sksurv.datasets.load_aids

sksurv.datasets.load_aids(endpoint='aids')

Load and return the AIDS Clinical Trial dataset

The dataset has 1,151 samples and 11 features. The dataset has 2 endpoints:

  1. AIDS defining event, which occurred for 96 patients (8.3%)
  2. Death, which occurred for 26 patients (2.3%)

See [1], [2] for further description.

Parameters:endpoint (aids|death) – The endpoint
Returns:
  • x (pandas.DataFrame) – The measurements for each patient.
  • y (structured array with 2 fields) – censor: boolean indicating whether the endpoint has been reached or the event time is right censored.

    time: total length of follow-up

    If endpoint is death, the fields are named censor_d and time_d.

References

[1]https://web.archive.org/web/20170114043458/http://www.umass.edu/statdata/statdata/data/
[2]Hosmer, D., Lemeshow, S., May, S.: “Applied Survival Analysis: Regression Modeling of Time to Event Data.” John Wiley & Sons, Inc. (2008)