pplot

Tool for quickly plotting things from commandline to matplotlib.

Quick start

$pplot.py data.txt 0 1 2

Plots columns 0 1 2 from file data.txt, assumes the file contains data separated by white spaces and contains atleast 3 columns.

$pplot.py data*.txt -c 0 1 2

Plots columns 0 1 2 from all files matched by data*.txt.

One file can be piped in through stdin (use -p to select which columns to plot from piped data):

$ cat data.txt | pplot.py -p 1 2 3

Possible to plot error bars -e, histograms -hist, 3d surfaces -surf, subplots -subf, etc.. See help for all possible commands:

$pplot.py -h

Plot Expressions:

Plot expressions from the read in data. The ith column of nth file is denoted as f[n]_[i] in the expressions. All numpy array functions and constants should be usable. More than one expression can be given at the same time.

$pplot.py file0.txt file1.txt -expr "( f0_0 - mean(f0_0) )/pi"  "sin(f1_1) - cos(f0_1)"

If expressions contain different file numbers file columns are assumed to have the same length.

Using in scripts

Can be used in scripts to edit a figure further. See pp_script_example.py.

Custom data reader:

It is possible to write a custom data reader method. The method should be put into data_readers/ folder with the following format: The file containing the code for the reader [reader_name].py should contain a function with the same name as the file and with the signature def [reader_name](pp: pplot, fname: str). The function is expected to return a tuple ( data: numpy.ndarray , labels: list\[str\] ). The data is expected to be 2d array with shape (num_rows, num_cols). Labels can be an empty list [], if not it is expected to be a list of strings with length num_cols containing the labels. You can do what ever in the method even read a custom binary format.

Then to use a specific reader use the command -rf [reader_name].

See data_readers/data_reader_example.py.

$pplot.py data.txt 1 -x0 -rf data_reader_example

Dependencies:

Python3:

• numpy (tested on v1.22.3)
• sympy (tested on v1.9)
• matplotlib (tested on v3.5.1)
• pandas (tested on v2.3.1)

Optional:

GNU utils:

• grep
• wc
• sed

Should be installed and in your $PATH. For now, used in default data reader for finding tag_data_start tag_data_end and tag_header tags. Can be disabled by setting defaults as tag_data_start=None, tag_data_end=None, tag_header=None.

Warning:

The script does not do much of error handling. For example, if you pass a column that does not exist the script will just crash and print python stack trace.

There are bound to be bugs.