factory for adding zoom callback to matplotlib graphs

simp_zoom.py

import matplotlib.pyplot as plt


def zoom_factory(ax,base_scale = 2.):
    def zoom_fun(event):
        # get the current x and y limits
        cur_xlim = ax.get_xlim()
        cur_ylim = ax.get_ylim()
        # set the range
        cur_xrange = (cur_xlim[1] - cur_xlim[0])*.5
        cur_yrange = (cur_ylim[1] - cur_ylim[0])*.5
        xdata = event.xdata # get event x location
        ydata = event.ydata # get event y location
        if event.button == 'up':
            # deal with zoom in
            scale_factor = 1/base_scale
        elif event.button == 'down':
            # deal with zoom out
            scale_factor = base_scale
        else:
            # deal with something that should never happen
            scale_factor = 1
            print event.button
        # set new limits
        ax.set_xlim([xdata - cur_xrange*scale_factor,
                     xdata + cur_xrange*scale_factor])
        ax.set_ylim([ydata - cur_yrange*scale_factor,
                     ydata + cur_yrange*scale_factor])
        ax.figure.canvas.draw() # force re-draw

    fig = ax.get_figure() # get the figure of interest
    # attach the call back
    fig.canvas.mpl_connect('scroll_event',zoom_fun)

    #return the function
    return zoom_fun

Very nice!

One modification I made for my PyQt4 GUI: instead of using ax as a parameter, I pass in the parent canvas and say ax = canvas.ax on the first line. Then, instead of plt.draw(), I use canvas.draw(). However, your version works better for subplot zooming.

@jdreaver I didn't see your comment till recently. I am glad it was helpful! If you think your changes would be helpful for other people you should fork the gist.

I also tweaked it a bit to not call plt.draw().

IMHO, I think we don't have to calculate the cur_xrange and cur_yrange. Instead we can calculate the relative distance from the cursor position to the frame like this:

        # Get distance from the cursor to the edge of the figure frame
        x_left = xdata - cur_xlim[0]
        x_right = cur_xlim[1] - xdata
        y_top = ydata - cur_ylim[0]
        y_bottom = cur_ylim[1] - ydata

then set the new xlim, ylim like this:

ax.set_xlim([xdata - x_left*scale_factor,
                    xdata + x_right*scale_factor])
ax.set_ylim([ydata - y_top*scale_factor,
                    ydata + y_bottom*scale_factor])

By this way, when zooming in and zooming out, the pixel which the cursor is pointing to won't be changed.

@giadang good point! That fixes one of the more annoying behaviors of this.

Thank you for this gist. Zooming in and out with the mousewheel (without moving the mouse pointer) does not restore the original view. It seems that the underlying view is paned.