ToBeReviewed/LECTURE/
read_ncdf.pro
Reading function for the file net_cdf.
This program is less universal than ncdf_lec (it appeal to
declared variables in common) but it is very easier to be used.
It considerate
the declaration of the different zooms which have been defined
(ixminmesh...premierx...), the declaration of the variable key_shift...
To put it in a nutshell, the result of read_ncdf can be directly used in
plt ...
This is also this program which is used by default in our reading widgets.
read_ncdf
Reading
result = read_ncdf(name, beginning, ending[, compatibility], BOXZOOM=BOXZOOM, FILENAME=string, PARENTIN=PARENTIN, TIMESTEP=scalar: 0 or 1, ADDSCL_BEFORE=scalar: 0 or 1, TOUT=scalar: 0 or 1, NOSTRUCT=scalar: 0 or 1, CONT_NOFILL=CONT_NOFILL, INIT=scalar: 0 or 1, GRID=GRID, CALLITSELF=scalar: 0 or 1, DIREC=DIREC, ZETAFILENAME=string, ZETAZERO=scalar: 0 or 1, ZINVAR=named variable, _EXTRA=_EXTRA)
Return value
Structure readable by litchamp or an array if NOSTRUCT is
activated.
Parameters
name
in
required
type: string
It define the field to be read.
beginning
in
required
Relative with the time axis.
These can be
- 2 dates of the type yyyymmdd and in this case, we select data
which are included between these two dates.
- 2 indexes which define between which and which time steps we have
to extract the temporal dimension.
ending
in
required
Relative with the time axis.
See BEGINNING.
compatibility
in
optional
Useless, defined for compatibility
Keywords
BOXZOOM
Contain the boxzoom on which we have to do the reading
FILENAME
required
type: string
It contains the file's name.
PARENTIN
TIMESTEP
type: scalar: 0 or 1
default: 0
Specify that BEGINNING and ENDING refer to indexes of the time axis and not
to dates
ADDSCL_BEFORE
type: scalar: 0 or 1
default: 0
put 1 to apply add_offset and scale factor on data before looking for
missing values
TOUT
type: scalar: 0 or 1
default: 0
We activate it if we want to read the file on the whole domain without
considerate the sub-domain defined by the boxzoom or
lon1,lon2,lat1,lat2,vert1,vert2.
NOSTRUCT
type: scalar: 0 or 1
default: 0
We activate it if we do not want that read_ncdf send back a structure
but only the array referring to the field.
CONT_NOFILL
INIT
type: scalar: 0 or 1
default: 0
To call automatically initncdf with filename as input argument
and thus redefine all the grid parameters
GRID
='[UTVWF]' to specify the type of grid. Default is (1)
based on the name of the file if the file ends by
GRID[._][TUVFW].NC (not case sensible) or (2) T if case (1)
is not found.
CALLITSELF
type: scalar: 0 or 1
default: 0
For ROMS outputs. Use by read_ncdf itself to access auxilliary data
(h and zeta).
DIREC
a string used to specify the direction along which we want to make
spatial and/or temporal mean. It could be: 'x' 'y' 'z' 't' 'xy' 'xz'
'yz' 'xyz' 'xt' 'yt' 'zt' 'xyt' 'xzt' 'yzt' or 'xyzt'
ZETAFILENAME
type: string
default: FILENAME
For ROMS outputs. The filename of the file where zeta variable should be read
ZETAZERO
type: scalar: 0 or 1
default: 0
For ROMS outputs. To define zeta to 0. instead of reading it
ZINVAR
type: named variable
Set this keyword to a named variable in which 1 is returned if a
vertical dimension is found in the variable. Returns 0 otherwise
_EXTRA
Used to pass keywords to isafile, initncdf,
ncdf_gettime, ncdf_getatt and domdef
Version history
Version
$Id: read_ncdf.pro 399 2009-07-02 08:38:54Z smasson $
History
Sebastien Masson (smasson@lodyc.jussieu.fr)
15/10/1999
Known issues
Restrictions
The field must have a temporal dimension.
Other attributes
Uses routines
common
Produced by IDLdoc 2.0.