Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

A bunch of errors when using categoricals #100

Open
OVPavlov opened this issue Jan 7, 2021 · 0 comments
Open

A bunch of errors when using categoricals #100

OVPavlov opened this issue Jan 7, 2021 · 0 comments

Comments

@OVPavlov
Copy link

OVPavlov commented Jan 7, 2021

Win 10
hpbandster Version: 0.7.4
python37

I use BOHB locally with parameters:
n_iterations=7,
min_budget=1,
max_budget=10,

I'm using custom wrapper for config for simplicity 
search_space_config = {
    'x': uniform(0, 50),
    'y': [1, 2],
}

simple compute function just to test the package
def compute(conf, budget)
    loss = conf['x'] + conf['y'] + (10 / budget)
   return loss

randomly generates errors like this:

...\Python37\site-packages\hpbandster\optimizers\config_generators\bohb.py:140: RuntimeWarning: invalid value encountered in true_divide
  minimize_me = lambda x: max(1e-32, g(x))/max(l(x),1e-32)

or like this:

...\Python37\site-packages\statsmodels\nonparametric\kernels.py:62: RuntimeWarning: divide by zero encountered in true_divide
  kernel_value = np.ones(Xi.size) * h / (num_levels - 1)

or like this:

WARNING:hpbandster:sampled vector: [0.23518812254059032, 0] has EI value nan
WARNING:hpbandster:data in the KDEs:
[[0.00980373 1.        ]
 [0.04901943 1.        ]
 [0.18627439 1.        ]]
[[0.20588224 1.        ]
 [0.77450991 1.        ]
 [0.85294131 1.        ]]
WARNING:hpbandster:bandwidth of the KDEs:
[0.06678006 0.001     ]
[0.25448712 0.001     ]
WARNING:hpbandster:l(x) = inf
WARNING:hpbandster:g(x) = inf

withount categoricals everything works fine

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant