Model.fromFile (Model)

Create new Model object from model source file(s)

Syntax

m = Model.fromFile(fileName, ...)

Input arguments

fileName [ string ]

Name(s) of model source file(s) that will be loaded, parsed, and converted to a new Model object.

Output arguments

m [ Model ]

New Model object based on the input model source file or files listed in fileNames.

General options

Assign=struct( ) [ struct | empty ]

Assign model parameters and/or steady states from this database at the time the model objects is being created.

AutoDeclareParameters=false [ true | false ]

If true, disregard any parameter declaration sections in the model file, and determine the list of parameters automatically as residual names found in equations but not declared.

BaseYear=@config [ numeric | @config ]

Base year for constructing deterministic time trends; @config means the base year will be read from iris configuration.

Comment='' [ char ]

Text comment attached to the model object.

CheckSyntax=true [ true | false ]

Perform syntax checks on model equations; setting CheckSyntax=false may help reduce load time for larger model objects (provided the model file is known to be free of syntax errors).

Epsilon=eps^(1/4) [ numeric ]

The minimum relative step size for numerical differentiation.

Linear=false [ true | false ]

Indicate linear models.

MakeBkw=@auto [ @auto | @all | string ]

Variables included in the list will be made part of the vector of backward-looking variables; @auto means the variables that do not have any lag in model equations will be put in the vector of forward-looking variables.

AllowMultiple=false [ true | false ]

Allow each variable, shock, or parameter name to be declared (and assigned) more than once in the model file.

Optimal={ } [ cell ]

Specify optimal policy options, see Optimal for policy options models below; only applicable when the keyword min is used in the model source file.

OrderLinks=true [ true | false ]

Reorder !links so that they can be executed sequentially.

RemoveLeads=false [ true | false ]

Remove all leads (aka forward-looking variables) from the state-space vector and keep included only current dates and lags; the leads are not a necessary part of the model solution and can dropped e.g. for memory efficiency reasons in larger model objects.

SteadyOnly=false [ true | false ]

Read in only the steady-state versions of equations (if available).

Std=@auto [ numeric | @auto ]

Default standard deviation for model shocks; @auto means 1 for linear models and log(1.01) for nonlinear models.

UserData=[ ] [ ... ]

Attach user data to the model object.

Options for optimal policy models

The following options for optimal policy models need to be nested within the 'Optimal=' option.

MultiplierPrefix='Mu_' [ char ]

Prefix used to create names for lagrange multipliers associated with the optimal policy problem; the prefix is followed by the equation number.

Nonnegative=[] [ string ]

List of variables constrained to be nonnegative.

Type="discretion" [ "commitment" | "discretion" ]

Type of optimal policy; "discretion" means leads (expectations) are taken as given and not differentiated w.r.t. whereas "commitment" means both lags and leads are differentiated w.r.t.

Description

Loading a model file

The Model.fromFile constructor can be used to read in a model source file named fileNames, and create a model object m based on these. If fileNames is an array of more than one file names then all these files get combined together in order of appearance.

Examples

Plain vanilla model constructor

Read in a model source file named some.model, and declare the model as linear:

m = Model.fromFile("some.model", Linear=true);

Construct model object and immedidately assign parameters

Read in a model source file named some.model, declare the model as linear, and assign some of the model parameters:

m = Model.fromFile("some.model", linear=true, assign=p);

Note that this is equivalent to

m = Model.fromFile("some.model", linear=true);
m = assign(m, p);

unless some of the parameters passed in to the Model.fromFile constructor are needed to evaluate Matlab expressions inside the model source files, such as conditions in the !if or !switch expressions, or angle bracket expressions <...>.