Matplotlib: get colors and x/y data from a bar plot

Question

I have a bar plot and I want to get its colors and x/y values. Here is a sample code:

import matplotlib.pyplot as plt
def main():
    x_values = [1,2,3,4,5]
    y_values_1 = [1,2,3,4,5]
    y_values_2 = [2,4,6,8,10]
    f, ax = plt.subplots(1,1)
    ax.bar(x_values,y_values_2,color='r')
    ax.bar(x_values,y_values_1,color='b')
    #Any methods?
    plt.show()
if __name__ == '__main__':
    main()

Are there any methods like ax.get_xvalues(), ax.get_yvalues(), ax.get_colors(), which I can use so I could extract back from ax the lists x_values, y_values_1, y_values_2 and the colors 'r' and 'b'?

Answer 1

The ax knows what geometric objects it's drawing, but nothing about it keeps track of when those geometric objects were added, and of course it doesn't know what they "mean": which patch comes from which bar-plot, etc. The coder needs to keep track of that to re-extract the right parts for further use. The way to do this is common to many Python programs: the call to barplot returns a BarContainer, which you can name at the time and use later:

import matplotlib.pyplot as plt
def main():
    x_values = [1,2,3,4,5]
    y_values_1 = [1,2,3,4,5]
    y_values_2 = [2,4,6,8,10]
    f, ax = plt.subplots(1,1)
    rbar = ax.bar(x_values,y_values_2,color='r')
    bbar = ax.bar(x_values,y_values_1,color='b')
    return rbar, bbar

if __name__ == '__main__':
    rbar, bbar = main()
    # do stuff with the barplot data:
    assert(rbar.patches[0].get_facecolor()==(1.0,0.,0.,1.)) 
    assert(rbar.patches[0].get_height()==2)