对数线性回归

Question

我有一个数据集，我正在尝试对数据集进行日志转换，但我不断收到错误消息，提示“预期 2D 数组，得到 1D 数组”：

dataset3 = df_sheet_map['Set 3']
dataset3

X2 = dataset3.x
Y2 = dataset3.Y
plt.plot(X2, Y2, 'o')
plt.xlabel('x')
plt.ylabel('y')
plt.show()
print('A logarthimic regression model will be used for this data set')

from sklearn.linear_model import LinearRegression
ln_Y2 = np.log(Y2)
plt.plot(X2, ln_Y2, 'o')
plt.xlabel('x')
plt.ylabel('y')
plt.show()

from sklearn.cross_validation import train_test_split
X2_train, X2_test, Y2_train, Y2_test = train_test_split(X2, Y2, test_size= 0.2, random_state=0)
from sklearn.linear_model import LinearRegression
X2_test = X2_test.reshape(1, -1)

regressor = LinearRegression()
regressor.fit(X2_train,Y2_train)

y_pred = regressor.predict([[X2_test]])

但我显示以下错误：

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-59-f9248c950ce4> in <module>()
      2 X2_train, X2_test, Y2_train, Y2_test = train_test_split(X2, Y2, test_size= 0.2, random_state=0)
      3 from sklearn.linear_model import LinearRegression
----> 4 X2_test = X2_test.reshape(1, -1)
      5 
      6 regressor = LinearRegression()

~\Anaconda3\lib\site-packages\pandas\core\generic.py in __getattr__(self, name)
   5065             if self._info_axis._can_hold_identifiers_and_holds_name(name):
   5066                 return self[name]
-> 5067             return object.__getattribute__(self, name)
   5068 
   5069     def __setattr__(self, name, value):

AttributeError: 'Series' object has no attribute 'reshape'

有人可以帮忙吗？我不确定如何像使用 X2_test = X2_test.reshape(1, -1) 之前那样重塑这段代码。我得到了预期的 2d 数组但得到了 1d 的错误。

Answer 1

就我而言

X = df['cases_log']

给了

ValueError：预期的二维数组，得到一维数组：

但是当我使用时

X = df[['cases_log']]

成功了。

它的问题是单括号返回一个熊猫系列obj，另一个返回熊猫数据框。