-----------------------------------------------------
lines 6-148 of file: at_cascade/create_all_node_db.py
-----------------------------------------------------

{xrst_begin create_all_node_db}

Create an All Node Database
###########################

Prototype
*********
{xrst_literal
   # BEGIN_DEF
   # END_DEF
}

all_node_database
*****************
is a python string containing the name of the
:ref:`all_node_db-name` that is created by this call.
This argument can't be ``None``.

option_all
**********
This argument can't be ``None``.
It is a ``dict`` with a key equal to each
:ref:`option_all_table@Table Format@option_name` that appears in the
option_all table.
The value corresponding to the key is the
:ref:`option_all_table@Table Format@option_value`
in the same row of the option_all_table.
Note that keys must have type ``str`` and all the values will be converted
to ``str``.

fit_goal_table
**************
This specifies the nodes in the :ref:`fit_goal_table-name`.
It must be a ``list`` of ``dict`` representation of the
:ref:`fit_goal_table-name` with one key for each ``dict``.
If the key is node_id (node_name) the corresponding value is
the ``int`` ( ``str`` ) representation of the node.
If this argument is ``None``, the fit_goal table is empty.

split_reference_table
*********************
This specifies the possible reference values for the splitting covariate.
It must be a ``list`` of ``dict`` representation of the
:ref:`split_reference_table-name`.
If this argument is ``None``, the split_reference table is empty.

node_split_table
****************
This specifies the nodes at which the cascade will split by the value
of the splitting covariate.
It must be a ``list`` of ``dict`` representation of the
:ref:`node_split_table-name` with one key for each ``dict``.
If the key is node_id (node_name) the corresponding value is
the ``int`` ( ``str`` ) representation of the node.
If this argument is ``None``, the node_split table is empty.

mulcov_freeze_table
*******************
This specifies the jobs at which the cascade will freeze specific
covariate multipliers.
If this argument is ``None``, the node_split table is empty.
Otherwise, it must be a ``list`` of ``dict`` representation of the
:ref:`mulcov_freeze_table-name` .
We use *row* below to refer to a ``dict`` that is an entry in this ``list`` .

fit_node_id, fit_node_name
==========================
#. If *row*\ ``["fit_node_id"]`` must be ``None``
   if and only if the :ref:`split_reference_table-name` is empty.
#. Otherwise if *row*\ ``["fit_node_id"]`` exists it must be an
   ``int`` representation of the fit_node_id for this job.
#. Otherwise *row*\ ``["fit_node_name"]`` is a
   ``str`` representation of the corresponding node name.

split_reference_id
==================
The value *row*\ ``["split_reference_id"]`` is an
``int`` representation of the split_reference_id for this job or
*row*\ ``["split_reference_name"]`` is a
``str`` representation of the corresponding split reference name.

mulcov_id
=========
The value *row*\ ``["mulcov_id"]`` is an
``int`` representation of the mulcov_id for the covariate multiplier
that is frozen.

omega_grid
**********
is a dictionary with two keys ``age`` and ``time``.
If this argument is ``None``,
the :ref:`omega_grid@omega_age_grid Table` and
the :ref:`omega_grid@omega_time_grid Table` are empty.

age
===
The value *omega_grid*\ [``age``] is a list containing the
age_id values for the omega_grid.
These are indices in the
:ref:`glossary@root_database` age table.
We use the notation *n_omega_age* for the length of the age list.

time
====
The value *omega_grid*\ [``time``] is a list containing the
time_id values for the omega_grid.
These are indices in the *root_database* time table.
We use the notation *n_omega_time* for the length of the time list.

omega_data
**********
This is a python dictionary with a key for each node name
for the :ref:`glossary@root_node` and its descendant.
The value *omega_data[node_name]* is a list.
For each *k*, *omega_data[node_name][k]* is a list.
For *i* equal 0, ..., *n_omega_age*-1
and *j* equal 0, ..., *n_omega_time*-1,

| |tab| *omega_data[node_name][k][ i * n_omega_time + j ]*

is the value of *omega* at the specified node,
the age corresponding to index *i* in *omega_grid*\ [``age``],
and time corresponding to index *j* in *omega_grid*\ [``time``].
If split_reference table is empty, *k* is zero.
Otherwise, let *n_split* be the length of the split_reference table.
For *k* equal 0, ... , *n_split*-1, it specifies the value of
:ref:`split_reference_table@split_reference_id` for the covariate value.

default
=======
The *omega_data* argument is ``None`` if and only if *omega_grid* is ``None``.
If *omega_data* is ``None`` the
:ref:`omega_all@omega_all Table` and
:ref:`omega_all@omega_index Table` will be empty.

cov_reference_table
*******************
This must be ``None`` or a ``list`` of ``dict`` representation of the
:ref:`cov_reference_table-name`.
If it `None``, the :ref:`com_cov_reference-name` routine
is used to create the cov_reference table.

{xrst_end create_all_node_db}