Matplotlib Output To Pdf For Corel Draw

Update: The fonts issue was actaully solved by using rc('pdf', fonttype=42) but unfortunately it is combined with another - whenever is used any type of marker I tried CorelDraw co

Solution 1:

It turned out there are two important issues which ruins import of Matplotlib generated PDF into CorelDraw.

1. Setting up the font type from the default rc("pdf", fonttype=3) to rc("pdf", fonttype=42)
2. Not using multiple markers. Only one marker per plot allowed! Possible to replace by text. (no pyplot.scatter, not in pyplot.plot etc.). When using any markers in number of 2 or more per plot CorelDraw finds the PDF corrupted and wont open it at all.

Code rewritten to plot only one marker per plot plus only one marker in the legend:

from matplotlib.backends.backend_pdf import PdfPages
import matplotlib.pyplot as plt
from matplotlib import rc
rc("pdf", fonttype=42)

with PdfPages("simple.pdf") as pdf:
    points_y = points_x = [1,2,3]
    plt.plot(points_x, points_y,color="r")
    # plot points one be one otherwise "File is corrupted" in CorelDraw# also plot first point out of loop to make appropriate legend
    plt.plot(points_x[0], points_y[0],marker="o",color="r",markerfacecolor="b",label="Legend label")
    for i inrange(1,len(points_x)):
        plt.plot(points_x[i], points_y[i],marker="o",markerfacecolor="b")
    plt.legend(numpoints=1) #Only 1 point in legend because in CorelDraw "File is corrupted" if default two or more 
    pdf.savefig()
    plt.close()

As a possible replacement for points (markers) pyplot.text can be used, for my example in question updated code looks like this:

from matplotlib.backends.backend_pdf import PdfPages
import matplotlib.pyplot as plt
from matplotlib import rc
rc("pdf", fonttype=42)

with PdfPages("simple.pdf") as pdf:
    points_y = points_x = [1,2,3]
    plt.plot(points_x, points_y)
    # print points as + symbolfor i inrange(len(points_x)):
        plt.text(points_x[i], points_y[i],"+",ha="center", va="center")
    pdf.savefig()
    plt.close()