CODEWITHSUNDEEP
pythonmatplotlib
bluePhlavio
bluePhlavio

Reputation: 547

Define style classes in matplotlib

I have to plot different curves in one matplotlib figure. Each curve must have its own style (color, thickness, etc..) and I would like to set the styles in one matplotlibrc file. I would like to use a name like line1, line2, etc.. to refer to different styles, and not to a color cycle. Is this possible in motplotlib?

Upvotes: 2

Views: 265

Answers (2)

umitu
umitu

Reputation: 3340

cycler() can be used to set cycles for linewidth, linestyle, etc. as well as color. The length of the list of each argument must be the same. The list of available properties is defined here.

# series 1: color=red, linewidth=1, linestyle='-'
# series 2: color=blue, linewidth=2, linestyle='--'
axes.prop_cycle: cycler(color=['red', 'blue'], linewidth=[1, 2], linestyle=['-', '--'])

Upvotes: 0

ImportanceOfBeingErnest
ImportanceOfBeingErnest

Reputation: 339560

The matplotlib rc file is meant to provide the default style for a plot, its not meant to provide styles in the way of Cascading Style Sheets (CSS) or classes.

So what is possible, is to create several rc files and use them in a context, as explained in the temporary-styling part of the customizing tutorial, e.g. using a file called line1.mplstyle you could do

with plt.style.context(('line1')):
    plt.plot([1,2,3])

Because this seems to be a little overkill for just setting some line properties, it might be sufficient to simply create some argument dictionaries to provide to the plot command, like so:

line1 = dict(lw=2, ls=":", color="red")
line2 = dict(lw=0.8, ls="-", color="blue")

ax.plot([1,2,3], **line1)
ax.plot([1,2,3], **line2)

Upvotes: 1

Related Questions