4677f1eb215c93a651b4bb5c491b2c20b9efb8b2,evaluation/GoodnessOfFit.py,GoodnessOfFit,kolmogorov_smirnov_2sample,#GoodnessOfFit#,83

Before Change

  def kolmogorov_smirnov_2sample(self):
    ks = []
    p = []
    for _ in range(self.repeat_kolmogorov):
      self.resample_new_conditional_samples()
      ks_new, p_new = ks_2samp(self.estimator_conditional_samples, self.proba_model_conditional_samples)
      ks.append(ks_new), p.append(p_new)
    return np.mean(ks), np.mean(p)

  def jarque_bera_test(self):
    jb, p = jarque_bera(self.estimator_conditional_samples)

After Change

    return np.mean(statistics[:,0]), np.mean(statistics[:,1])

  def kolmogorov_smirnov_2sample(self):
    samples = np.asarray([self.sample_conditional_values() for _ in range(self.repeat_kolmogorov)])
    estimator_cond_samples = samples[:, 0]
    probabilistic_cond_samples = samples[:, 1]

    statistics = np.asarray(Parallel(n_jobs=-1)(delayed(ktest_2sample)(estimator_cond_samples[i], probabilistic_cond_samples[i]) for i in range(
      self.repeat_kolmogorov)))
    return np.mean(statistics[:,0]), np.mean(statistics[:,1])

  def kl_divergence(self):
    P = self.probabilistic_model.pdf

Instances