@@ -28,6 +28,10 @@ Currently in RICK we have not yet implemented the following steps:
<br>
## Preparing the dataset
To be digested by RICK, the input Measurement Set must be written in binary. To do this, a Python script named `create_binMS.py` can be found in the `/scripts`directory. The supported format for MS input is the Measurement Set V2.0 standard. Given that the script needs to read some columns of the input data, we reccommend to use a Singularity image to deal with the `casacore` dependencies. You can find one [here](https://lofar-webdav.grid.sara.nl/software/shub_mirror/tikk3r/lofar-grid-hpccloud/lofar_sksp_v3.5_x86-64_generic_noavx512_ddf.sif?action=show). The script has to be modified with the input Measurement Set.
## How to compile the code
The latest version of the code is in the *merge* branch. To compile it, you need first to specify in the Makefile the required configuration for the RICK execution.<br>
...
...
@@ -106,19 +110,23 @@ To use the **cuFFTMp** with **nvhpc 23.5** you have to add the following paths t
Once you have compiled the code, run it simply with the command:
```
> mpirun -np [n] [executable] data/paramfile.txt
```
When CPU hybrid MPI+OpenMP parallelism is requested, you have to select the number of threads by setting **--cpus-per-task=** in your bash script, then add the following lines:
```
> export OMP_NUM_THREADS=${SLURM_CPUS_PER_TASK}
> export OMP_PLACES=cores
> export OMP_PROC_BIND=close
```
Once you have compiled the code, run it simply with the command:
```
> mpirun -np [n] [executable] data/paramfile.txt
```
Then, to run the code in order to fully fill all the available cores in the node please add **--ntasks-per-socket=** in your bash script and then run: