The Model Coupling Toolkit API Reference Manual: MCT v. 2.8

J. W. Larson
R. L. Jacob
E. Ong
R. Loy


Mathematics and Computer Science Division, Argonne National Laboratory

This paper has not been published and should be regarded as an Internal Report from MCS. Permission to quote from this Technical Note should be obtained from the MCS Division of Argonne National Laboratory.

This Technical Note was produced for the Scientific Discovery through Advanced Computing (SciDAC) project.

Version	Version	Pages Affected/	Aproval
Number	Date	Extent of Changes	Authority
Version 1	December 13, 2000	First draft (before review)
Version 1	February 16, 2001	Add more routines
Version 1	June 6, 2001	Convert to pure API's doc
Version 1	Apr 24, 2002	Update with latest source
Version 1.0	Nov 14, 2002	1.0 Version
Version 2.0.0	Apr 23, 2004	2.0.0 Version
Version 2.0.1	May 18, 2004	2.0.1 Version
Version 2.1.0	Feb 11, 2005	2.1.0 Version
Version 2.2.0	Dec 01, 2005	2.2.0 Version
Version 2.2.1	Apr 22, 2006	2.2.1 Version
Version 2.2.2	Sep 08, 2006	2.2.2 Version
Version 2.2.3	Oct 16, 2006	2.2.3 Version
Version 2.3.0	Jan 10, 2007	2.3.0 Version
Version 2.4.0	Aug 17, 2007	2.4.0 Version
Version 2.4.1	Nov 21, 2007	2.4.1 Version
Version 2.5.0	Jan 28, 2008	2.5.0 Version
Version 2.5.1	May 20, 2008	2.5.1 Version
Version 2.6.0	Mar 05, 2009	2.6.0 Version
Version 2.7.0	Jan 05, 2010	2.7.0 Version
Version 2.7.1	Feb 28, 2010	2.7.1 Version
Version 2.7.2	Nov 30, 2010	2.7.2 Version
Version 2.7.3	Jan 25, 2011	2.7.3 Version
Version 2.7.4	Mar 07, 2012	2.7.4 Version
Version 2.8.0	Apr 30, 2012	2.8.0 Version
Version 2.8.1	Jul 05, 2012	2.8.1 Version
Version 2.8.2	Sep 12, 2012	2.8.2 Version
Version 2.8.3	Dec 17, 2012	2.8.3 Version

This document describes the Application Program Interfaces (APIs) for the Model Coupling Toolkit (MCT).

For functions that take a Fortran90 real argument, either a scalar or a vector, MCT provides both double and single precision versions. Only the single precision version are described here denoted by SP. The double precision versions are otherwise identical.

• Contents
• 1 Basic API's and associated communication routines
  • 1 MCTWorld
    • 1.1 Module m_MCTWorld - MCTWorld Class (Source File: m_MCTWorld.F90)
  • 2 The Attribute Vector
    • 2.1 Module m_AttrVect - Multi-field Storage (Source File: m_AttrVect.F90)
    • 2.2 Module m_AttrVectComms - MPI Communications Methods for the AttrVect (Source File: m_AttrVectComms.F90)
    • 2.3 Module m_AttrVectReduce - Local/Distributed AttrVect Reduction Ops. (Source File: m_AttrVectReduce.F90)
  • 3 Global Segment Map
    • 3.1 Module m_GlobalSegMap - a nontrivial 1-D decomposition of an array. (Source File: m_GlobalSegMap.F90)
    • 3.2 Module m_GlobalSegMapComms - GlobalSegMap Communications Support (Source File: m_GlobalSegMapComms.F90)
  • 4 The Router
    • 4.1 Module m_Router - Router class (Source File: m_Router.F90)
  • 5 The General Grid
    • 5.1 Module m_GeneralGrid - Physical Coordinate Grid Information Storage (Source File: m_GeneralGrid.F90)
    • 5.2 Module m_GeneralGridComms - Communications for the GeneralGrid type. (Source File: m_GeneralGridComms.F90)
  • 6 The Navigator
    • 6.1 Module m_Navigator - An Object for Indexing Segments of a Vector (Source File: m_Navigator.F90)
  • 7 The Global Map
    • 7.1 Module m_GlobalMap - One-Dimensional Domain Decomposition Descriptor (Source File: m_GlobalMap.F90)
  
  
• 2 High Level API's
  • 8 Sending and Receiving Attribute Vectors
    • 8.1 Module m_Transfer - Routines for the MxN transfer of Attribute Vectors (Source File: m_Transfer.F90)
  • 9 Rearranging Attribute Vectors
    • 9.1 Module m_Rearranger - Remaps an AttrVect within a group of processes (Source File: m_Rearranger.F90)
  • 10 Sprase Matrix Support
    • 10.1 Module m_SparseMatrix - Sparse Matrix Object (Source File: m_SparseMatrix.F90)
    • 10.2 Module m_SparseMatrixComms - sparse matrix communications methods. (Source File: m_SparseMatrixComms.F90)
    • 10.3 Module m_SparseMatrixDecomp - Parallel sparse matrix decomposition. (Source File: m_SparseMatrixDecomp.F90)
    • 10.4 Module m_SparseMatrixToMaps - Maps from the Sparse Matrix (Source File: m_SparseMatrixToMaps.F90)
    • 10.5 Module m_SparseMatrixPlus - Class Parallel for Matrix-Vector Multiplication (Source File: m_SparseMatrixPlus.F90)
  • 11 Matrix Vector Multiplication
    • 11.1 Module m_MatAttrVectMul - Sparse Matrix AttrVect Multipication. (Source File: m_MatAttrVectMul.F90)
  • 12 Spatial Integration and Averaging
    • 12.1 Module m_SpatialIntegral - Spatial Integrals and Averages using a GeneralGrid (Source File: m_SpatialIntegral.F90)
    • 12.2 Module m_SpatialIntegralV - Spatial Integrals and Averages using vectors of weights (Source File: m_SpatialIntegralV.F90)
  • 13 Merging of Flux and State Data from Multiple Sources
    • 13.1 Module m_Merge - Merge flux and state data from multiple sources. (Source File: m_Merge.F90)
  • 14 Time Averaging
    • 14.1 Module m_Accumulator - Time Averaging/Accumlation Buffer (Source File: m_Accumulator.F90)
    • 14.2 Module m_AccumulatorComms - MPI Communication Methods for the Accumulator (Source File: m_AccumulatorComms.F90)
  • 15 Global To Local Index Translation
    • 15.1 Module m_GlobalToLocal - Global to Local Index Translation (Source File: m_GlobalToLocal.F90)
  • 16 Convert From Global Map To Global Segment Map
    • 16.1 Module m_ConvertMaps - Conversion Between MCT Domain Decomposition Descriptors (Source File: m_ConvertMaps.F90)
  
  
• 3 Documentation of MPEU Datatypes Used to Define MCT Datatypes
  • 17 The String Datatype
    • 17.1 Module m_String - The String Datatype (Source File: m_String.F90)
  • 18 The List Datatype
    • 18.1 Module m_List - A List Manager (Source File: m_List.F90)
  • 19 InPackF90
  • 20 Package Overview
    • 20.1 Resource Files
    • 20.2 A Quick Stroll
    • 20.3 Main Routine/Functions
    • 20.4 Package History
    • 20.5 Bugs
    • 20.6 Availability
    • 20.7 I90_LoadF() -- Loads resource file into memory. (Source File: m_inpak90.F90)
    • 20.8 I90_Label() -- Selects a label (record). (Source File: m_inpak90.F90)
    • 20.9 I90_GLine() -- Selects next line. (Source File: m_inpak90.F90)
    • 20.10 I90_GToken() -- Gets next token. (Source File: m_inpak90.F90)
    • 20.11 i90_GFloat() -- Returns next float number. (Source File: m_inpak90.F90)
    • 20.12 I90_GInt() -- Returns next integer number. (Source File: m_inpak90.F90)
    • 20.13 i90_AtoF() -- Translates ASCII (string) to float. (Source File: m_inpak90.F90)
    • 20.14 I90_AtoI() -- Translates ASCII (strings) to integer. (Source File: m_inpak90.F90)
    • 20.15 I90_Len() -- Returns length of string. (Source File: m_inpak90.F90)
    • 20.16 I90_Lua() -- Returns available logical unit number. (Source File: m_inpak90.F90)
    • 20.17 I90_Pad() -- Pad strings. (Source File: m_inpak90.F90)
    • 20.18 I90_Trim() - Removes leading blanks from strings. (Source File: m_inpak90.F90)
    • 20.19 Lablin() -- Selects a Label (Inpak 77) (Source File: m_inpak90.F90)
    • 20.20 FltGetsp() -- Returns next float (Inpak 77, single precision) (Source File: m_inpak90.F90)
    • 20.21 FltGetdp() -- Returns next float (Inpak 77) (Source File: m_inpak90.F90)
    • 20.22 IntGet() -- Returns next integer (Inpak 77). (Source File: m_inpak90.F90)
    • 20.23 ChrGet() -- Returns next character (Inpak 77). (Source File: m_inpak90.F90)
    • 20.24 TokGet() -- Gets next token (Inpakk 77 like). (Source File: m_inpak90.F90)
  
  
• About this document ...