In Situ Python Analysis - hyschive/gamer-fork GitHub Wiki

What is in situ analysis?

In situ analysis is to analyze ongoing simulation data without outputting data to disk before conducting analysis. It enables us to explore these ongoing simulation data through accessing memory directly using Python.

Dependencies

• Install libyt and yt_libyt: the in situ analysis library and its yt frontend.
• Install yt: the core analysis tool.

Compilation Options

libyt has the following compilation options:

• --libyt
• --libyt_patchgroup
• --libyt_interactive
• --libyt_reload
• --libyt_jupyter

Must set LIBYT_PATH in the machine configuration file to the path containing the folders include and lib of libyt.

Runtime Parameters

libyt has the following runtime parameters:

• YT_SCRIPT: The Python script name to be loaded. Do not include the file extension .py.
• YT_VERBOSE: Log level of libyt (0=off, 1=info, 2=warning, 3=debug).
• YT_FIG_BASENAME: Figure basename of the outputs from yt.

For example, Input__Parameter with the following lines will read the Python script inline_script.py, set log level to information only, and set the figure basename to MHD.

YT_SCRIPT         inline_script
YT_VERBOSE        1
YT_FIG_BASENAME   MHD

Quick Demos

1. Plummer

FAQs

If you encounter the following OpenMPI error messages related to MPI remote memory access operation,

ompi_osc_ucx_win_attach: Assertion ......

try adding OMPI_MCA_osc=sm,pt2pt before mpirun:

OMPI_MCA_osc=sm,pt2pt mpirun ...

This is something libyt will fix in the near future.