Binary data organized as a sequence of numbers

Many binary formats contain internal "meta-data," that is the files will contain information about the structure and data type of the dataset the files contain. Other binary files may be "raw" dumps of an array, with little or no "meta-data." The binary sequence reader can read binary files that contain simple sequences of numbers. This type of data often results from a utility program that writes out the contents of an array as binary numbers. You will have to supply information about the dimensions and the data type of the data.

The following is an example of a program that creates a small binary file. IT writes out a 4 by 3 2D array of binary numbers.

#include <stdio.h>

#define X_SIZE 4
#define Y_SIZE 3

int main( int argc, char *argv[] )
{
int arr[Y_SIZE][X_SIZE];
int i,j;

for( j = 0; j < Y_SIZE; ++j )
for( i = 0; i < X_SIZE; ++i )
arr[j][i] = i + j;

for( j = 0; j < Y_SIZE; ++j )
for( i = 0; i < X_SIZE; ++i )
fwrite( &(arr[j][i]), 4, 1, stdout );

return 0;
}

When you use the wizard to select this reader, the sequence reader will be created and its user interface panel will appear in the Module Editor section of the application window. The reader's user interface asks you several question about your file in order to choose the correct reader and settings.

The skip lines asks how many bytes to skip at the beginning. Sometimes files will have information at the beginning of the file which is not data. This asks how many bytes to skip before it reaches the actual data.

The second box asks you how many dimensions you have in your file. In the case of our example we read it in as a 4 by 3 2 dimensional array.

Choose a byte order option that is appropriate for the machine that wrote the data file Windows PCs, Linux PCs and Compaq/HP Alphas are little-endian machines. Suns, SGIs, IBM workstations running AIX and HP workstation running HP-UX are big-endian machines. If the type of machine that wrote the data file is the same as the machine you are using to run AVS/Express, then you can choose "same as machine".

This reader can actually read multiple arrays from the data file. Although many uses of this reader will involve data files with only one array. The user interface used the term "variable." If you are familiar with AVS/Express' terminology for fields, a "variable" corresponds to a node data component. The user interface allows you to specify information for each node data component, such as vector length, data type, (optionally) label and units. For binary data, it is crucial that you supply the correct data type. If you supply the wrong data type, at a minimum your data will be garbled. It is also possible that AVS/Express will malfunction and need to be restarted.

Binary Data Organized as a Sequence of Numbers