Plotting with Matplotlib

Several months ago, I built a pen plotter from an open-source design from thingiverse and modified it to hold a different servo and be wall-mounted.

The goal of this project was to take a simple 2d plot generated in python, and export it to a gcode format that is readable by the pen plotter device.

The workflow from start to finish is as follows:

  1. Create a plot figure using commonly used python libraries
  2. Export that plot to an svg file
  3. Generate a gcode file for my plotter using the vpype-gcode plugin

Creating a plot

First, a plot was created using matplotlib to handle the figures, and numpy for access to math and array operations. (Plot copied from here)

import matplotlib.pyplot as plt
import numpy as np

Fs = 8000
f = 5
sample = 8000
x = np.arange(sample)
y = np.sin(2 * np.pi * f * x / Fs)
plt.plot(x, y)
plt.xlabel('sample(n)')
plt.ylabel('voltage(V)')
plt.show()

Result:

png

Saving a Figure

Figures can be saved by calling plt.savefig("filename.svg") before running plt.show()

Fs = 8000
f = 5
sample = 8000
x = np.arange(sample)
y = np.sin(2 * np.pi * f * x / Fs)
plt.figure(figsize=(6,6),dpi=100)
plt.plot(x, y)
plt.xlabel('sample(n)')
plt.ylabel('voltage(V)')
plt.savefig("sine_wave.svg", pad_inches=0.1)
plt.show()

Result:

png

This saved the figure to sine_wave.svg in the same directory as this notebook. Using optional args in plt.figure and plt.savefig, the plotted image will have the following parameters:

  • 100 dpi (dots per inch) plt.figure(dpi=100)
  • 6.0 x 6.0 inches in size (W x H) plt.figure(figsize=(6,6))
  • 0.1 inch padding around the image plt.savefig("sine_wave.svg", pad_inches=0.1)

Generating Gcode

The following shell command takes the output svg file and writes it to a gcode file

vpype read sine_wave.svg gwrite sine_wave.gcode

Before running this command, I set up a default profile for my machine that will handle flipping the y axis and setting home as the current position on start.

Since this is a shell command, it must be called in this notebook using the subprocess module.

import subprocess

vpype_export = "vpype read sine_wave.svg gwrite sine_wave.gcode"
print(subprocess.Popen(vpype_export, shell=True, stdout=subprocess.PIPE).stdout.read())

Pen Plotter Results

Running the resulting gcode through the pen plotter yielded a clean image with suprisingly clear text.

pen=plot

The only issues I noticed were the doubling of lines along the plot box, and the outer rectangle that was a result of including the white backgound in the exported SVG. I think I can resolve this in future plots.

Conclusion and Future Work

All in all, I consider this a promising start to future plots. For my next steps I would like to try splitting the plot into multiple colors, and having the plotter prompt for pen changes within a single gcode file.