Perform a decomposition of space-time field in a set of Empirical Orthogonal Functions (EOFs) and Principal components (PCs). The input data set can be optionally weighted in space. By default, the analysis computes "temporal" (T) or classical spatial (S) EOFs depending on if the space dimension is greater than the time dimension. This default behavior can be overridden.
Perform a reconstruction using a set of components previously computed with a PCA. All the reconstructed components are summed. A reconstructed component is simply the "product" of an EOF by its PC. The sum of all reconstructed component is the original field.
Perform a decomposition of space-time field in a set of space-time Empirical Orthogonal Functions (EOFs) and time Principal components (PCs), according to a window parameter.
Performs phase composites of S-T oscillatory field. This field is typically a reconstructed pair of MSSA modes. Composites are evaluated according to an index defined by the first PC of the input field and its derivative. Space weights can be optionally used to compute the PC. A minimal normalized amplitude can be also used: when the index is under value, data are not used to compute phases. It is also possible so specify the angle of the first phase in the 360 degrees phase diagram circle: zero means the the first phase conincides with the maximmum.