Revision history [back]

This is a bug, I think. Here is the documentation for Piecewise plotting.

def plot(self, *args, **kwds):
    """
    Returns the plot of self.

    Keyword arguments are passed onto the plot command for each piece
    of the function. E.g., the plot_points keyword affects each
    segment of the plot.

So that includes the legend ones, which however, is not what you want, as opposed to some of the other options. I've added this to the piecewise plotting bug ticket.

This is a bug, I think. Here is the documentation for Piecewise plotting.

def plot(self, *args, **kwds):
    """
    Returns the plot of self.

    Keyword arguments are passed onto the plot command for each piece
    of the function. E.g., the plot_points keyword affects each
    segment of the plot.

So that includes the legend ones, which however, is not what you want, as opposed to some of the other options. I've added this to the piecewise plotting bug ticket.

Update: Here is a workaround of sorts. Unfortunately, it means plotting each part separately, only labeling one of them.

sage: P = plot(Piecewise([[(0,1),x]]),legend_label='f(x)')
sage: Q = plot(Piecewise([[(1,2),x^2]]),legend_label='')
sage: R = plot(Piecewise([[(0,1),x^2]]),legend_label='g(x)',color='red')
sage: S = plot(Piecewise([[(1,2),x]]),legend_label='',color='red')
sage: P+Q+R+S

Or you could just skip the legend_label option totally on Q and S.