bench_plot_incremental_pca.py 文件源码

def variable_batch_size_comparison(data):
    batch_sizes = [i.astype(int) for i in np.linspace(data.shape[0] // 10,
                                                      data.shape[0], num=10)]

    for n_components in [i.astype(int) for i in
                         np.linspace(data.shape[1] // 10,
                                     data.shape[1], num=4)]:
        all_times = defaultdict(list)
        all_errors = defaultdict(list)
        pca = PCA(n_components=n_components)
        rpca = RandomizedPCA(n_components=n_components, random_state=1999)
        results_dict = {k: benchmark(est, data) for k, est in [('pca', pca),
                                                               ('rpca', rpca)]}

        # Create flat baselines to compare the variation over batch size
        all_times['pca'].extend([results_dict['pca']['time']] *
                                len(batch_sizes))
        all_errors['pca'].extend([results_dict['pca']['error']] *
                                 len(batch_sizes))
        all_times['rpca'].extend([results_dict['rpca']['time']] *
                                 len(batch_sizes))
        all_errors['rpca'].extend([results_dict['rpca']['error']] *
                                  len(batch_sizes))
        for batch_size in batch_sizes:
            ipca = IncrementalPCA(n_components=n_components,
                                  batch_size=batch_size)
            results_dict = {k: benchmark(est, data) for k, est in [('ipca',
                                                                   ipca)]}
            all_times['ipca'].append(results_dict['ipca']['time'])
            all_errors['ipca'].append(results_dict['ipca']['error'])

        plot_batch_times(all_times, n_components, batch_sizes, data)
        # RandomizedPCA error is always worse (approx 100x) than other PCA
        # tests
        plot_batch_errors(all_errors, n_components, batch_sizes, data)