pca

Principal component analysis

Tools for performing PCA, found in Tools/. Note: some related tools are also found in the network models section (enm).

coverlap

Compute the covariance and subspace overlaps between either ENM results or PCA results (or any combination). The covariance overlap was defined in Hess, Phys. Rev. E, 65, 031910, 2002

big-svd

Computes the SVD/PCA for a trajectory. This differs from the svd below in that it uses single precision math and it computes fewer right singular vectors (e.g. if your structure has N atoms and you have T timestemps in the trajectory, then the RSV matrix will be T x 3N as opposed to T x T in the svd tool). See also svd.

kurskew

Kurtosis and skewness for SVD/PCA results

svd

Performs principal component analysis for a trajectory using singular value decomposition, writing out the eigenvalues, left singular vectors (eigenvectors) and the right singular vectors (projection timeseries) as OCTAVE-formatted text files.

svdcolmap

Map the magnitude of a left singular vector onto a PDB file's B-value column. Useful for visualizing which portions of a molecule are mobile for a given SVD mode.