raw2vdf - Transform a raw data volume into a VDC
raw2vdf [options] vdffile datafile
raw2vdf reads a 3D or 2D volume of data from disk, wavelet transforms the data, and stores it in the VDC associated with the file vdffile. The data volume must be stored on disk as a contiguous array of unformatted binary floating point values (32bit precision by default) with no header or trailer information. The dimensions of the raw data volume must match those specified in vdffile.
The -ts option specifies the integer time step of the data volume. The default value is 0, which corresponds to the first allowable time step in a VDC.
This option specifies name of the variable that the raw data volume corresponds too. The name must match either a 2D or 3D variable name in the VDF indicated by vdffile. If name matches a 2D variable, a 2D slice of data expected to be contained in datafile. If name matches a 3D variable, raw2vdf will attempt to read a 3D data volume from datafile.
This option can be used to limit the wavelet approximation hierarchy written to a Type 1 VDC. By default, level == -1, and all levels of the approximation hierarchy computed by the wavelet transform are stored in the VDC. If level is 0, only the coarsest approximation is saved. If level is 1, the coarsest approximation, and first refinement are saved. And so on. This option is not supported for VDC Type 2 and is silently ignored if present.
This option can be used to specify the number of execution threads to be employed when transforming data. The default value of n is 0, which causes the application to query the operating system to find the number of processors available, and then use that value. For most platforms the number of processors available is the result of the system call, sysconf(_SC_NPROCESSORS_ONLN).
The compression level which the data will be saved at. 0=> coaresst, 1 => next refinement ... (VDC2 only).
Operate quietly without reporting any informational messages.
As data are read from the file datafile their byte ordering is reversed, making "big endian" numbers "little endian" numbers, or vice-versa.
By default data values stored in the file datafile are assumed to be 32bit. If this option is present 64bit floating point values are read.
This option limits the reading of the raw data to a subvolume, whose minimum and maximum integer voxel bounds along the X dimension are given by min and max, respectively. The smallest allowable value of min is 0. The largest allowable value of max is NX-1, where NX is the dimension of the X axis.
See -xregion
See -xregion
This option informs raw2vdf that one of the raw data volume dimensions is "staggered": data samples are computed on cell faces, not cell centers. A staggered dimension will have N+1 sample points, where N is the axis dimension. A value of 1 indicates that the X dimension is staggered, a value of 2 indicates a Y staggered dimesion, and so on. By default, no dimensions are staggered.
Staggered dimensions are resampled to cell centers by raw2vdf prior to wavelet transforming the data.
The command
raw2vdf -ts 0 -varname vx foo.vdf vx.float
would transform the volume stored in the file vx.float and write it into the VDC associated with the foo.vdf VDF file. The time step and variable would be 0 and vx, respectively.
vdfcreate, vdf2raw
An Overview of VAPOR Data Collections
Last updated on $Date$