当我在 jupyter 记事本上运行以下代码时，我得到了(属性错误："数据帧"对象没有属性"as_matrix"(，当我在 jupyter 记事本上运行以下代码时，我得到了 2 行 #create x 和 y 变量
X = features_df.as_matrix()
y = df['Price'].as_matrix()
我的整个代码如下
#developing model to predict houses prices in Australia
#importing needed libraries
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn import ensemble
from sklearn.metrics import mean_absolute_error
import sklearn.externals
# importing the file path
df = pd.read_csv('~/mypython/machine_learning/machine_learning/housing/Melbourne_housing_FULL.csv')
#removing less related dimentions
del df['Address']
del df['Method']
del df['SellerG']
del df['Date']
del df['Postcode']
del df['Lattitude']
del df['Longtitude']
del df['Regionname']
del df['Propertycount']
#delete raws with any empty value
df.dropna(axis = 0 ,how = 'any' ,thresh = None, subset = None, inplace = True)
#converting non-numerical values to numerical values using pandas
features_df = pd.get_dummies(df, columns=['Suburb', 'CouncilArea', 'Type'])
# delete price because it's the dependant varilable
del features_df['Price']
#create x & y variables
X = features_df.as_matrix()
y = df['Price'].as_matrix()
X_train, X_test, y_train, y_test=train_test_split(X, y, test_size=0.3,random_state=0)
model = ensembel.GradientBoostingRegressor(
n_estimators=150,
learning_rate=0.1,
max_depth=30,
min_sample_split= 4,
min_samples_leaf=6,
max_features=0.6,
loss="huber")
model.fit(X_train,y_train)
joblib.dumb(model,"house_train_model.pkl")
mse=mean_absolute_error(y_train_model, model.predict(X_train))
print("Training set mean absolute error:%.2f"%mse)
你应该使用这个
X = features_df.values
y = df['Price'].values