statsmodels.emplike.aft_el.emplikeAFT#
- class statsmodels.emplike.aft_el.emplikeAFT(endog, exog, censors)[source]#
Class for estimating and conducting inference in an AFT model
- Parameters:
- endog
ndarray Response variables that are subject to random censoring
- exog
ndarray Matrix of covariates
- censorsarray_like
Array with entries 0 or 1. 0 indicates a response was censored.
- endog
- Attributes:
- nobs
int Number of observations
- endog
ndarray Endog array
- exog
ndarray Exogenous variable matrix
- censors
ndarray Censors array but sets the max(endog) to uncensored
- nvar
int Number of exogenous variables
- uncens_nobs
int Number of uncensored observations
- uncens_endog
ndarray Uncensored response variables
- uncens_exog
ndarray Exogenous variables of the uncensored observations
- nobs
Methods
fit()Fits an AFT model and returns results instance
predict(params[, exog])Return the linear predictor, params multiplied by exog
Notes
The data is immediately sorted in order of increasing endogenous variables
The last observation is assumed to be uncensored which makes estimation and inference possible.
Methods