SIDD Simulator of Individual Dynamic Decisions

Downloads.

Model variants.

All versions of SIDD can be downloaded from this page; click on a model version to start.  Each download delivers a single installer file; details concerning model installation can be found here. Please note that the model downloads accessible from this page include limited access to the underlying source code. Individuals who would like access to the full source code should submit a request via the form found here. Enjoy! NOTE: Versions of the model prior to 3.21 (03 Mar 2018)  are parameterised for the UK to data reported by a withdrawn public access file of the Wealth and Assets Survey.

Code snippets.

Selected components of the model code, organised by subheading, can be downloaded here (by clicking on the blue down arrow down-load icons).  The code snippets provided here are not subject to licensing, and can be used in any way (including porting into third-party applications). This code is made available without warranty, and subject to the usual disclaimer indemnifying  the developers from any and all associated damages incurred.

Optimisation

Search for a local minimum in a single dimension

Uses Brent’s method for finding a (local) minimum in a single dimension.

Search for a local minimum in n dimensions

Uses Powell’s method to finding a (local) minimum in multiple dimensions, via calls to Brent’s method (see above).

Linear evaluation of a zero using only function calls

Uses the Bus-Dekker algorithm to search for a zero to a function along a single dimension

Miscellaneous

Gauss Hermite Quadrature

Given the order of the Hermite polynomial, this subroutine returns vectors for abscissae and weights implied by the Gauss Hermite Quadrature.

Function to calculate the cumulative Normal density

Given an observation X, and the mean and standard deviation, the function evaluates the probability associated with a value less than or equal to X.

Function to calculate the inverse of the cumulative Normal

The function returns the value X, under which a normal Given a probability P, and the mean and standard deviation, the function evaluates the value X  associated with a value less than or equal to X.

FORTRAN communications with Excel

Code managing read and write routines between FORTRAN and Excel. FORTRAN communication is undertaken by C# routines accessed via C++ native and a C/CLI wrapper.  While an Intel Fortran module wizard exists to facilitate direct communication with Excel via COM (the Autodice example), we have found the routines generated by the wizard to be unreliable, which motivated development of the code provided here.  C# was selected for Excel communication because it is a low-level Microsoft language, which we hope will provide reliable communications with Excel into the future.  The C/CLI wrapper provides a useful bridge between managed (C#) and unmanaged (C++) code, and communication between C/C++ and FORTRAN is straight-forward.  The code that is provided here could be easily adapted to omit the FORTRAN extension.  A full walk-through of the Visual Studio 2015 Solution set-up is also provided to clarify compiler settings, along with the associated VS2015 source code.

Interpolation

Linear spline interpolation.

A method for finding a (local) minimum in a single dimension.

Cubic spline interpolation.

A method for finding a (local) minimum in multiple dimensions - uses calls to Brent’s method (see above).

Latest version:

Older versions:

Model description:

SIDD Simulator of Individual Dynamic Decisions

Downloads.

Model variants.

All versions of SIDD can be downloaded from this page; click on a model version to start.  Each download delivers a single installer file; details concerning model installation can be found here. Please note that the model downloads accessible from this page include limited access to the underlying source code. Individuals who would like access to the full source code should submit a request via the form found here. Enjoy! NOTE: Versions of the model prior to 3.21 (03 Mar 2018)  are parameterised for the UK to data reported by a withdrawn public access file of the Wealth and Assets Survey.

Code snippets.

Selected components of the model code, organised by subheading, can be downloaded here (by clicking on the blue down arrow down-load icons).  The code snippets provided here are not subject to licensing, and can be used in any way (including porting into third-party applications). This code is made available without warranty, and subject to the usual disclaimer indemnifying  the developers from any and all associated damages incurred.

Optimisation

Brent’s method.

A method for finding a (local) minimum in a single dimension.

Powell’s method.

A method for finding a (local) minimum in multiple dimensions - uses calls to Brent’s method (see above).

Linear evaluation of a zero using only

function calls

Uses the Bus-Dekker algorithm to search for a zero to a function along a single dimension

Interpolation

Linear spline interpolation.

A method for finding a (local) minimum in a single dimension.

Cubic spline interpolation.

A method for finding a (local) minimum in multiple dimensions - uses calls to Brent’s method (see above).

Miscellaneous

Gauss Hermite Quadrature

Given the order of the Hermite polynomial, this subroutine returns vectors for abscissae and weights implied by the Gauss Hermite Quadrature.

Function to calculate the cumulative

Normal density

Given an observation X, and the mean and standard deviation, the function evaluates the probability associated with a value less than or equal to X.

Function to calculate the inverse of the

cumulative Normal

The function returns the value X, under which a normal Given a probability P, and the mean and standard deviation, the function evaluates the value X  associated with a value less than or equal to X.

FORTRAN communications with Excel

Code managing read and write routines between FORTRAN and Excel. FORTRAN communication is undertaken by C# routines accessed via C++ native and a C/CLI wrapper.  While an Intel Fortran module wizard exists to facilitate direct communication with Excel via COM (the Autodice example), we have found the routines generated by the wizard to be unreliable, which motivated development of the code provided here.  C# was selected for Excel communication because it is a low-level Microsoft language, which we hope will provide reliable communications with Excel into the future.  The C/CLI wrapper provides a useful bridge between managed (C#) and unmanaged (C++) code, and communication between C/C++ and FORTRAN is straight-forward.  The code that is provided here could be easily adapted to omit the FORTRAN extension.  A full walk-through of the Visual Studio 2015 Solution set-up is also provided to clarify compiler settings, along with the associated VS2015 source code.
Downloads can only be accessed via the desktop variant of this website.