Search code examples
pythonlogistic-regressionstatsmodels

statsmodels.api convergence failure

I am trying to get a particular value from: 

result = logit.fit()
print result.summary()

                           Logit Regression Results
==============================================================================
Dep. Variable:                      y   No. Observations:                 8039
Model:                          Logit   Df Residuals:                     8033
Method:                           MLE   Df Model:                            5
Date:                Tue, 15 Sep 2015   Pseudo R-squ.:                 0.01873
Time:                        10:54:35   Log-Likelihood:                -1851.0
converged:                       True   LL-Null:                       -1886.4
                                        LLR p-value:                 7.422e-14

I want to get the value from converged (For above it is True) I could not find a way to get that value out of summary. Is there a way to get whether the model was converged or not?

Solution

  • The additional information provided by the optimization routine are stored in mle_retvals. This should work for you: 

    from statsmodels.discrete.discrete_model import Logit
import numpy as np
np.random.seed(42)

n, d = 20, 10
y = np.random.randint(2, size=n)
X = np.random.rand(n, d)

res = Logit(y, X).fit()

did_converge = res.mle_retvals["converged"]