Scikit-survival

This release adds the *ties* argument to [sksurv.linear_model.CoxPHSurvivalAnalysis](https://scikit-survival.readthedocs.io/en/latest/generated/sksurv.linear_model.CoxPHSurvivalAnalysis.html#sksurv.linear_model.CoxPHSurvivalAnalysis) to choose between Breslow’s and Efron’s likelihood in the presence of tied event times. Moreover, [sksurv.compare.compare_survival()](https://scikit-survival.readthedocs.io/en/latest/generated/sksurv.compare.compare_survival.html#sksurv.compare.compare_survival) has been added, which implements the log-rank hypothesis test for comparing the survival function of 2 or more groups.

Enhancements

- Update API doc of predict function of boosting estimators (75).
- Clarify documentation for GradientBoostingSurvivalAnalysis (78).
- Implement Efron’s likelihood for handling tied event times.
- Implement log-rank test for comparing survival curves.
- Add support for scipy 1.3.1 (66).

Bug fixes

- Re-add *baseline_survival_* and *cum_baseline_hazard_* attributes to [sksurv.linear_model.CoxPHSurvivalAnalysis](https://scikit-survival.readthedocs.io/en/latest/generated/sksurv.linear_model.CoxPHSurvivalAnalysis.html#sksurv.linear_model.CoxPHSurvivalAnalysis) (76).

This release adds support for sklearn 0.21 and pandas 0.24.

Enhancements

- Add reference to IPCRidge (65).
- Use scipy.special.comb instead of deprecated scipy.misc.comb.
- Add support for pandas 0.24 and drop support for 0.20.
- Add support for scikit-learn 0.21 and drop support for 0.20 (71).
- Explain use of intercept in ComponentwiseGradientBoostingSurvivalAnalysis (68)
- Bump Eigen to 3.3.7.

Bug fixes

- Disallow scipy 1.3.0 due to scipy regression (66).

Enhancements

- Add `sksurv.linear_model.CoxnetSurvivalAnalysis.predict_survival_function`
and `sksurv.linear_model.CoxnetSurvivalAnalysis.predict_cumulative_hazard_function`
(46).
- Add `sksurv.nonparametric.SurvivalFunctionEstimator`
and `sksurv.nonparametric.CensoringDistributionEstimator` that
wrap `sksurv.nonparametric.kaplan_meier_estimator` and provide
a `predict_proba` method for evaluating the estimated function on
test data.
- Implement censoring-adjusted C-statistic proposed by Uno et al. (2011)
in `sksurv.metrics.concordance_index_ipcw`.
- Add estimator of cumulative/dynamic AUC of Uno et al. (2007)
in `sksurv.metrics.cumulative_dynamic_auc`.
- Add flchain dataset (see `sksurv.datasets.load_flchain`).

Bug fixes

- The `tied_time` return value of `sksurv.metrics.concordance_index_censored`
now correctly reflects the number of comparable pairs that share the same time
and that are used in computing the concordance index.
- Fix a bug in `sksurv.metrics.concordance_index_censored` where a
pair with risk estimates within tolerance was counted both as
concordant and tied.

This release adds support for Python 3.7 and sklearn 0.20.

Changes:
* Add support for sklearn 0.20 (48).
* Migrate to py.test (50).
* Explicitly request ECOS solver for `sksurv.svm.MinlipSurvivalAnalysis` and `sksurv.svm.HingeLossSurvivalSVM`.
* Add support for Python 3.7 (49).
* Add support for cvxpy >=1.0.
* Add support for numpy 1.15.

This release adds support for numpy 1.14 and pandas up to 0.23. In addition, the new class `sksurv.util.Surv` makes it easier to construct a structured array from numpy arrays, lists, or a pandas data frame.

Changes:

- Support numpy 1.14 and pandas 0.22, 0.23 (36).
- Enable support for cvxopt with Python 3.5+ on Windows (requires cvxopt >=1.1.9).
- Add `max_iter` parameter to `sksurv.svm.MinlipSurvivalAnalysis` and `sksurv.svm.HingeLossSurvivalSVM`.
- Fix score function of `sksurv.svm.NaiveSurvivalSVM` to use concordance index.
- `sksurv.linear_model.CoxnetSurvivalAnalysis` now throws an exception if coefficients get too large (47).
- Add `sksurv.util.Surv` class to ease constructing a structured array (26).

This release adds support for scikit-learn 0.19 and pandas 0.21. In turn, support for older versions is dropped, namely Python 3.4, scikit-learn 0.18, and pandas 0.18.