AMBER Archive (2009)

Subject: Re: [AMBER] bintraj and amber 10

From: Naser Alijabbari (na3m_at_virginia.edu)
Date: Thu Apr 23 2009 - 19:37:06 CDT

Hi Chris, just going from three digit of precision to 6 digits of precision
using Netcdf files makes my results different. Is there any way of
increasing “float coordinates(frame, atom, spatial)” to “double” or “long
double” just to see what the limit of this value does to my results. This is
what I was looking through so please bear with me:

“Trajectory files shall use 64 bit offsets

This can be accomplished by setting a flag during file creation; refer to
API docs for details."

So I google 64 bit offset. This gives
http://www.unidata.ucar.edu/software/netcdf/docs/netcdf/NetCDF-64-bit-Offset-Format-Limitations.html

“A 64-bit offset format netCDF file can have up to 2^32 - 1 fixed sized
variables, each under 4GiB in size. If there are no record variables in the
file the last fixed variable can be any size.”

So I look up record variable:

https://www.unidata.ucar.edu/software/netcdf-java/formats/RecordsInNetcdf3.html

“In the NetCDF version 3 data model, *record variables* are ones that use
the *unlimited dimension* (also known as the *record dimension*). These are
laid out on disk differently than non-record variables. All of the data for
a non-record variable is stored contiguously on disk. Record variables, in
contrast, are divided up into*records*, and each record contains the data
for all record variables for the ith record dimension index. You can append
an unlimited number of records to a NetCDF file.”

I am still not sure what “If there are no record variables in the file the
last fixed variable can be any size.” means.

 This is my float.h:

#define FLT_RADIX 2

   /* Number of base-FLT_RADIX digits in the significand of a float */

#define FLT_MANT_DIG 24

   /* Number of decimal digits of precision in a float */

#define FLT_DIG 6

   /* Addition rounds to 0: zero, 1: nearest, 2: +inf, 3: -inf, -1: unknown
*/

#define FLT_ROUNDS 1

   /* Difference between 1.0 and the minimum float greater than 1.0 */

#define FLT_EPSILON 1.19209290e-07F

   /* Minimum int x such that FLT_RADIX**(x-1) is a normalised float */

#define FLT_MIN_EXP (-125)

   /* Minimum normalised float */

#define FLT_MIN 1.17549435e-38F

   /* Minimum int x such that 10**x is a normalised float */

#define FLT_MIN_10_EXP (-37)

   /* Maximum int x such that FLT_RADIX**(x-1) is a representable float */

#define FLT_MAX_EXP 128

   /* Maximum float */

#define FLT_MAX 3.40282347e+38F

   /* Maximum int x such that 10**x is a representable float */

#define FLT_MAX_10_EXP 38

   /* Number of base-FLT_RADIX digits in the significand of a double */

#define DBL_MANT_DIG 53

   /* Number of decimal digits of precision in a double */

#define DBL_DIG 15

   /* Difference between 1.0 and the minimum double greater than 1.0 */

#define DBL_EPSILON 2.2204460492503131e-16

   /* Minimum int x such that FLT_RADIX**(x-1) is a normalised double */

#define DBL_MIN_EXP (-1021)

   /* Minimum normalised double */

#define DBL_MIN 2.2250738585072014e-308

   /* Minimum int x such that 10**x is a normalised double */

#define DBL_MIN_10_EXP (-307)

   /* Maximum int x such that FLT_RADIX**(x-1) is a representable double */

#define DBL_MAX_EXP 1024

   /* Maximum double */

#define DBL_MAX 1.7976931348623157e+308

   /* Maximum int x such that 10**x is a representable double */

#define DBL_MAX_10_EXP 308

   /* Number of base-FLT_RADIX digits in the significand of a long double */

#define LDBL_MANT_DIG 64

   /* Number of decimal digits of precision in a long double */

#define LDBL_DIG 18

   /* Difference between 1.0 and the minimum long double greater than 1.0 */

#define LDBL_EPSILON 1.08420217248550443401e-19L

   /* Minimum int x such that FLT_RADIX**(x-1) is a normalised long double
*/

#define LDBL_MIN_EXP (-16381)

   /* Minimum normalised long double */

#define LDBL_MIN 3.36210314311209350626e-4932L

   /* Minimum int x such that 10**x is a normalised long double */

#define LDBL_MIN_10_EXP (-4931)

   /* Maximum int x such that FLT_RADIX**(x-1) is a representable long
double */

#define LDBL_MAX_EXP 16384

   /* Maximum long double */

#define LDBL_MAX 1.18973149535723176502e+4932L

   /* Maximum int x such that 10**x is a representable long double */

#define LDBL_MAX_10_EXP 4932

Is “float coordinates(frame, atom, spatial)” is float because of the “64 bit
offset”? Is there a way of increasing this to something else? Thanks

On Thu, Apr 23, 2009 at 1:46 PM, Chris Moth <chris.moth_at_vanderbilt.edu>wrote:

> > Is there any way to see the precision of what is written in the netcdf
> > file?
>
> Google gives this
>
> http://ambermd.org/netcdf/nctraj.html
>
> And, the ncdump tool which comes with netcd gives some clues.
>
> mycomputer> ncdump -h 16301-16400.crd
> netcdf 16301-16400 {
> dimensions:
> frame = UNLIMITED ; // (63 currently)
> spatial = 3 ;
> atom = 35196 ;
> label = 5 ;
> cell_spatial = 3 ;
> cell_angular = 3 ;
> variables:
> char spatial(spatial) ;
> float time(frame) ;
> time:units = "picosecond" ;
> float coordinates(frame, atom, spatial) ;
> coordinates:units = "angstrom" ;
> char cell_spatial(cell_spatial) ;
> char cell_angular(cell_angular, label) ;
> double cell_lengths(frame, cell_spatial) ;
> cell_lengths:units = "angstrom" ;
> double cell_angles(frame, cell_angular) ;
> cell_angles:units = "degree" ;
>
> // global attributes:
> :title = "" ;
> :application = "AMBER" ;
> :program = "pmemd" ;
> :programVersion = "10.0" ;
> :Conventions = "AMBER" ;
> :ConventionVersion = "1.0" ;
>
>
>
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber