cost: Copenhagen Stroke Study
In pec: Prediction Error Curves for Risk Prediction Models in Survival Analysis

cost	R Documentation

Copenhagen Stroke Study

Description

This data set contains a subset of the data from the Copenhagen stroke study.

Format

This data frame contains the observations of 518 stroke patients :

age: Age of the patients in years.
sex: A factor with two levels female and male.
hypTen: Hypertension, a factor with two levels no and yes.
ihd: History of ischemic heart disease at admission, a factor with two levels no and yes.
prevStroke: History of previous strokes before admission, a factor with two levels no and yes.
othDisease: History of other disabling diseases (e.g. severe dementia), a factor with two levels no and yes.
alcohol: Daily alcohol consumption, a factor with two levels no and yes.
diabetes: Diabetes mellitus status indicating if the glucose level was higher than 11 mmol/L, a factor with two levels no and yes.
smoke: Daily smoking status, a factor with two levels no and yes.
atrialFib: Atrial fibrillation, a factor with two levels no and yes.
hemor: Hemorrhage (stroke subtype), a factor with two levels no (infarction) and yes (hemorrhage).
strokeScore: Scandinavian stroke score at admission to the hospital. Ranges from 0 (worst) to 58 (best).
cholest: Cholesterol level
time: Survival time (in days).
status: Status (0: censored, 1: event).

References

Joergensen HS, Nakayama H, Reith J, Raaschou HO, and Olsen TS. Acute stroke with atrial fibrillation. The Copenhagen Stroke Study. Stroke, 27(10):1765-9, 1996.

Mogensen UB, Ishwaran H, and Gerds TA. Evaluating random forests for survival analysis using prediction error curves. Technical Report 8, University of Copenhagen, Department of Biostatistics, 2010.

pec documentation built on April 11, 2023, 5:55 p.m.