dataplot.py

Plot numerical data found in human readable weakly formatted logfiles.

This program extracts numerical data from arbitrary text files, typically logfiles. It plots the data in a graph which is written to a file in PNG/JPG/PDF format (really anything supported by matplotlib).

Input lines are first optionally filtered with --filter. Each line forms a data row. Numeric values in each line are extracted using a regex (--num-regex). The X and Y values of each record are extracted from fixed columns specified by --xcol and --ycol, respectively. When --xcol is not specified the row index (the line number in the file after filtering) is used as X value.

Example: Plotting roundtrip times of ping:

sudo ping example.com -c 1000 -i 0.001 > log.txt  
dataplot.py -f time= -x 2 -y 4 -s . log.txt -o log.png  

Try adding --sort.

Help message

This help message may be outdated. Use dataplot.py --help to get an up-to-date help.

> dataplot.py --help

positional arguments:
  FILES               Files to process.

optional arguments:
  -h, --help          show this help message and exit
  --version           show program's version number and exit
  -o F, --outfile F   Output image. Default is 'out.png'. PNG, JPG, PDF and
                      others are supported.
  -x N, --xcol N      X column. Use -1 for 'index' (if no X column is present
                      in file).
  -y N, --ycol N      Y column. Use -vv to figure out column indices of data.
  -c C, --colors C    Set colors. One character per graph. Try rbyg.
  -s S, --shapes S    Set dot shapes (try oO.,+x).
  -a S, --addstyle S  Add additional style to all graphs (use -a - to add
                      lines).
  -f RE, --filter RE  Only use lines which match regex RE.
  --num-regex RE      Regex used to extrat numeric values in line.
  --xlog              Use logscale for X.
  --xdiv N            Divide X values by N (float).
  --ymax MAX          Set Y range to MAX (float).
  --ymin MIN          Set Y range to MIN (float).
  --ylog              Use logscale for Y.
  --sort              Sort Y values. Only makes sense without --xcol.
  --hist B            Build histogram over data with the specified binsize B.
                      Try --bar and --alpha 0.5.
  --legend POS        Set legend position (default "upper left").
  --bar               Draw filled bars.
  --alpha ALPHA       Set transparency. Useful for --bar with multiple plots.
  --fig-width W       Width of output image in inch at 100 dpi.
  --fig-height H      Height of output image in inch at 100 dpi.
  --print-high N      Print lines with Y values higher than N.
  --print-stats       Print statistics of all Y values.
  -v, --verbose       Be more verbose.

dataplot.py version 0.2.1 *** Copyright (c) 2010-2023 Johannes Overmann ***
https://github.com/jovermann/dataplot

Installation

dataplot.py depends on the following:

• Python 3
  • Version 0.2.2 tested with Python 2.7.18, 3.8.2, 3.9.10 and 3.11.1.
• matplotlib Python module

Usually you can install matplotlib by doing pip3 install matplotlib. On my Mac that failed, because it installed matplotlib for the system Python 3.8, and not for the Python 3.9 on my path which I installed using Mac ports. To install modules for the Python interpreter on your path use the Python interpreter to execute pip:

python3 -m pip install matplotlib

This worked well for me.