current version: 1.0 - 4th July 2008
| version | date | comment |
|---|---|---|
| 1.0 | 4/Jul/2008 | Original code |
license: GNU GPL http://www.gnu.org/licenses/
This file is part of
MOSAICO -- MOdular library for raSter bAsed hydrologIcal appliCatiOn.
Copyright (C) 2011 Giovanni Ravazzani
Language: Fortran 90.
Software Standards: "European Standards for Writing and
Documenting Exchangeable Fortran 90 Code".
Motivation:
Portability refers to the ease with which source code can be moved from machine to machine. A portable program requires little or no change to the source code when compiled and run on a different machine. Portability of numerical code is important for several reasons. Fortran 90 introduces several new mechanisms to aid in porting numerical code to other machines. The most difficult problem in porting numerical programs is in the portable selection of precision. Selecting the precision of a computation in a portable way was impossible with Fortran 77. However, with the introduction of kind values as well as intrinsic environmental inquiry functions for selecting and inquiring about precision, Fortran 90 programs should be much easier to port to other machines. Kind values are integer constants that can be used to further specify the characteristics of an intrinsic type, such as integer or real. For example, real( 2 ) selects a real type with kind value 2. Unfortunately, kind values are processor-dependent and are therefore not standardized. However, there are several portable ways to select kind values based on the precision desired. Two such functions are selected_int_kind and selected_real_kind
Description:
Provide named kind parameters for use in declarations of real and integer variables with specific byte sizes (i.e. one, two, four, and eight byte integers; four and eight byte reals). The parameters can then be used in (KIND = XX) modifiers in declarations. A single function (SizesOK()) is provided to ensure that the selected kind parameters are correct.
References and Credits:
Adapted from netCDF library http://www.unidata.ucar.edu/software/netcdf/ Written by Robert Pincus
Design Notes:
Fortran 90 doesn't allow one to check the number of bytes in a real variable; we check only that four byte and eight byte reals have different kind parameters.
| Type | Visibility | Attributes | Name | Initial | |||
|---|---|---|---|---|---|---|---|
| integer, | public, | parameter | :: | double | = | SELECTED_REAL_KIND(P=13, R=307) |
64 bits real type floating point numbers with numerical precision at least 13 digits and one decimal exponent range between -307 and +307 equivalent to EightByteReal type in netCDF library (double precision) |
| integer, | public, | parameter | :: | float | = | SELECTED_REAL_KIND(P=6, R=37) |
32 bits real type floating point numbers with numerical precision at least 6 digits and one decimal exponent range between -37 and +37 equivalent to FourByteReal type in netCDF library |
| integer, | public, | parameter | :: | long | = | SELECTED_INT_KIND(18) |
64 bits integer type equivalent to EightByteInt type in netCDF library |
| integer, | public, | parameter | :: | short | = | SELECTED_INT_KIND(9) |
32 bits integer type equivalent to FourByteInt type in netCDF library |
Users may call this function once to ensure that the kind parameters the module defines are available with the current compiler. We can't ensure that the two REAL kinds are actually four and eight bytes long, but we can ensure that they are distinct. Early Fortran 90 compilers would sometimes report incorrect results for the bit_size intrinsic, but I haven't seen this in a long time.