eigenvalues, EIGEN_VALUES
  • KMEANS: KMEANS, Kmeans, KMeans, kmeans
  • SUM_CLASSES: SUMCLASSES, SUM-CLASSES, Sum-Classes, sum-classes, SumClasses, sumclasses, SUM_CLASSES
  • SUM_PARTICLES: SUMPARTICLES, SUM-PARTICLES, Sum-Particles, sum-particles, SumParticles, sumparticles, SUM_PARTICLES
  • WEIGHT_MATRIX: WEIGHTMATRIX, WEIGHT-MATRIX, Weight-Matrix, weight-matrix, WeightMatrix, weightmatrix, WEIGHT_MATRIX

MotiveList

The list of particles to classify.
Type: string
Option is mandatory: true

MotiveListResult

The filename of the resulting motive list.
Type: string
Option is mandatory: true

Particles

Particle filename. Two naming conventions are possible: filename_PartNr.em/mrc or filename_TomoNr_PartInTomoNr.em/mrc.
Type: string
Option is mandatory: true

Wedges

Wedge filename. Naming convention: filename_WedgeNr.em/mrc.
Type: string
Option is mandatory: true

Mask

Mask filename.
Type: string
Option is mandatory: true

MaskFSC

Mask filename for the mask to use in FSC calculation.
Type: string
Option is mandatory: true

FourierFilterFileName

If set, the particles are not low and high pass filtered but filtered with the provided filter mask.
Type: string
Option is mandatory: false
Default value if not set: ’’

NumberOfEigenVectors

The number of eigen vectors to compute.
Type: int
Option is mandatory: true

NumberOfClasses

The number of classes.
Type: int
Option is mandatory: true

NumberOfParticlesInPCA

Limit the number of particles used for the covariance matrix by randomly choosing X particles from the motivelist. Uses all particles if set to 0.
Type: int
Option is mandatory: true

BlockSize

To speed up caluclations, the covariance matrix is computed in blocks of the given size.
Type: int
Option is mandatory: true

CovarMatrixFilename

The filename for the covariance matrix (.em format).
Type: string
Option is mandatory: true

WeightMatrixFilename

The filename for the weight matrix (.em format).
Type: string
Option is mandatory: true

EigenImagesFilename

The filename for the eigen images. Naming convention: filename_Nr.em/mrc.
Type: string
Option is mandatory: true

EigenVectorsFilename

The filename for the eigen vector (.em format).
Type: string
Option is mandatory: false
Default value if not set: ’’

EigenValuesFilename

The filename for the eigen values (.em format).
Type: string
Option is mandatory: false
Default value if not set: ’’

AverageParticleFile

The filename for the average of all particles (.em or .mrc format).
Type: string
Option is mandatory: false
Default value if not set: ’’

SumClassesFilename

The filename for the sum of each class. Naming convention: filename_ClassNr_IterationNr.em/mrc.
Type: string
Option is mandatory: false
Default value if not set: ’’

SumClassesIterationNumber

The iteration number to use in SumClassesFilename. Default: 0.
Type: int
Option is mandatory: false
Default value if not set: 0