Summarize parameter objects

Summarize the coefficients of a parameter object by computing the mean, standard deviation, and quantiles for each model term. This is a convenient way to check whether a parameter object has been specified correctly and sampling distributions of the coefficients are as expected.

# S3 method for class 'params_lm'
summary(object, probs = c(0.025, 0.975), ...)

# S3 method for class 'params_mlogit'
summary(object, probs = c(0.025, 0.975), ...)

# S3 method for class 'params_mlogit_list'
summary(object, probs = c(0.025, 0.975), ...)

# S3 method for class 'params_surv'
summary(object, probs = c(0.025, 0.975), ...)

# S3 method for class 'params_surv_list'
summary(object, probs = c(0.025, 0.975), ...)

Arguments

object

An object of the appropriate class.

probs

A numeric vector of probabilities with values in [0,1] used to compute quantiles. By default, the 2.5th and 97.5th percentiles are computed.

...

Additional arguments affecting the summary. Currently unused.

Value

A data.table::data.table that always contains the following columns:

term

The regression term.

mean

The mean value of the regression term.

sd

The standard deviation of the values of the regression term.

In addition, the probs argument determines the quantiles that are computed. By default, the columns 2.5% and 97.5% are returned corresponding to the 2.5th and 97.5th percentiles.

Finally, the following columns may also be present:

parameter

The name of the parameter of interest. This is relevant for any parametric model in which the underlying probability distribution has multiple parameters. For instance, both params_surv and params_surv_list store regression coefficients that are used to model the underlying parameters of the survival distribution (e.g., shape and scale for a Weibull model). Similarly, there are two parameters (mean and sd) for params_lm objects.

model

The name of the statistical model. This is used for a params_surv_list object, where each list element represents a separate model. In a state transition model, each model is a unique health state transition and in a partitioned survival model, there is a separate model for each curve.

to

The health state that is being transitioned to. In params_mlogit and params_mlogit_list objects, there are coefficients for each health state that can be transitioned to.

from

The health state that is being transitions from. This is used for a params_mlogit_list objects where a different multinomial logistic regression is used for each state that can be transitioned from.

See also

For examples, see the the underlying parameter object functions: params_lm(), params_surv(), params_surv_list(), params_mlogit(), and params_mlogit_list().