avgint_table¶

The avgint Table: Defines Average Integrand Cases¶

Purpose¶

Given a value for the model_variables , this table contains the other information necessary so that the predict_command can compute average integrand values that correspond to any possible case, (not just the cases in the data_table ). For an example of how to use this table, see the discussion of parent and child rates below.

Usage¶

This table is only used by the predict_command . It can be changed, and the predict command can be re-run, without needing to re-run any other commands.

Parent Rates¶

The parent rates \(q_k (a, t)\) can be computed by using the following settings in this table:

Set node_id to the parent node id .
Set integrand_id to the integrand corresponding to rate k .
Set age_lower and age_upper set to \(a\).
Set time_lower and time_upper to \(t\).
The weight_id does not matter because we are not averaging over age or time.
Set the Covariates to null.

Note that the rates and integrands have the following correspondence:

( iota , Sincidence ), ( rho , remission ), ( chi , mtexcess ), ( omega , mtother ).

Child Rates¶

The child rates can be computed as adjusted rates \(r_{i,k} (a, t)\) using the following modifications to the parent rate settings above:

Set node_id to the node table node_id for this child.
Set each covariate to its average value for this child.
Set the subgroup_id to the subgroup of interest for this child.

avgint_id¶

This column has type integer and is the primary key for the avgint table. Its initial value is zero, and it increments by one for each row.

integrand_id¶

This column has type integer and is the integrand_id that identifies the integrand for this case.

node_id¶

This column has type integer and is the node_id that identifies the node for this case. If the integrand_name begins with mulcov_ , node_id should be null. Otherwise node_id should not be null.

subgroup_id¶

This column has type integer and is the subgroup_id that identifies the subgroup for this case. The chosen subgroup affects the results through its affect on the covariate multipliers; see mulcov_table .

weight_id¶

This column has type integer and is the weight_id that identifies the weighting used for this case.

null¶

If weight_id is null , the constant weighting is used for this data point.

age_lower¶

This column has type real and is the lower age limit for this case. It must be greater than or equal the minimum age_table value.

age_upper¶

This column has type real and is the upper age limit for this case. It must be greater than or equal the corresponding age_lower and less than or equal the maximum age_table value.

time_lower¶

This column has type real and is the lower time limit for this case. It must be greater than or equal the minimum time_table value.

time_upper¶

This column has type real and is the upper time limit for this case. It must be greater than or equal the corresponding time_lower and less than or equal the maximum time_table value.

Covariates¶

The covariate columns have type real and column names that begin with the two characters x_ . For each valid covariate_id , column x_ covariate_id contains the value, for this measurement, of the covariate specified by covariate_id . The covariate value null is interpreted as the reference value for the corresponding covariate.

Example¶

The file avgint_table.py contains an example avgint table.