遍历每个元素并使用该元素作为GBM中的输出列名称

Question

I want to loop through a list of parameters in gbm and generate a dataframe that records results of each parameter combination.

Below is my code:

from sklearn.ensemble import GradientBoostingRegressor
import pandas as pd

totalreturn_annual = []
params = {'n_estimators': [1, 10, 50, 100, 200], 'max_depth': [1,3,5,7,9],
        'learning_rate': [0.01,0.05,0.1,0.2,0.3], 'min_samples_split ':[0.1,0.3,0.5,0.7,0.9]}

params = pd.Dataframe(params)

for p in range(16):

        model_cape = GradientBoostingRegressor(random_state = 10, max_features = 'sqrt',
                                               n_estimators = params.iloc[p,0], learning_rate = params.iloc[p,2], 
                                            alpha = params.iloc[p,3], max_depth = params.iloc[p,1],).fit(xs, ys_cape)
totalreturn_annual[p] = np.append(totalreturn_annual, totalreturn_annual_temp)

This is the error that I got:

    totalreturn_annual[p] = totalreturn_annual.append(totalreturn_annual)

IndexError: list assignment index out of range

I wonder why I got the error.

Answer 1

You initialize totalreturn_annual as an empty list in the first line of your code, thus it will not take an index. This is resulting in the index error. This should work:

np.append(totalreturn_annual, totalreturn_annual_temp)

instead of:

totalreturn_annual[p] = np.append(totalreturn_annual, totalreturn_annual_temp)

The np.append(a,b) function appends b to a; a is modified in place.