init_command#

The Initialize Command#

Syntax#

dismod_at database init

Purpose#

This command initializes the data flow. To be specific, it begins by deleting any existing output tables, except for the log table, and then creates new versions of the following tables:

age_avg_table	The Age Average Table
bnd_mulcov_table	The Covariate and Multiplier Bound Table
data_subset_table	The Data Subset Table
start_var_table	Starting Values Used During a Fit
scale_var_table	Scaling Fixed Effects Objective and Constraints
var_table	Identifying Model Variables

Deleted Tables#

This routine begins by deleting any existing output tables , except for the log_table .

Changing Values#

You must re-run this command when ever any of the input tables is changed. The option table is an exception to this rule; see Changing Values .

database#

Is an http://www.sqlite.org/sqlite/ database containing the dismod_at input tables which are not modified.

age_avg_table#

A new age_avg_table is created when this command is run. In fact, this table is recreated when any command is run except for the python commands and the set_command .

var_table#

A new var_table is created with the information that maps a var_id to its meaning in terms of the model_variables .

data_subset_table#

A new data_subset_table is created. This makes explicit exactly which rows of the data table are used.

hold_out#

This column is set to zero.

density_id#

This column is set to the corresponding density_id in the data table.

eta#

This column is set to the corresponding eta in the data table.

nu#

This column is set to the corresponding nu in the data table.

start_var_table#

A new start_var_table is created using the means of the priors for the model variables.

scale_var_table#

A new scale_var_table is created using the means of the priors for the model variables.

bnd_mulcov_table#

A new bnd_mulcov table is created using null for max_mulcov ; i.e., plus infinity.

Example#

The file init_command.py contains an example and test using this command.