Modules

Sarcomp follows a consistent set of conventions for its data handling. In the context of Synthetic Aperture Radar (SAR) processing, the azimuth dimension (or the along-track direction) is referred to as sx, and is considered the first dimension. The range (or cross-track or radial line-of-sight) dimension is referred to as sy, and is considered the second dimension.

Sometimes, you may find nx and ny used in place of sx and sy. This notation represents the zero-padded power of 2 representation of sx and sy. For instance, if sx equals 2010, then nx would be its nearest power of 2, i.e., 2048.

The following modules are currently implemented in Sarcomp:

Modules:

• Memory
  • Complex Array Types
    • Interleaved-Complex Representation
    • Split-Complex Representation
      • sarcomp.empty
        • empty()
      • sarcomp.copy
        • copy()
      • sarcomp.zeros
        • zeros()
      • sarcomp.convert
        • convert()
• Utility Functions
  • Functions
    • cvt_bytes()
• Math Functions
  • Module API
• Fast Fourier Transform (FFT)
  • FFT Class
    • FFT
      • FFT.supported_sizes
      • FFT.FORWARD
      • FFT.BACKWARD
      • FFT.r2c
      • FFT.c2c
      • FFT.fft_plan
      • FFT.fft_exec
      • FFT.plan_isset
      • FFT.__init__()
      • FFT.__call__()
      • FFT.free()
      • FFT.free()
      • FFT.get_exec_function()
      • FFT.supported_sizes
  • Functions
    • fft()
    • ifft()
    • fft2()
    • ifft2()
  • Usage
    • 1D FFT Example
    • 2D FFT Example
• Multiprocessing with SARComp
  • Enabling Multiprocessing Support
  • Using SARComp with Multiple Threads
  • Important Note on Thread Management
  • Best Practices and Tips
  • Troubleshooting
  • Additional Resources

Multi-Core / Multi-Thread Processing

Sarcomp supports multi-core and multi-thread processing through OpenMP. You can set the number of threads used by Sarcomp by exporting an environment variable, OMP_NUM_THREADS. Here is an example of how to set this variable in a Unix-like system:

export OMP_NUM_THREADS=4

This command sets OMP_NUM_THREADS to 4, instructing OpenMP to use 4 threads. Adjust the number as needed for your system and workload.

Integration in Caramba

Sarcomp is currently integrated in the following Caramba workflows:

If you wish to use Sarcomp within Caramba, you need to set the CARAMBA_USE_SARCOMP environment variable to 1. You can do this by executing the following command in your terminal:

export CARAMBA_USE_SARCOMP=1