GEMPACK Frequently Asked Questions

Some of the answers contain references to the GEMPACK user documentation. References are made using the GPD number (for example, GPD-1).
Current versions of these GEMPACK documents can be downloaded..

If you are having a software problem with GEMPACK which is not listed below, please contact GEMPACK Support.

Different Versions of GEMPACK

• What are the main differences between the Source-Code version and the Unlimited Executable-Image version of GEMPACK?
• How much slower is GEMSIM than a TABLO-generated program?
• I am interested in GEMPACK because I want to work with the GTAP model. Which version of GEMPACK do I need?

Getting Started with GEMPACK

• How can I get started using GEMPACK?
• I have used GAMS for general equilibrium models. How similar is GEMPACK?
• Why might I prefer GEMPACK to GAMS for teaching?
• I am familiar with Tom Rutherford's MPSGE language for writing down GE models. Can I solve models written down this way via GEMPACK?

GEMPACK Licence Problems

• I have a GEMPACK licence but a GEMPACK program says it cannot find the licence. What should I do?
• I have purchased an Introductory GEMPACK licence. How do I ensure that GEMPACK programs find the licence?
• ViewHAR, ViewSOL or TABmate is not finding my new GEMPACK licence?

Modelling Issues

• How can I add equations to an existing model?
• How can I model quotas (volume quotas or tariff rate quotas) explicitly using GEMPACK?
• How can I model other inequalities explicitly using GEMPACK?
• Where can I find more about intertemporal models implemented and solved using GEMPACK?
• I want to create a new, large industry from a tiny one. Is this possible?

Solution Problems

• What to do if an NDP error occurs?
• When I display the values of a matrix produced by a TABLO-generated program or GEMSIM, why are there values I did not put there? I expected zeros.
• Why aren't the linearised equations satisfied by a multi-step solution?
• In my model are 2 endogenous variables X and Y which have the particular property that percentage changes in them are equal whenever exogenous variable A is shocked but exogenous variable B is not shocked. I carried out a multi-step simulation in which A and B were both shocked and found that the A-subtotal results for X and Y are not equal, even though there are no B shocks when the A-subtotal is calculated Is this an error?
• My model is structurally singular. What should I do?
• My model is not converging. Would it converge better if GEMPACK used double precision when doing calculations?
• My model is numerically singular. What should I do?
• There is not enough memory to solve my model. What should I do?
• I think that my model would run more quickly if I added more memory. What is the memory limit for GEMPACK?
• My model takes a long time to run a simulation. How can I speed it up?
• Will condensation speed up my model or slow it down?
• One formula (or submatrix, or update) seems to take a long time. How can I find out how long it takes?
• A TABLO-generated program or GEMSIM tells me it cannot open a file which I know is present. What can I do?
• A TABLO-generated program or GEMSIM tells me it cannot open a new file. What can I do?
• Does the order of applying shocks affect the results?
• Can I use GEMPACK to invert a matrix?

Solution Warnings

• There are warnings about equations not being satisfied very accurately. Is this serious?
• There are warnings about variation in endogenous values between passes. Does this mean that my model is not converging?

Converting Data into GEMPACK

• When I convert data from a spreadsheet into GEMPACK the numbers change. Why?
• My data was balanced in the spreadsheet but is no longer balanced when I convert the data to GEMPACK. Why?
• MODHAR tells me there is not enough data on the text file but I know there is.
• I have a complicated data modification task. I am having difficulty building a Stored-input file for MODHAR to do this.
• I have a complicated data modification task. How can I make sure that I can repeat it at some time in the future?
• When I display the values of a matrix produced by a TABLO-generated program or GEMSIM, why are there values I did not put there? I expected zeros.
• How can I convert data from a GAMS model into GEMPACK or vice versa?

Header Array (and Other Binary Files) on Windows PCs

• Are Header Array files produced using LF95 identical to those produced using LF90 or F77L3?
• A friend who uses LF95 gave me a Header Array file which my Release 6.0 GEMPACK programs cannot read. Why is this?
• When a GEMPACK Fortran program (for example, GEMSIM, a TABLO-generated program or CMPHAR) tries to read a Header Array file, I get the error: "E-UNEXPECTED STORAGE METHOD". What can I do?
• When ViewHAR or another Windows programs tries to read a Header Array file, I get the error: "Cannot open file .. Unrecognized Storage Method". What can I do?

WinGEM Problems

• After running TABLO, the TG-program is not found. Why?

Windows Program Problems

• ViewSOL says it cannot open a Solution file. What might the problem be?
• AnalyseGE says it cannot open a Solution file or cannot find the TAB file. What might the problem be?
• TABmate says it cannot find the INF file. What might the problem be?
• I am using Windows 2000 or Windows XP. What should I do about various obscure problems?
• With Windows Vista, I get warnings during installation, and the on-line help is not working. What should I do?

RunMONASH Problems

• How do I decode the Dynamic Information (DIN) file for the MONASH model?
• After installing Release 9 of GEMPACK, my RunMONASH (or Rundynam) is not working. How do I update RunMONASH for Release 9?

Installation Problems

• How do I fix the PATH and Environment variable GPDIR after installation?
• I am using some GEMPACK programs or utilities on a Non-English Language Windows PC. What should I do to avoid various obscure problems?
• A warning appears after installing GEMPACK on Vista. Should I worry?

Source-Code Installation and LAHEY Fortran Compiler Problems

• LTG produces the compiler error "ABORT -- Cannot recognize command line". How do I fix this?
• When I build GEMPACK libraries, LF90 or LF95 produces the compiler error "ABORT -- Cannot recognize command line" . How do I fix this?
• When I build GEMPACK libraries under Windows XP, F77L3 produces an error message. What can I do about this?
• When I use LTG to compile and link with the LF90 compiler, many .MOD, .LIB and .FOR files are left in my directory. What can I do about this?

Intel Fortran

• Should I move to Intel Fortran?

Writing Code

• Can I use GEMPACK subroutines in writing my own programs?
  (Also, some Borland Delphi pascal routines are available for manipulating HAR files.)

Miscellaneous Problems

• A GEMPACK program cannot open a new file. What can be happening?
• I am using Windows 2000 or Windows XP. What should I do about various obscure problems?
• Under Windows 2000 or Windows XP, how do I see all the file extensions or suffixes such as .zip?
• I am using some GEMPACK programs or utilities on a Non-English Language Windows PC. What should I do to avoid various obscure problems?
• GEMPACK does not seem to recognise file or directory names with more than 8 characters. Why?
• GEMPACK does not seem to recognise file or directory names containing spaces. Why?
• The online help files for GEMPACK Windows programs have changed. What's going on?

Appropriate Hardware

• I want a new PC to speed up long simulations. What sort should I get?