split_dataset_to_data_and_labels in qiskit not working correctly

Question

I am using source code of split_dataset_to_data_and_labels as I am unable to install qiskit-aqua because of error. But in that source code, "keys" is not being identified. I am using qiskit's tutorials as I am learning, and code I wrote from it, is-:

feature_dim = 2
training_dataset_size = 20
testing_dataset_size = 10
shot = 10000
sample_Total, training_input, test_input, class_labels = ad_hoc_data(training_size=training_dataset_size,
                                                                    test_size=testing_dataset_size,
                                                                    gap=0.3,
                                                                    n=feature_dim,
                                                                    plot_data=True)
datapoints, class_to_label = split_dataset_to_data_and_labels(test_input)
print(class_to_label)

It is giving this error-:

AttributeError                            Traceback (most recent call last)
Input In [6], in <cell line: 10>()
      4 shot = 10000
      5 sample_Total, training_input, test_input, class_labels = ad_hoc_data(training_size=training_dataset_size,
      6                                                                     test_size=testing_dataset_size,
      7                                                                     gap=0.3,
      8                                                                     n=feature_dim,
      9                                                                     plot_data=True)
---> 10 datapoints, class_to_label = split_dataset_to_data_and_labels(test_input)
     11 print(class_to_label)

Input In [4], in split_dataset_to_data_and_labels(dataset, class_names)
     22 labels = []
     23 if class_names is None:
---> 24     sorted_classes_name = sorted(list(dataset.keys()))
     25     class_to_label = {k: idx for idx, k in enumerate(sorted_classes_name)}
     26 else:

AttributeError: 'numpy.ndarray' object has no attribute 'keys'

The link to source code, I am using is https://github.com/Qiskit/qiskit-aqua/blob/main/qiskit/aqua/utils/dataset_helper.py . I found at https://stackoverflow.com/questions/66591189/can-you-help-me-in-this-error-attributeerror-numpy-ndarray-object-has-no-att that flags can be used but it also gives error.

Answer 1

Qiskit Aqua (now deprecated) contains the function ad_hoc_data^[1] which returns sample_total, training_input, test_input, class_labels. When using it, you need to call split_dataset_to_data_and_labels to split the retuned dataset to data and labels.

On the other hand, Qiskit Machine Learning contains a new version of ad_hoc_data^[2] whcih returns training_features, training_labels, test_features, test_labels. There is no need to use split_dataset_to_data_and_labels with it.

It looks like you are using functions from old qiskit.ml with functions from new qiskit_machine_learning. You shouldn't. Try to follow the new Qiskit Machine Learning tutorials (here) to avoid these errors. The new version of ad_hoc_data is used in this tutorial.