maxfuse.model.Fusor.refine_pivots

Fusor.refine_pivots(wt1=0.5, wt2=0.5, svd_components1=None, svd_components2=None, cca_components=None, filter_prop=0, n_iters=1, randomized_svd=False, svd_runs=1, cca_max_iter=2000, verbose=True)[source]

Perform refined matching.

Parameters:

  • wt1 (float, default=0.3) – The shrinkage weight to put on the raw data for arr1.

  • wt2 (float, default=0.3) – The shrinkage weight to put on the raw data for arr2.

  • svd_components1 (None or int, default=None) – If not None, perform SVD to reduce the dimension of self.active_arr1 before feeding it to CCA.

  • svd_components2 (None or int, default=None) – If not None, perform SVD to reduce the dimension of self.active_arr2 before feeding it to CCA.

  • cca_components (None or int, default=None) – Number of CCA components. If None, it is set to 100 or self.active_arr1.shape[1] or self.active_arr2.shape[1], whichever is smaller.

  • filter_prop (float, default=0.) – CCA is performed on top 1-filter_prop slice of the data on which the matched distances are smallest.

  • n_iters (int, default=1) – Number of refinement iterations.

  • randomized_svd (bool, default=False) – Whether to perform randomized SVD.

  • svd_runs (int, default=1) – Perform multiple runs of SVD and the one with lowest Frobenious reconstruction error is selected.

  • cca_max_iter (int, default=2000) – Maximum iteration number for CCA.

  • verbose (bool, default=True) – Whether to print the progress.

Returns:

None