np.int was a deprecated alias for the builtin int.

[a3678911]

My version is R1.10.1. If my numpy's version is beyond or equal to R1.24.0, an error will output like this when skopt.space.Integer is used:
AttributeError: module 'numpy' has no attribute 'int'.
np.int was a deprecated alias for the builtin int. To avoid this error in existing code, use int by itself.
This is how to fix it:
https://stackoverflow.com/questions/74844262/how-can-i-solve-error-module-numpy-has-no-attribute-float-in-python
Hope next version will fix it.

[mitramir55]

same issue.

[samihamine]

I have the same error.

return np.round(X_orig).astype(np.int)
Change it to:

return np.round(X_orig).astype(int)
in the file : skopt/space/transformers.py

EDIT : There is a PR #1159 but didn't pass the CI checks 🤔

[martin-kokos]

Another workaround:

np.int = np.int64

has worked for me for now.

[gorj-tessella]

Same issue

[jwelch1324]

There are several open PRs to fix this issue -- do we know if this project is being actively maintained/monitored?

[tfortimmy]

Seems like it is not: https://snyk.io/advisor/python/scikit-optimize

No update in two years

May be worth looking at other packages e.g. optuna

[rajdecca]

Same issue

[SmokinCaterpillar]

Seems like it is not: https://snyk.io/advisor/python/scikit-optimize

No update in two years

May be worth looking at other packages e.g. optuna

Isn't optuna also relying (at least partly) on skopt?

[boemer00]

Hey, I have tried fixing this issue --> #1187
Hopefully, it will be sorted soon :)

[a3678911]

Hey, I have tried fixing this issue --> #1187 Hopefully, it will be sorted soon :)

Good job! When will be the next Release Version released?

[Y-Srivaishnavi]

Tried fixing this issue too. Would not pass the tests - due to some version name inconsistency.

[baslia]

same issue here :(

[breoganpardo]

Any information on when the issue will be solved in the library main branch?

[cloudy-sfu]

Same issue.

[baslia]

The easiest way is to fork the repo or copy locally and fix it.

[RNarayan73]

A workaround that worked for me is to put these lines before fitting the estimator:

import numpy as np
np.int = int

[svdcvt]

Same issue, Is it hard to update just one line?