Get coordinate values and keyboard input values by clicking on the python / matplotlib diagram

There are times when you want to display data with matplotlib on python or jupyter and analyze it manually with a mouse, range specification, value acquisition, etc. Also, I'm sorry for the appropriate sentence, but it is a memo at that time. I use it to handle astronomical data and music code, but it is useful when I want to create a data analysis tool with a suitable GUI like that. This is the reference page. http://matplotlib.org/users/event_handling.html

In the case of jupyter, it is not possible to do% matplotlib inline, so please use it as it is or% matplotlib notebook.

Preparation

import numpy as np
import pylab
import matplotlib.pyplot as plt

Mouse click and get data

def onclick(event):
    print 'event.button=%d,  event.x=%d, event.y=%d, event.xdata=%f, \
    event.ydata=%f'%(event.button, event.x, event.y, event.xdata, event.ydata)

Let's display the appropriate x, y array data

fig=plt.figure()
ax=fig.add_subplot(111)
ax.plot(x,y)
cid = fig.canvas.mpl_connect('button_press_event', onclick)
plt.show()

Click the figure with the mouse

event.button=1,  event.x=234, event.y=229, event.xdata=496.774194, event.ydata=1.011211
event.button=1,  event.x=257, event.y=182, event.xdata=570.967742, event.ydata=1.005703
event.button=1,  event.x=193, event.y=135, event.xdata=364.516129, event.ydata=1.000195
event.button=3,  event.x=193, event.y=135, event.xdata=364.516129, event.ydata=1.000195

It will display the button number and position you clicked like this. If you left-click, event.button = 1, 2 in the middle, and 3 on the right.

This event.xdata is the value of x, so let's find the corresponding index in the array and plot it with a red dot so that the index appears in the title. For this, you can use np.searchsorted, which will give you the index to the right of that value.

def oncpaint(event):
    ind=np.searchsorted(x,event.xdata)
    plt.title("You clicked index="+str(ind))
    ax.plot([x[ind]],[y[ind]],".",color="red")
    fig.canvas.draw()

Is there a clicked index in the title? Basically, you can get information interactively like this. It looks like Lego. Now are you using a mac now? Doesn't the mac have only one button and give too much priority to fashion? so. I want to do various operations such as enlargement by left-clicking (event.button = 1), and I want to do the operations I defined with other buttons. So you can use a cheap mouse, and then connect the one with 3 buttons.

Next, let's make something like this.

Prepare all true masks
Right-click and select a point twice to mark between the two points
Left double-click to confirm, middle button to cancel
The mask of the determined mark part becomes false

All true masks

mask=np.ones(len(x),dtype=bool)

Definition part of the above operation

def oncmask(event):
    global stat
    global leftind, rightind
    ind=np.searchsorted(x,event.xdata)
    plt.title("You clicked index="+str(ind))
    if event.button==3 and stat==1:
        leftind=ind
        ax.plot([x[ind]],[y[ind]],".",color="red")
        stat=2
    elif event.button==3 and stat==2:
        rightind=ind
        ax.plot(x[leftind:rightind],y[leftind:rightind],color="red")
        stat=3
        print leftind, rightind
    elif event.button==1 and event.dblclick==1 and stat==3:
        plt.title("Approved")
        mask[leftind:rightind]=False
        stat=1
    elif event.button==2 and stat==3:  
        plt.title("Canceled")
        ax.plot(x[leftind:rightind],y[leftind:rightind],color="blue")
        ax.plot([x[leftind]],[y[leftind]],".",color="green")
        stat=1
    fig.canvas.draw()

Right-click twice to select an area, then double-click left to Approve! Press the middle button to start over. Like this

Let's check if it was masked

fig=plt.figure()
ax=fig.add_subplot(111)
plt.xlim(200,205)
ax.plot(x[mask],y[mask])
plt.show()

It has been done.

Get keyboard value

You can also get keyboard input values. Let's erase it with flush this time.

import sys

Definition part. I set it to end when I press q.

def onkey(event):
    print('you pressed', event.key, event.xdata, event.ydata)
    if event.key == 'q':
        plt.close(event.canvas.figure)
    sys.stdout.flush()

Take out the figure and try to use the keyboard on the figure.

fig=plt.figure()
ax=fig.add_subplot(111)
plt.xlim(200,205)
plt.title("q:quit")
ax.plot(x,y)
cid = fig.canvas.mpl_connect('key_press_event', onkey)
plt.show()

It looks like this.

('you pressed', u'a', 201.38104838709677, 1.0182421874999998)
('you pressed', u'b', 201.38104838709677, 1.0182421874999998)
('you pressed', u'c', 201.38104838709677, 1.0182421874999998)
('you pressed', u'shift', 201.33064516129031, 1.0173046875)
('you pressed', u'A', 201.33064516129031, 1.0173046875)
('you pressed', u'shift', 201.32056451612902, 1.0168359375)
('you pressed', u'R', 201.32056451612902, 1.0164843749999999)
('you pressed', u'q', 201.21975806451613, 1.0033593750000001)

See you soon.

Postscript:

`func.py`


def oncpaint(event):
    ind=np.searchsorted(x,event.xdata)
    ax.plot([x[ind]],[y[ind]],".",color="red")
    fig.canvas.draw()

`main.py`


import builtins

builtins.x=np.linspace(0,1,100)
builtins.y=np.cos(x)
builtins.fig=plt.figure()
ax=fig.add_subplot(111)
ax.plot(x,y)
cid = fig.canvas.mpl_connect('button_press_event', onclick)
plt.show()

If you want to write this in a separate file and call it with a module, you may want to make x or fig a built-in scope with builtins at the call destination (please tell me if there is a better way).