use jupyter widgets to save clicks on a pandas dataframe

Question

Basically what I'm trying to do is make a pandas dataframe clickable and saving the clicks the user made.

Thought about using widgets for that.

so far I have this:

import pandas as pd
from IPython.display import display, HTML
from ipywidgets import widgets

df = pd.DataFrame([[1,'car'],[2,'bus'],[3,'train']])

click_list = []

display(df[:1]) 
button = widgets.Button(description='click')
display(button)
def obc(b):
    click_list.append((pd.to_datetime('now'),1)) 
button.on_click(obc)

button2 = widgets.Button(description='click')
display(button2)
def obc2(b):
    click_list.append((pd.to_datetime('now'),2))
display(df[1:2])  
button2.on_click(obc2)

button3 = widgets.Button(description='click')
display(button3)
def obc3(b):
    click_list.append((pd.to_datetime('now'),3))
display(df[2:3])  
button3.on_click(obc3)

which give out this:

it has the the basic functionality I'm after, but it is displayed in a confusing way (especially if we are talking about tens of rows) I would like to have something like this:

or similar, any help will be much appriciated

Answer 1

It seems VBoxing html displays does a nice job (notice that IPython.display.HTML and ipywidgets.HTML are not the same)

import pandas as pd
from IPython.display import display, HTML
from ipywidgets import Button, HBox, VBox,widgets
import ipywidgets


df = pd.DataFrame([[1,'car'],[2,'bus'],[3,'train']])

click_list = []

button = widgets.Button(description='click')
def obc(b):
    click_list.append((pd.to_datetime('now'),1)) 
button.on_click(obc)

button2 = widgets.Button(description='click')
def obc2(b):
    click_list.append((pd.to_datetime('now'),2))
button2.on_click(obc2)

button3 = widgets.Button(description='click')
def obc3(b):
    click_list.append((pd.to_datetime('now'),3)) 
button3.on_click(obc3)

display(HBox([VBox([widgets.Button(description=''),button,button2,button3]),ipywidgets.
                    HTML(df.style.set_table_attributes('class="table"').render())]))

Answer 2

Is there anyway to automate this? So it can be used with anunknown amount of checkboxes/rows? The code belows gives an errror: AttributeError: 'list' object has no attribute '_handle_displayed'

import pandas as pd
from IPython.display import display, HTML
from ipywidgets import Checkbox, HBox, VBox,widgets
import ipywidgets

df = pd.DataFrame(data=[['a',1],['b',32]], columns=['J1','J2'])

mydict = {}
t=0
for ts in df.J1:
    mydict[str('c')+ str(t)] =  widgets.Checkbox(value=False, description = 'Accepted')
    t=t+1

    display(HBox([VBox([widgets.Checkbox(description=''),mydict.values()]),ipywidgets.
                    HTML(df.style.set_table_attributes('class="table"').render())]))