netcdf4-js

NodeJS addon for reading and writing the files in the
Network Common Data Form (NetCDF)
version <= 4,
built upon the C-library for netcdf.

Installation

netcdf4-js is built with nodejs >= 4.x

Install using npm:

``
$ npm install netcdf4
`

Prerequisites:

You will need libnetcdf >= 4.x installed.

$3

- Install NetCDF4 using your package manager, e.g., on Ubuntu/Debian:

`
$ sudo apt-get install libnetcdf-dev
`

or download it here

- Make sure your system fulfills all the prerequisites of node-gyp

$3

- Install NetCDF4 from here
- Make sure to select at least "dependencies", "headers" and "libraries" to install in the NetCDF installation wizard
- Install the build tools as described here
- Set the environment variable NETCDF_DIR to your NetCDF installation, e.g.,

`bash
C:\> set NETCDF_DIR=C:\Program Files\netCDF 4.6.1
`

Usage

Open files with

`
var netcdf4 = require("netcdf4");

var file = new netcdf4.File("test/testrh.nc", "r");
`

File modes are "r" for "reading", "w" for "writing", "c" for
"creation", and "c!" for "overwriting".

Then you can read variables using read or readSlice. The following example reads values at positions 5 to 15:

`
console.log(file.root.variables['var1'].readSlice(5, 10));
`

$3

Properties marked _(r/w)_ can be read and will be written to the file
when set.

#### File

Properties:

- root : Main Group-object in file

Methods:

- close() : Close file
- sync() : Sync (or "flush") file to disk

#### Group

Properties:

- id : ID used by C-library
- name : Name
- fullname : Full name (path in file)
- variables : Associative array of variables in group
- dimensions : Associative array of dimensions in group
- unlimited : Associative array of unlimited Dimensions in group
- attribute : Associative array of attributes of group
- subgroups : Associative array of subgroups of group

Methods:

- addVariable(name, type, dimensions) : Add a new variable in
group. type is one of "byte", "char", "short", "int", "ubyte", "ushort", "uint", "float", "double". dimensions is an array of ids of dimensions
for the new variable. Returns new variable.
- addDimension(name, length) : Add new dimension of length length
(can be "unlimited" for unlimited dimension). Returns new dimension.
- addSubgroup(name) : Add subgroup. Returns new subgroup.
- addAttribute(name, type, value) : Add and set new attribute. Returns new attribute.

$3

Properties:

- id : ID used by C-library
- name : Name (r/w)
- length : Length or currently used length for unlimited dimensions

$3

Properties:

- id : ID used by C-library
- name : Name (r/w)
- value : Value (r/w)

Methods:

- delete() : Delete attribute

$3

Properties:

- id : ID used by C-library
- name : Name (r/w)
- type : Type of variable
- attributes : Associative array of attributes
- dimensions : Array of dimensions used by variable
- endianness : Endianness: "little", "big", or "native" (r/w)
- checksummode : Checksum mode: "none", or "fletcher32" (r/w)
- chunkmode : Chunk mode: "contiguous", or "chunked" (r/w)
- chunksizes : Array of chunk sizes (one size per dimension) (r/w)
- fillmode : Boolean switch for fill mode (r/w)
- fillvalue : Fill value (r/w)
- compressionshuffle : Boolean switch for shuffle (r/w)
- compressiondeflate : Boolean switch for compression (r/w)
- compressionlevel : Compression level (1-9) (r/w)

Methods:

- read(pos....) : Reads and returns a single value at positions
given as for write.
- readSlice(pos, size....) : Reads and returns an array of values (cf.
"Specify a Hyperslab")
at positions and sizes given for each dimension, readSlice(pos1, size1, pos2, size2, ...) e.g. readSlice(2, 3, 4, 2) gives an
array of the values at position 2 for 3 steps along the first
dimension and position 4 for 2 steps along the second one.
- readStridedSlice(pos, size, stride....) : Similar to readSlice(), but it
adds a _stride_ (interval between indices) parameter to each dimension. If stride is 4,
the function will take 1 value, discard 3, take 1 again, etc.
So for instance readStridedSlice(2, 3, 2, 4, 2, 1) gives an
array of the values at position 2 for 3 steps with stride 2 (i.e.
every other value) along the first dimension and position 4 for 2 steps
with stride 1 (i.e. with no dropping) along the second dimension.
- write(pos..., value) : Write value at positions given,
e.g. write(2, 3, "a") writes "a" at position 2 along the first
dimension and position 3 along the second one.
- writeSlice(pos, size..., valuearray) : Write values in valuearray
(must be a typed array) at positions and sizes given for each
dimension, e.g. writeSlice(2, 3, 4, 2, new Int32Array([0, 1, 2, 3, 4, 5])) writes the array at position 2 for
3 steps along the first dimension and position 4 for 2 step along
the second one (cf.
"Specify a Hyperslab").
- writeStridedSlice(pos, size, stride..., valuearray) : Similar to
writeSlice(), but it adds a _stride_ parameter to each dimension.
So for instance writeStridedSlice(2, 3, 2, 4, 2, 1), new Int32Array([0, 1, 2, 3, 4, 5]) writes the array
at position 2 for 3 steps with stride 2 (i.e.
every other value) along the first dimension and position 4 for 2 steps
with stride 1 (i.e. with no dropping) along the second dimension.
- addAttribute(name, type, value)` : Adds and sets new attribute. Returns
new attribute.