Proportional hazards regression with piecewise constant hazards and tabular data.

Source: R/tpchreg.R

Proportional hazards regression with piecewise constant hazards and tabular data.

tpchreg(formula, data, time, weights, last, subset, na.action, 
contrasts = NULL, start.coef = NULL, 
control = list(epsilon = 1.e-8, maxit = 200, trace = FALSE))

Arguments

formula

a formula with 'oe(count, exposure) ~ x1 + ...'
data

a data frame with occurrence/exposure data plus covariates.
time

the time variable, a factor character vector indicating time intervals, or numeric, indicating the start of intervals.
weights

Case weights.
last

If time is numeric, the closing of the last interval.
subset

subset of data, not implemented yet.
na.action

Not implemented yet.
contrasts

Not implemented yet.
start.coef

For the moment equal to zero, not used.
control

list of control parameters for the optimization.

Note

The interpretation of cuts is different from that in link{hpch}. This is intentional.

See also

oe.

Examples


sw <- swepop
sw$deaths <- swedeaths$deaths
fit <- tpchreg(oe(deaths, pop) ~ strata(sex) + I(year - 1990), time = age, last = 101, data = sw)
summary(fit)

#> Covariate             Mean       Coef     Rel.Risk   S.E.    LR p
#> I(year - 1990)        4.489    -0.016     0.984     0.000   0.0000 
#> 
#> Events                    4729403 
#> Total time at risk        457210264 
#> Max. log. likelihood      -18975534 
#> LR test statistic         247379.01 
#> Degrees of freedom        1 
#> Overall p-value           0
#> 
#> Restricted mean survival in (0, 101] : 
#>    women      men 
#> 80.63921 75.59998 
#>