smooth_table

View page source

The Smoothing Table

Purpose

Each row of the smooth table defines a statistical prior for a function of age and time. To be more specific, each row corresponds to a set of model model_variables that, using bilinear interpolation, define a function of age and time. For each variable in the set its statistical prior is defined by a row of the smooth_grid_table which specifies an age, time, value prior, age difference prior, and time difference prior. (The difference priors represent the smoothing in a mathematical sense.)

Empty

The smooth table must not be empty because there could be no model_variables in this case.

smooth_id

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

smooth_name

This column has type text and has a different value for every row; i.e., the names are unique and can act as substitutes for the primary key. The names are intended to be easier for a human to remember than the ids.

n_age

This column has type integer and is the number of age values in this smoothing grid. It must be greater than zero.

n_time

This column has type integer and is the number of _time values in this smoothing grid. It must be greater than zero.

mulstd_value_prior_id

This column has type integer and its value is the prior_id for the variable that multiplies the value_prior_id standard deviations for this smooth_id .

null

If this variable is not needed, use the value null (which corresponds to a multiplier of one).

mulstd_dage_prior_id

This column has type integer and its value is the prior_id for the variable that multiplies the dage_prior_id standard deviations for this smooth_id .

null

If this variable is not needed, use the value null (which corresponds to a multiplier of one).

mulstd_dtime_prior_id

This column has type integer and its value is the prior_id for the variable that multiplies the dtime_prior_id standard deviations for this smooth_id .

null

If this variable is not needed, use the value null (which corresponds to a multiplier of one).

Removing an Age or Time

It is possible to remove an age (time) from the Rectangular Grid for a specific smooth_id . This is done by decreasing n_age ( n_time ) by one and setting smooth_id to null for the points in the smooth_grid_table that are being removed.

Example

user_change_grid.py

Example

The file smooth_grid_table.py contains an example smooth table.