Weibull Parametric Survival Model
The Weibull parametric survival model is a regression model for time-to-event data that assumes event times follow a Weibull distribution, whose shape parameter allows hazards that rise, fall, or stay constant over time.
와이블 모수적 생존모형은 사건 발생 시간이 와이블 분포를 따른다고 가정하는 생존시간 회귀모형으로, 형상모수에 따라 위험률이 시간에 따라 증가, 감소, 또는 일정할 수 있습니다.
How do I model time to an event when I want a smooth, fully specified hazard over time?
시간에 따른 위험률을 매끄럽고 완전하게 지정해서 사건 발생 시간을 모형화하려면 어떻게 하나요?
The Weibull survival model estimates how predictors affect time to an event while assuming a Weibull distribution, which yields a monotonic hazard and allows extrapolation and smooth survival predictions. It can be expressed as either a proportional hazards or an accelerated failure time model and fits when the Weibull shape is supported by the data.
When to use Weibull Parametric Survival Model
- Time-to-event data with censoring and covariates
- Need to predict or extrapolate survival beyond the observed follow-up
- Reliability or failure-time data where the hazard rises or falls steadily over time
When to use something else
- No distributional assumption is wanted: use Cox proportional hazards regression
- Hazard rises then falls: use a log-normal, log-logistic, or flexible parametric model
Assumptions and requirements
- Event times follow a Weibull distribution (checked with a log-log survival plot)
- Non-informative censoring
- Proportional hazards for hazard ratio interpretation
- Independent subjects
What the analysis reports
- Hazard ratios or time ratios with confidence intervals
- Shape and scale parameters
- Log-likelihood and AIC
- Predicted survival curves and median survival times
How to report it
In the Weibull model, [predictor] was associated with a higher hazard of [event], HR = [value], 95% CI [[value], [value]], p = [value], with an estimated shape parameter of [value].
Related methods
- Cox Proportional Hazards Regression
- Accelerated Failure Time Model
- Kaplan-Meier Estimator
Run Weibull Parametric Survival Model on Quorant
- Data you need
- a table (CSV or Excel) with a time variable, an event indicator and predictor (feature) columns
- Price
- 250 credits for one analysis up to 10,000 rows (complexity level 2); larger or wider data costs more. Optional methodologist review: 1200 credits.
- Checking
- Python results are cross-checked against R reference outputs in every deployment
- Before you pay
- A free pre-analysis diagnosis checks your data and gives an exact quote; nothing is deducted until you approve it.