Merge pull request statsmodels#1825 from josef-pkt/phreg_1312_rebased

ENH: Cox Proportional Hazard Model, Phreg rebased closes statsmodels#1312

docs/source/duration.rst

@@ -0,0 +1,81 @@
+.. currentmodule:: statsmodels.duration.hazard_regression
+
+
+.. _duration:
+
+Models for Survival and Duration Analysis
+=========================================
+
+currently contains Cox's Proportional Hazard Model.
+
+
+Examples
+--------
+
+::
+
+  url = "http://vincentarelbundock.github.io/Rdatasets/csv/survival/flchain.csv"
+  data = pd.read_csv(url)
+  del data["chapter"]
+  data = data.dropna()
+  data["lam"] = data["lambda"]
+  data["female"] = 1*(data["sex"] == "F")
+  data["year"] = data["sample.yr"] - min(data["sample.yr"])
+
+  status = np.asarray(data["death"])
+  mod = PHreg.from_formula("futime ~ 0 + age + female + creatinine + " +
+                           "np.sqrt(kappa) + np.sqrt(lam) + year + mgus",
+                           data, status=status, ties="efron")
+  rslt = mod.fit()
+  print(rslt.summary())
+
+Detailed examples can be found here:
+
+.. toctree::
+    :maxdepth: 2
+
+    examples/notebooks/generated/    not yet
+
+There some notebook examples on the Wiki:
+`Wiki notebooks for PHReg and Survival Analysis <https://github.com/statsmodels/statsmodels/wiki/Examples#survival-analysis>`_
+
+
+
+Technical Documentation
+-----------------------
+
+TODO
+
+
+References
+^^^^^^^^^^
+
+References for Cox proportional hazards regression model::
+
+T Therneau (1996).  Extending the Cox model.  Technical report.
+http://www.mayo.edu/research/documents/biostat-58pdf/DOC-10027288
+
+G Rodriguez (2005).  Non-parametric estimation in survival models.
+http://data.princeton.edu/pop509/NonParametricSurvival.pdf
+
+B Gillespie (2006).  Checking the assumptions in the Cox proportional
+hazards model.
+http://www.mwsug.org/proceedings/2006/stats/MWSUG-2006-SD08.pdf
+
+
+Module Reference
+----------------
+
+The model class is:
+
+.. autosummary::
+   :toctree: generated/
+
+   PHReg
+
+The result classe is:
+
+.. autosummary::
+   :toctree: generated/
+
+   PHRegResults

docs/source/index.rst

@@ -97,6 +97,7 @@ Table of Contents
    discretemod
    anova
    tsa
+   duration
    stats
    nonparametric
    gmm

docs/source/release/version0.6.rst

@@ -14,7 +14,7 @@ Major changes:
 Addition of Generalized Estimating Equations GEE
 
 Generalized Estimating Equations
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+--------------------------------
 
 Generalized Estimating Equations (GEE) provide an approach to handling
 dependent data in a regression analysis.  Dependent data arise
@@ -63,9 +63,8 @@ errors that are approximately correct even if the working dependence
 structure is misspecified.
 
 
-
 Seasonality Plots
-~~~~~~~~~~~~~~~~~
+-----------------
 
 Adding functionality to look at seasonality in plots. Two new functions are :func:`sm.graphics.tsa.month_plot` and :func:`sm.graphics.tsa.quarter_plot`. Another function :func:`sm.graphics.tsa.seasonal_plot` is available for power users.
 
@@ -106,7 +105,7 @@ We added a naive seasonal decomposition tool in the same vein as R's ``decompose
 Addition of Linear Mixed Effects Models (MixedLM)
 
 Linear Mixed Effects Models
-~~~~~~~~~~~~~~~~~~~~~~~~~~~
+---------------------------
 
 Linear Mixed Effects models are used for regression analyses involving
 dependent data.  Such data arise when working with longitudinal and

setup.py

@@ -361,7 +361,7 @@ def get_data_files():
     # add all the tests and results files
     for r, ds, fs in os.walk(pjoin(curdir, "statsmodels")):
         r_ = relpath(r, start=curdir)
-        if r_.endswith('results') and 'sandbox' not in r_:
+        if r_.endswith('results'):
             data_files.update({r_.replace(sep, ".") : ["*.csv",
                                                        "*.txt"]})
 

statsmodels/api.py

@@ -14,6 +14,7 @@
 from .discrete.discrete_model import (Poisson, Logit, Probit,
                                       MNLogit, NegativeBinomial)
 from .tsa import api as tsa
+from .duration.hazard_regression import PHReg
 from .nonparametric import api as nonparametric
 from . import distributions
 from .__init__ import test