Typeerror: float object is not subscribable is reported in parameter adjustment of xgboost algorithm based on Python Bayesian optimization

be based on Python Bayesian optimization XGBoost The operation of algorithm parameter adjustment is as follows ：

itertargetcolsam…gammamax_depthmin_ch…subsample10.93980.80430.74836.0276.70.651420.94050.72310.26767.382347.60.788630.93880.80480.71676.818708.60.609640.94210.86760.47568.235155.30.669350.93990.90020.97147.254569.20.9067

Report the following error ：

Traceback (most recent call last):
......
suggestion = acq_max(
File "/usr/local/python3/lib/python3.8/site-packages/bayes_opt/util.py", line 65, in acq_max
if max_acq is None or -res.fun[0] >= max_acq:
TypeError: 'float' object is not subscriptable

Reference key codes are as follows ：

def _xgb_logistic_evaluate(max_depth, subsample, gamma, colsample_bytree, min_child_weight):
import xgboost as xgb
params = {

'objective': 'binary:logistic', # The problem of logistic regression dichotomy 
'eval_metric': 'auc',
'max_depth': int(max_depth),
'subsample': subsample, # 0.8
'eta': 0.3,
'gamma': gamma,
'colsample_bytree': colsample_bytree,
'min_child_weight': min_child_weight}
cv_result = xgb.cv(params, self.dtrain,
num_boost_round=30, nfold=5)
return 1.0 * cv_result['test-auc-mean'].iloc[-1]
def evaluate(self, bo_f, pbounds, init_points, n_iter):
bo = BayesianOptimization(
f=bo_f, # Objective function 
pbounds=pbounds, # Value space 
verbose=2, # verbose = 2 Print all when ,verbose = 1 Print the maximum value found in the run ,verbose = 0 Will print nothing 
random_state=1,
)
bo.maximize(init_points=init_points, # Number of steps of random search 
n_iter=n_iter, # Number of Bayesian Optimization iterations performed 
acq='ei')
print(bo.max)
res = bo.max
params_max = res['params']
return params_max

Reference resources stackoverflow Explanation above ：

This is related to a change in scipy 1.8.0, One should use -np.squeeze(res.fun) instead of -res.fun[0]

https://github.com/fmfn/BayesianOptimization/issues/300

The comments in the bug report indicate reverting to scipy 1.7.0 fixes this,

UPDATED: It seems the fix has been merged in the BayesianOptimization package, but the new maintainer is unable to push a release to pypi https://github.com/fmfn/BayesianOptimization/issues/300#issuecomment-1146903850

therefore , Uninstall current scipy 1.8.1, Return to scipy 1.7.0.

[[email protected] bin]# pip3 uninstall scipy
......
Successfully uninstalled scipy-1.8.1
[[email protected] bin]# pip3 install -i https://pypi.tuna.tsinghua.edu.cn/simple scipy==1.7
Successfully installed scipy-1.7.0

Successfully run the Bayesian optimization parameter adjuster again .

Reference resources ：

seul233. python When using Bayesian to optimize random forest TypeError: ‘float’ object is not subscriptable. CSDN Blog . 2022.03

https://stackoverflow.com/questions/71460894/bayesianoptimization-fails-due-to-float-error