Dear Colleagues
I hope all is well.
I am doing penalised cox regression in python
I have managed to get the coefficients.
However, I am struggling to get the confidence interval.
I would really appreciate your help
The full code is in this link ( from line code 10 till line code 13)
https://scikit-survival.readthedocs.io/en/stable/user_guide/coxnet.html
The codes are:
best_model = gcv.best_estimator_.named_steps["coxnetsurvivalanalysis"]
best_coefs = pd.DataFrame(
best_model.coef_,
index=Xt.columns,
columns=["coefficient"]
)
non_zero = np.sum(best_coefs.iloc[:, 0] != 0)
print("Number of non-zero coefficients: {}".format(non_zero))
non_zero_coefs = best_coefs.query("coefficient != 0")
coef_order = non_zero_coefs.abs().sort_values("coefficient").index
_, ax = plt.subplots(figsize=(6, 8))
non_zero_coefs.loc[coef_order].plot.barh(ax=ax, legend=False)
ax.set_xlabel("coefficient")
ax.grid(True)
Dear Hatem Ali , here is an implementation of cox reg with l2 penalty in Python that (supposedly) returns confidence intervals:
https://github.com/CamDavidsonPilon/lifelines/blob/9be26a9a8720e8536e9828e954bb91d559a3016f/lifelines/fitters/coxph_fitter.py
Hope that helps! Good luck with your project!
https://towardsai.net/p/machine-learning/survival-analysis-with-python-tutorial-how-what-when-and-why-19a5cfb3c312,u can follow this link.
https://machinelearningmastery.com/lasso-regression-with-python/,this link may be useful.
https://stackoverflow.com/questions/36021730/cox-regression-python,pls. check this link.
https://www.kdnuggets.com/2020/07/guide-survival-analysis-python-part-3.html,this link may be useful.
https://medium.com/the-researchers-guide/survival-analysis-in-python-km-estimate-cox-ph-and-aft-model-5533843c5d5d,also this link may be useful.
see this article to read out more
https://escholarship.org/content/qt3fs5m9tx/qt3fs5m9tx_noSplash_580654820a1845de1673b3c183b66437.pdf
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python