lifelines 0.3.2.0.2

Survival analysis in Python, including Kaplan Meier, Nelson Aalen and regression

What is survival analysis and why should I learn it? Survival analysis was originally developed and applied heavily by the actuarial and medical community. Its purpose was to answer why do events occur now versus later under uncertainity (where events might refer to deaths, disease remission, etc.). This is great for researchers who are interested in measuring lifetimes: they can answer questions like what factors might influence deaths?

But outside of medicine and actuarial science, there are many other interesting and exciting applications of this lesser-known technique, for example: - SaaS providers are interested in measuring customer lifetimes, or time to first behaviours. - sociologists are interested in measure political parties lifetimes, or relationships, or marriages - Businesses are interested in what variables affect lifetime value

lifelines is a pure Python implementation of the best parts of survival analysis. We’d love to hear if you are using lifelines, please ping me at [@cmrn_dp](https://twitter.com/Cmrn_DP) and let me know your thoughts on the library.

Installation:

Dependencies:

The usual Python data stack: Numpy, Scipy, Pandas (a modern version please), Matplotlib

Installing

You can install lifelines using

pip install lifelines

Or getting the bleeding edge version with:

pip install git+https://github.com/CamDavidsonPilon/lifelines.git

from the command line.

Intro to lifelines and survival analysis

Situation: 500 random individuals are born at time 0, currently it is time 12, so we have possibly not observed all death events yet.

censor_after = 12
C = (actual_lifetimes < censor_after) #boolean array

Non-parametrically fit the survival curve:

from lifelines import KaplanMeierFitter

kmf = KaplanMeierFitter()

# fitter methods have an internal plotting method.
# plot the curve with the confidence intervals
kmf.plot()

It looks like 50% of all individuals are dead before time 7.

time            KM-estimate
0.000000        1.000
0.038912        0.998
0.120667        0.996
0.125719        0.994
0.133778        0.992

Non-parametrically fit the cumulative hazard curve:

from lifelines import NelsonAalenFitter

naf = NelsonAalenFitter()

#plot the curve with the confidence intervals
naf.plot()

time       NA-estimate
0.000000     0.000000
0.038912     0.002000
0.120667     0.004004
0.125719     0.006012
0.133778     0.008024

Compare two populations using the logrank test:

from lifelines.statistics import logrank_test

print summary

Results
df: 1
alpha: 0.95
t 0: -1
test: logrank
null distribution: chi squared

__ p-value ___|__ test statistic __|__ test results __
0.00000 |              268.465 |     True

(Less Quick) Intro to lifelines and survival analysis

If you are new to survival analysis, wondering why it is useful, or are interested in lifelines examples and syntax, please check out the Documentation and Tutorials page

Alternatively, you can use the IPython notebooks tutorials, located in the main directory of the repo:

More examples

There are some IPython notebook files in the repo, and you can view them online here.