3e38fe09a87e6ef05289f3cbe6ffa03e2dc716e8,metric_learn/rca.py,RCA,fit,#RCA#Any#Any#,78

Before Change

    // PCA projection to remove noise and redundant information.
    M_pca = None
    if self.params["pca_comps"] is not None:
      pca = decomposition.PCA(n_components=self.params["pca_comps"],
                              svd_solver="full")
      data = pca.fit_transform(data)
      d = data.shape[1]
      M_pca = pca.components_
    else:
      data -= data.mean(axis=0)

    chunk_mask, chunk_data = self._process_chunks(data, chunks)
    inner_cov = np.cov(chunk_data, rowvar=0, bias=1)
    rank = np.linalg.matrix_rank(inner_cov)

After Change

    
    data, M_pca = self._process_data(data)

    chunks = np.asanyarray(chunks, dtype=int)
    chunk_mask, chunked_data = _chunk_mean_centering(data, chunks)

    inner_cov = np.cov(chunked_data, rowvar=0, bias=1)
    dim = self._check_dimension(np.linalg.matrix_rank(inner_cov))

Instances