形状(i，j)和(i，j)不对齐:J (dim 1) != i (dim 0)。Statmodels OLS汇总错误

Question

from statsmodels.regression.linear_model import OLS
import numpy as np
X = np.array([[1,2,3],[4,7,5]])
y = np.array([1,2])

mod = OLS(X,y)
res = mod.fit()
print(res.summary())

出现以下错误：

ValueError                                Traceback (most recent call last)
<ipython-input-78-5e3dfbfe5426> in <module>
      6 mod = OLS(X,y)
      7 res = mod.fit()
----> 8 print(res.summary())

~\AppData\Local\Continuum\anaconda3\lib\site-packages\statsmodels\regression\linear_model.py in summary(self, yname, xname, title, alpha)
   2483 
   2484         rsquared_type = '' if self.k_constant else ' (uncentered)'
-> 2485         top_right = [('R-squared' + rsquared_type + ':', ["%#8.3f" % self.rsquared]),
   2486                      ('Adj. R-squared' + rsquared_type + ':', ["%#8.3f" % self.rsquared_adj]),
   2487                      ('F-statistic:', ["%#8.4g" % self.fvalue]),

~\AppData\Local\Continuum\anaconda3\lib\site-packages\statsmodels\tools\decorators.py in __get__(self, obj, type)
     91         _cachedval = _cache.get(name, None)
     92         if _cachedval is None:
---> 93             _cachedval = self.fget(obj)
     94             _cache[name] = _cachedval
     95 

~\AppData\Local\Continuum\anaconda3\lib\site-packages\statsmodels\regression\linear_model.py in rsquared(self)
   1636             return 1 - self.ssr/self.centered_tss
   1637         else:
-> 1638             return 1 - self.ssr/self.uncentered_tss
   1639 
   1640     @cache_readonly

~\AppData\Local\Continuum\anaconda3\lib\site-packages\statsmodels\tools\decorators.py in __get__(self, obj, type)
     91         _cachedval = _cache.get(name, None)
     92         if _cachedval is None:
---> 93             _cachedval = self.fget(obj)
     94             _cache[name] = _cachedval
     95 

~\AppData\Local\Continuum\anaconda3\lib\site-packages\statsmodels\regression\linear_model.py in ssr(self)
   1582         """Sum of squared (whitened) residuals."""
   1583         wresid = self.wresid
-> 1584         return np.dot(wresid, wresid)
   1585 
   1586     @cache_readonly

ValueError: shapes (2,3) and (2,3) not aligned: 3 (dim 1) != 2 (dim 0)

似乎只有当X的维数是n乘1，n是观察值的数量时，它才有效。

下面的代码片段运行起来没有任何问题：

from statsmodels.regression.linear_model import OLS
import numpy as np
X = np.array([[1,2,3],[4,7,5]])
y = np.array([1,2])

mod = OLS(X,y)
res = mod.fit()
print(res.params)

从而产生预期的参数。当X使用k>1时，为什么汇总(例如f_test)会抛出错误？

Answer 1

您刚刚颠倒了OLS参数，请尝试：

mod = OLS(y,X)

那么一切都应该像预期的那样工作，即：

from statsmodels.regression.linear_model import OLS
import numpy as np
X = np.array([[1,2,3],[4,7,5]])
y = np.array([1,2])

mod = OLS(y,X)
res = mod.fit()
print(res.summary())
                            OLS Regression Results                            
==============================================================================
Dep. Variable:                      y   R-squared:                       1.000
Model:                            OLS   Adj. R-squared:                    nan
Method:                 Least Squares   F-statistic:                     0.000
Date:                Sun, 15 Mar 2020   Prob (F-statistic):                nan
Time:                        13:16:38   Log-Likelihood:                 68.110
No. Observations:                   2   AIC:                            -132.2
Df Residuals:                       0   BIC:                            -134.8
Df Model:                           1                                         
Covariance Type:            nonrobust                                         
==============================================================================
                 coef    std err          t      P>|t|      [0.025      0.975]
------------------------------------------------------------------------------
x1             0.0234        inf          0        nan         nan         nan
x2             0.0760        inf          0        nan         nan         nan
x3             0.2749        inf          0        nan         nan         nan
==============================================================================
Omnibus:                          nan   Durbin-Watson:                   1.000
Prob(Omnibus):                    nan   Jarque-Bera (JB):                0.333
Skew:                           0.000   Prob(JB):                        0.846
Kurtosis:                       1.000   Cond. No.                         7.83
==============================================================================

Warnings:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The input rank is higher than the number of observations.