Minimal example attempting to save dataframe from Bokeh selection

save_dataframe.py

# Minimal example attempting to save dataframe from Bokeh selection (Sep 21, 2018)
# https://gist.github.com/jni/73a2a8b6961df6a4b1a63b1e217e66d8
# save_dataframe.py

import pandas as pd
import numpy as np

from bokeh.server.server import Server
from bokeh.application import Application
from bokeh.application.handlers.function import FunctionHandler
from bokeh.plotting import figure
from bokeh.layouts import widgetbox, layout
from bokeh.models import ColumnDataSource
from bokeh.models.widgets import Button


def scatterplot(source):
    tools = ['pan', 'box_select', 'wheel_zoom', 'reset']
    scatter = figure(title='scatterplot', tools=tools,
                     active_drag='box_select')
    scatter.circle(source=source, x='x', y='y')
    return scatter


def save_selected(dataframe, selection):
    data = dataframe.iloc[selection]
    def save():
        print(f'data.shape: {data.shape}')
        # data.to_csv(r'data.csv', index = False)
        data.to_csv(r'data.csv', index=False, encoding='utf-8')
    return save


def clear_click_callbacks(button):
    button._callbacks['clicks'].clear()     # error


def generate_makedoc_function(dataframe):
    def makedoc(doc):
        source = ColumnDataSource(dataframe)
        scatter = scatterplot(source)
        savebutton = Button(label='download selected points')
        savebutton.on_click(save_selected(dataframe, selection=[]))

        def update_selection(attr, old, new):
            # clear_click_callbacks(savebutton)  # on_click appends, want replace
            savebutton.on_click(save_selected(dataframe,
                                              selection=new))


        source.selected.on_change('indices', update_selection)
        page_content = layout([[scatter], [savebutton]])
        doc.title = 'save selected points!'
        doc.add_root(page_content)
    return makedoc


def main(path='/', port=5006):
    df = pd.DataFrame({'x': np.random.random(100),
                       'y': np.random.random(100),
                       'useful': np.random.randint(0, 100, size=100)})
    apps = {path: Application(FunctionHandler(generate_makedoc_function(df)))}
    server = Server(apps, port=port)
    server.run_until_shutdown()


if __name__ == '__main__':
    main()

save_dataframe_notebook.py

#!/usr/bin/env python
# coding: utf-8

# Minimal example attempting to save dataframe from Bokeh selection
# 
# https://gist.github.com/planetscape/a4d587f6668117321f2c0b24264eb4f6

# In[ ]:


import pandas as pd
import numpy as np

from bokeh.server.server import Server
from bokeh.application import Application
from bokeh.application.handlers.function import FunctionHandler
from bokeh.plotting import figure
from bokeh.layouts import column, widgetbox, layout
from bokeh.models import ColumnDataSource, Slider
from bokeh.models.widgets import Button, DataTable, TableColumn
from bokeh.themes import Theme
from bokeh.io import show, output_notebook

output_notebook()


# In[ ]:


selected_points = []

def scatterplot(source):
    tools = ['pan', 'box_select', 'wheel_zoom', 'reset']
    scatter = figure(title='scatterplot',
                     tools=tools,
                     active_drag='box_select')
    scatter.circle(source=source, x='x', y='y')
    return scatter


def save_selected(dataframe, selection):
    data = dataframe.iloc[selection]

    def save():
        print(f'Saving selected points: {data.shape}')
        data.to_csv('data.csv', index=False, encoding='utf-8')

    return save


def clear_click_callbacks(button):
    button._callbacks['clicks'].clear()


def generate_makedoc_function(dataframe):
    columns = [
        TableColumn(field="x", title="X"),
        TableColumn(field="y", title="Y"),
        TableColumn(field="useful", title="Useful"),
    ]
    
    all_points = dataframe.index.values.tolist()

    def makedoc(doc):
        source = ColumnDataSource(dataframe)
        scatter = scatterplot(source)
        savebutton = Button(label='Download Selected Points')

        points_filename = ".".join(["selected_points", "csv"])
        savebutton.on_click(save_selected(dataframe, selection=[]))

        table = DataTable(
            source=source,
            columns=columns,
            sortable=True,
            selectable=True,
            width=800,
        )

        def update_selection(attr, old, new):
            global selected_points
            savebutton.on_click(save_selected(dataframe, selection=new))
            selected_points = new
            update_table(attr=None, old=None, new=None)

        def update_dataframe(dataframe, selected_points):
            dataframe = dataframe[['x', 'y', 'useful']].loc[selected_points, :]
            return dataframe

        def update_table(attr, old, new):
            global selected_points
            if len(selected_points) == 0:
                selected_points = all_points
            new_df = update_dataframe(dataframe, selected_points)
            table.source = ColumnDataSource(new_df)
            table.columns = columns

        source.selected.on_change('indices', update_selection)
        page_content = column([scatter, savebutton, table])
        doc.add_root(page_content)

    return makedoc


# In[ ]:


df = pd.DataFrame({'x': np.random.random(100),
                   'y': np.random.random(100),
                   'useful': np.random.randint(0, 100, size=100)})

show(generate_makedoc_function(df), notebook_url="http://127.0.0.1:8888")


# In[ ]:

save_dataframe_py.ipynb