darshan 0.0.5

Python tools to interact with darshan log records of HPC applications.

Python utilities to interact with Darshan log records of HPC applications. pydarshan requires darshan-utils (3.2.2+) to be installed.

Features

• Darshan Report Object Wrapper

• CFFI bindings to access darshan log files

• Plots typically found in the darshan reports (matplotlib)

• Auto-discover darshan-util.so (via darshan-parser in $PATH)

Usage

For examples and Jupyter notebooks to get started with pydarshan make sure to check out the examples subdirectory.

A brief examples showing some of the basic functionality is the following:

import darshan

# Open darshan log
report = darshan.DarshanReport('example.darshan')

# Load some report data
report.mod_read_all_records('POSIX')
report.mod_read_all_records('MPI-IO')
# or fetch all
report.read_all_generic_records()

# ...
# Generate summaries for currently loaded data
# Note: aggregations are still experimental and have to be activated:
darshan.enable_experimental()
report.summarize()

Installation

To install in most cases the following will work:

pip install darshan

For alternative installation instructions and installation from source refer to <docs/install.rst>

Testing

Targets for various tests are included in the makefile. To run the normal test suite use:

make test

Or to test against different version of Python using Tox:

make test-all

Coverage tests can be performed using:

make coverage

Conformance to PEPs can be tested using flake8 via:

make lint

Documentation

Documentation for the python bindings is generated seperatedly from the darshan-utils C library in the interest of using Sphinx. After installing the developement requirements using pip install -r requirements_dev.txt the documentation can be build using make as follows:

pip install -r requirements_dev.txt
make docs