Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

unhashable type: 'list' by plot parallel coordinate

See original GitHub issue

I get this weird error that I don’t receive with optuna optimizations that are quite the same and can not explain how this error occurs.

This is how I implemented the optimization:

    study = optuna.create_study(storage=optuna.storages.RDBStorage("sqlite:///" + database_name + ".db"), 
                                study_name = database_name + "_study", direction="maximize", load_if_exists=True) 
    study.optimize(lambda trial: objective(trial, preprocessed_dataset, dataset_val, num_samples, database_name, *shift_numbers), n_trials)
    
    # plots of optuna optimization
    fig1 = optuna.visualization.matplotlib.plot_optimization_history(study)
    fig2 = optuna.visualization.matplotlib.plot_parallel_coordinate(study)
    fig3 = optuna.visualization.matplotlib.plot_contour(study)
    fig4 = optuna.visualization.matplotlib.plot_slice(study)
    fig5 = optuna.visualization.matplotlib.plot_param_importances(study)
    fig6 = optuna.visualization.matplotlib.plot_edf(study)
    plt.show()

Environment

Optuna version: 2.10.0
Python version: 3.8.10
OS: Linux-5.4.0-81-generic-x86_64-with-glibc2.29
(Optional) Other libraries and their versions:

Error messages, stack traces, or logs

This is the error I get:

Traceback (most recent call last):
  File "CTGAN_noscale.py", line 286, in <module>
    run_CTGAN(num_samples, n_trials, database_name, *shift_numbers)
  File "CTGAN_noscale.py", line 243, in run_CTGAN
    fig4 = optuna.visualization.matplotlib.plot_parallel_coordinate(study)
  File "/usr/local/lib/python3.8/dist-packages/optuna/_experimental.py", line 68, in new_func
    return func(*args, **kwargs)  # type: ignore
  File "/usr/local/lib/python3.8/dist-packages/optuna/visualization/matplotlib/_parallel_coordinate.py", line 90, in plot_parallel_coordinate
    return _get_parallel_coordinate_plot(study, params, target, target_name)
  File "/usr/local/lib/python3.8/dist-packages/optuna/visualization/matplotlib/_parallel_coordinate.py", line 159, in _get_parallel_coordinate_plot
    values = [vocab[v] for v in values]
  File "/usr/local/lib/python3.8/dist-packages/optuna/visualization/matplotlib/_parallel_coordinate.py", line 159, in <listcomp>
    values = [vocab[v] for v in values]
TypeError: unhashable type: 'list'

Has anyone else already got this problem?

Issue Analytics

State:
Created 2 years ago
Comments:8 (1 by maintainers)

Top GitHub Comments

1reaction

nzw0301commented, Nov 13, 2021

I suppose that sampling index might not be useful for visualisation that is the main concern in this issue. As a workaround, I suggest that choices is converted to str’s list: ["16, 16", "32, 32", ...] and transform sampled value to the tuple of int, eg., (16, 16) before passing to your GANs.

0reactions

Zepp3commented, Nov 13, 2021

Okay, thank you for your tips, they’ll be very useful for me in the future.

Top Results From Across the Web

Unhashable type: 'Point'" while trying to use GeoPandas

I am currently trying to plot a series of ...

Python: TypeError: unhashable type: 'list' - Net-Informations.Com

This error shows that the my_dict key [1,2,3] is List and List is not a hashable type in Python . Dictionary keys must...

getting "TypeError: unhashable type: 'list'" when merging

I'm merging 2 datasets in pandas but i get the error "TypeError: unhashable type: 'list'". The command im running is :

Python TypeError: unhashable type: 'list' Solution

The “TypeError: unhashable type: 'list'” error is raised when you try to assign a list as a key in a dictionary. To solve...

What's New - Xarray

Fix Dataset.assign_coords() resetting all dimension coordinates to default ... Fixed “unhashable type” error trying to read NetCDF file with variable having ...