Merge pull request #422 from javadnoorb/patch-1

Update Survival analysis with lifelines.rst
CamDavidsonPilon · Mar 8, 2018 · 4d22dd9 · 4d22dd9
2 parents bbd7e5a + 54fa651
commit 4d22dd9
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 2 deletions.
diff --git a/docs/Survival analysis with lifelines.rst b/docs/Survival analysis with lifelines.rst
@@ -259,7 +259,7 @@ see that very few leaders make it past 20 years in office. Of course,
 like all good stats, we need to report how uncertain we are about these
 point estimates, i.e., we need confidence intervals. They are computed in
 the call to ``fit``, and located under the ``confidence_interval_``
-property. (The mathematics are found in `these notes <http://courses.nus.edu.sg/course/stacar/internet/st3242/handouts/notes2.pdf>`_.)
+property. (The method uses exponential Greenwood confidence interval. The mathematics are found in `these notes <https://www.math.wustl.edu/%7Esawyer/handouts/greenwood.pdf>`_.)
 
 .. math::  S(t) = Pr( T > t) 
 

diff --git a/lifelines/fitters/kaplan_meier_fitter.py b/lifelines/fitters/kaplan_meier_fitter.py
@@ -90,7 +90,9 @@ def fit(self, durations, event_observed=None, timeline=None, entry=None, label='
         return self
 
     def _bounds(self, cumulative_sq_, alpha, ci_labels):
-        # See http://courses.nus.edu.sg/course/stacar/internet/st3242/handouts/notes2.pdf
+        # This method calculates confidence intervals using the exponential Greenwood formula.
+        # See https://www.math.wustl.edu/%7Esawyer/handouts/greenwood.pdf
+
         alpha2 = inv_normal_cdf((1. + alpha) / 2.)
         df = pd.DataFrame(index=self.timeline)
         v = np.log(self.__estimate.values)