pcm-graph

Tool that plots graphs from the CSV output of Intel PCM (Processor Counter Monitor)

Requirements

This script has been tested with Python 2.7, but should also work with Python 3.

The matplotlib library is required. Either install it with pip install matplotlib or use the requirements.txt file in this repository:

pip install -r requirements.txt

Usage

You can generate CSV output with the Intel PCM tool like so:

sudo ./pcm.x -csv=results.csv -- ./my_benchmark

This CSV file can then be plotted with this tool like this:

./pcm_graph.py -o output.png results.csv

The file extension of the output file will determine the file type. So if you want to generate a PDF, simply use something like output.pdf.

Further command line arguments can be inquired with ./pcm_graph.py --help:

usage: pcm_graph.py [-h] [-o OUTPUT] [-n NODES] [-p] [-q] [-s STYLE] [-t TITLE] input

positional arguments:
  input                 Path to the CSV file that contains the PCM results

optional arguments:
  -h, --help            show this help message and exit

  -o OUTPUT, --output OUTPUT
                        Path to output file. Defaults to %input%.png

  -n NODES, --nodes NODES
                        List of nodes to plot (e.g., 0,1,2,3)

  -p, --percentages     Use the percentage values for traffic instead of absolute values

  -q, --separate-qpi    Plot traffic for all QPI links separately

  -s STYLE, --style STYLE
                        Define a custom matplotlib style to use, see `matplotlib.style`

  -t TITLE, --title TITLE
                        Title of the figure

Styling

If you would like to customize the style of the output plot, I would recommand using the --style parameter with a matplotlib stylesheet.

For example, create a style sheet named custom.mplstyle with the following content:

figure.figsize : 10, 4
xtick.labelsize : 16
ytick.labelsize : 16

And this style sheet can then be invoked with --style:

./pcm_graph.py results.csv --style=custom.mplstyle