77550cc1f3ccf9537265d127b0f25aa790d2b4b5,skopt/acquisition.py,,gaussian_pi,#,149

Before Change


            mu, std = model.predict(X, return_std=True)

    // make sure mu and std have the same shape so we can divide them below
    std = std.reshape(mu.shape)

    values = np.zeros_like(mu)
    mask = std > 0
    improve = y_opt - xi - mu[mask]

After Change


            mu, std = model.predict(X, return_std=True)

    // check dimensionality of mu, std so we can divide them below
    if (mu.ndim != 1) or (std.ndim != 1):
        raise ValueError("mu and std are {}-dimensional and {}-dimensional, "
                         "however both must be 1-dimensional. Did you train "
                         "your model with an (N, 1) vector instead of an "
                         "(N,) vector?"
                         .format(mu.ndim, std.ndim))

    values = np.zeros_like(mu)
    mask = std > 0
    improve = y_opt - xi - mu[mask]
    scaled = improve / std[mask]
Italian Trulli
In pattern: SUPERPATTERN

Frequency: 3

Non-data size: 9

Instances


Project Name: scikit-optimize/scikit-optimize
Commit Name: 77550cc1f3ccf9537265d127b0f25aa790d2b4b5
Time: 2018-11-07
Author: hinnefe2@illinois.edu
File Name: skopt/acquisition.py
Class Name:
Method Name: gaussian_pi


Project Name: scikit-optimize/scikit-optimize
Commit Name: 77550cc1f3ccf9537265d127b0f25aa790d2b4b5
Time: 2018-11-07
Author: hinnefe2@illinois.edu
File Name: skopt/acquisition.py
Class Name:
Method Name: gaussian_ei


Project Name: yzhao062/pyod
Commit Name: 19985b2eadb39a85e1787298c5e8e9d670bd55f3
Time: 2019-07-17
Author: zhaoy@cmu.edu
File Name: pyod/models/combination.py
Class Name:
Method Name: average