我想使用 sklearn.compose.TransformedTargetRegressor
,如 this answer 中所示.但是,转换器是自定义转换器,我遇到了错误。
在这个最小的示例中,目标值应乘以 10,然后在预测时再次除以 10。 (在我的实际应用中,目标值必须从非数字格式转换为数字格式。)
import numpy as np
import sklearn
from sklearn.compose import TransformedTargetRegressor
from sklearn.linear_model import LinearRegression
class MyTransform(sklearn.base.TransformerMixin):
def fit(self, *_, **__):
return self
def transform(self, X):
return np.array(X)*10
def inverse_transform(self, X):
return np.array(X)/10
def MyLinearRegression():
return TransformedTargetRegressor(
regressor=LinearRegression(),
transformer=MyTransform()
)
if __name__ == '__main__':
model = MyLinearRegression()
model.fit(X=[[1], [2], [3]], y=[1, 2, 3]) # raises TypeError
这引发了:
Traceback (most recent call last):
File "<input>", line 1, in <module>
File "C:\Program Files\JetBrains\PyCharm Community Edition 2019.3.3\plugins\python-ce\helpers\pydev\_pydev_bundle\pydev_umd.py", line 197, in runfile
pydev_imports.execfile(filename, global_vars, local_vars) # execute the script
File "C:\Program Files\JetBrains\PyCharm Community Edition 2019.3.3\plugins\python-ce\helpers\pydev\_pydev_imps\_pydev_execfile.py", line 18, in execfile
exec(compile(contents+"\n", file, 'exec'), glob, loc)
File "C:/Users/me/.PyCharmCE2019.3/config/scratches/scratch.py", line 26, in <module>
model.fit(X=[[1], [2], [3]], y=[1, 2, 3]) # raises TypeError
File "C:\Users\me\.virtualenvs\project--3333Ox_\lib\site-packages\sklearn\compose\_target.py", line 185, in fit
self._fit_transformer(y_2d)
File "C:\Users\me\.virtualenvs\project--3333Ox_\lib\site-packages\sklearn\compose\_target.py", line 127, in _fit_transformer
self.transformer_ = clone(self.transformer)
File "C:\Users\me\.virtualenvs\project--3333Ox_\lib\site-packages\sklearn\base.py", line 64, in clone
raise TypeError("Cannot clone object '%s' (type %s): "
TypeError: Cannot clone object '<__main__.MyTransform object at 0x000001653B2FA9A0>' (type <class '__main__.MyTransform'>): it does not seem to be a scikit-learn estimator as it does not implement a 'get_params' methods.
最佳答案
您只需要继承 sklearn.base.BaseEstimator 以及 transformermixin :)。类型错误说:
it does not seem to be a scikit-learn estimator
所以你只需要让它成为一个 :D。下面的代码应该可以工作。
import numpy as np
import sklearn
from sklearn.compose import TransformedTargetRegressor
from sklearn.linear_model import LinearRegression
class MyTransform(sklearn.base.BaseEstimator, sklearn.base.TransformerMixin):
def fit(self, *_, **__):
return self
def transform(self, X):
return np.array(X)*10
def inverse_transform(self, X):
return np.array(X)/10
def MyLinearRegression():
return TransformedTargetRegressor(
regressor=LinearRegression(),
transformer=MyTransform()
)
model = MyLinearRegression()
model.fit(X=[[1], [2], [3]], y=[1, 2, 3])
关于python - 如何将 sklearn 的 TransformedTargetRegressor 与自定义数据转换器一起使用?,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/61365612/