4.1.7 Number literals

The table below lists the postfix character(s) to add to a number literal for type cohesion. To use the new netcdf4 types nco must be compiled/linked to the netcdf4 library and the Output file must be hdf5.

     n1[$time]=1UL;   // n1 will now by type NC_UINT
     n2[$lon]=4b;     // n2 will be of type NC_BYTE
     n3[$lat]=5ull;   // n3 will be of type NC_UINT64
     n3@a1=6.0d;      // attribute will be type NC_DOUBLE
     n3@a2=-666L;     // attribute will be type NC_INT

A floating point number without a postfix will default to NC_DOUBLE. An integer without a postfix will default to type NC_INT. Thre is no postfix for characters. Use a quoted string.

     n4[$rlev]=.1      // n4 will be of type NC_DOUBLE
     n5[$lon_grd]=2.   // n5 will be of type NC_DOUBLE
     n6[$gds_crd]=2e3; // n6 will be of type NC_DOUBLE
     n6@a1=41;         // attribute will be type NC_INT
     n6@a2=-21;        // attribute will be type NC_INT
     n6@units="kelvin" // attribute will be type NC_CHAR

netCDF3/4 Types

b|B

NC_BYTE a signed 1 byte integer

none

NC_CHAR a ISO/ASCII character

s|S

NC_SHORT a signed 2 byte integer

l|L

NC_INT a signed 4 byte integer

f|F

NC_FLOAT a single precision floating point number

d|D

NC_DOUBLE a double precision floating point number

netCDF4 Types

ub|UB

NC_UBYTE a unsigned 1 byte int

us|US

NC_USHORT a unsigned 2-byte int

u|U|ul|UL

NC_UINT a unsigned 4-byte int

ll|LL

NC_INT64 a signed 8-byte int

ull|ULL

NC_UINT64 a unsigned 8-byte int