问题
I am trying to add interactivity to a plotly 3d scatter plot which I am hosting using dash. My question contains two related parts:
(i) I would like to manually select points in my 3d scatter and change the selected points' colour to red. Selection should include click event and selection event.
(ii) I would like to be able to delete these points from the plot on pressing of a key e.g. 'delete' key.
Part (i) is similar to an example in the plotly guide https://plot.ly/python/click-events/, however, on_click is not available as a method for 3d scatter charts.
I have been attempting to use FigureWidget as that apparently provides methods to capture clicks and selections, but after 2 days I am struggling to make headway.
Example data (copy below to clipboard and run df = pd.read_clipboard(sep=','))
id,x_val,z_val,y_val
a,11.313449728149418,0.13039110880256777,0.5386387766748618
b,11.321463427315383,0.2360697833061771,1.32441455152796
c,10.127132005050942,0.23085014016641864,1.0961116175427044
d,11.639819269465233,0.0958798324712593,0.6506370305953094
e,8.892696370438149,0.08223988244819926,0.6440321391968353
f,6.711586646011124,0.3657515974938044,
g,7.095030650760687,,0.5723062047617504
h,6.4523124528415,,1.293852184258803
i,7.165105300812886,0.4151365420301895,-0.5920674079031845
j,7.480703395137295,0.14284429977557123,1.0600936940126982
k,5.570775744372319,,
l,4.358946555449826,,
I have example code below which I hope is almost there (but not quite). The idea is that 'handle_click' should capture click events. I should also handle 'selection' events, although I am not sure how to do this as 3d scatter plot does not provide a selection box or lassoo tool. With the callback, I'm not even sure how to initiate, as there is no clickData/selectionData event with 3d scatterplot that I can utilize (so [Input('subindustry-dropdown', 'value')]) is incorrect and note that subindustry-dropdown is not something that I have provided in the example, but I select my ids from a dropdown which returns a subindustry value in my development version. )
import dash
import dash_core_components as dcc
import dash_html_components as html
from dash.dependencies import Input, Output
import plotly.graph_objs as go
import pandas as pd
external_stylesheets = ['https://codepen.io/chriddyp/pen/bWLwgP.css']
app = dash.Dash(__name__, external_stylesheets=external_stylesheets)
app.css.append_css({
"external_url": "https://codepen.io/chriddyp/pen/bWLwgP.css"
})
app.layout = html.Div([html.Div(id = '3d-scatter'),
])
# Calculate and store data
@app.callback(Output('3d-scatter', 'children'),
[Input('subindustry-dropdown', 'value')])
def chart_3d():
f = go.FigureWidget(px.scatter_3d(df, x = 'x_val', y = 'y_val', z = 'z_val', hover_name = 'company_nm'))
f.layout.clickmode = 'event+select'
f.data[0].on_click(handle_click) # if click, then update point/df.
return dcc.Graph(id = '3d_scat', figure=f)
def handle_click(trace, points, selector):
c = list(f.data[0].marker.color)
s = list(f.data[0].marker.size)
for i in points.point_inds:
c[i] = '#bae2be'
s[i] = 20
with f.batch_update():
f.data[0].marker.color = c
f.data[0].marker.size = s
return f.data[0]
回答1:
Here is a solution which allows:
selecting points by clicking individually
delete selected points by pressing html button
clear selection by pressing html button
according to this issue selecting multiple points in a 3d plot is currently not supported
(The use of FigureWidget does not seem to make a difference, so I removed it)
import dash
import plotly.express as px
import dash_core_components as dcc
import dash_html_components as html
from dash.dependencies import Input, Output, State
import plotly.graph_objs as go
import pandas as pd
import json
external_stylesheets = ['https://codepen.io/chriddyp/pen/bWLwgP.css']
app = dash.Dash(__name__, external_stylesheets=external_stylesheets)
app.css.append_css({
"external_url": "https://codepen.io/chriddyp/pen/bWLwgP.css"
})
import pandas as pd
df = pd.DataFrame(columns=['x_val','z_val','y_val'])
df.loc['a'] = [11.313449728149418,0.13039110880256777,0.5386387766748618]
df.loc['b'] = [11.321463427315383,0.2360697833061771,1.32441455152796]
df.loc['c'] = [10.127132005050942,0.23085014016641864,1.0961116175427044]
df.loc['d'] = [11.639819269465233,0.0958798324712593,0.6506370305953094]
df.loc['e'] = [8.892696370438149,0.08223988244819926,0.6440321391968353]
df.loc['f'] = [6.711586646011124,0.3657515974938044,0]
df.loc['g'] = [7.095030650760687,0,0.5723062047617504]
df.loc['h'] = [6.4523124528415,0,1.293852184258803]
df.loc['i'] = [7.165105300812886,0.4151365420301895,-0.5920674079031845]
df.loc['j'] = [7.480703395137295,0.14284429977557123,1.0600936940126982]
df.loc['k'] = [5.570775744372319,0,0]
df.loc['l'] = [4.358946555449826,0,0]
def create_figure(skip_points=[]):
dfs = df.drop(skip_points)
return px.scatter_3d(dfs, x = 'x_val', y = 'y_val', z = 'z_val')
f= create_figure()
app.layout = html.Div([html.Button('Delete', id='delete'),
html.Button('Clear Selection', id='clear'),
dcc.Graph(id = '3d_scat', figure=f),
html.Div('selected:'),
html.Div(id='selected_points'), #, style={'display': 'none'})),
html.Div('deleted:'),
html.Div(id='deleted_points') #, style={'display': 'none'}))
])
@app.callback(Output('deleted_points', 'children'),
[Input('delete', 'n_clicks')],
[State('selected_points', 'children'),
State('deleted_points', 'children')])
def delete_points(n_clicks, selected_points, delete_points):
print('n_clicks:',n_clicks)
if selected_points:
selected_points = json.loads(selected_points)
else:
selected_points = []
if delete_points:
deleted_points = json.loads(delete_points)
else:
deleted_points = []
ns = [p['pointNumber'] for p in selected_points]
new_indices = [df.index[n] for n in ns if df.index[n] not in deleted_points]
print('new',new_indices)
deleted_points.extend(new_indices)
return json.dumps(deleted_points)
@app.callback(Output('selected_points', 'children'),
[Input('3d_scat', 'clickData'),
Input('deleted_points', 'children'),
Input('clear', 'n_clicks')],
[State('selected_points', 'children')])
def select_point(clickData, deleted_points, clear_clicked, selected_points):
ctx = dash.callback_context
ids = [c['prop_id'] for c in ctx.triggered]
if selected_points:
results = json.loads(selected_points)
else:
results = []
if '3d_scat.clickData' in ids:
if clickData:
for p in clickData['points']:
if p not in results:
results.append(p)
if 'deleted_points.children' in ids or 'clear.n_clicks' in ids:
results = []
results = json.dumps(results)
return results
@app.callback(Output('3d_scat', 'figure'),
[Input('selected_points', 'children'),
Input('deleted_points', 'children')],
[State('deleted_points', 'children')])
def chart_3d( selected_points, deleted_points_input, deleted_points_state):
global f
deleted_points = json.loads(deleted_points_state) if deleted_points_state else []
f = create_figure(deleted_points)
selected_points = json.loads(selected_points) if selected_points else []
if selected_points:
f.add_trace(
go.Scatter3d(
mode='markers',
x=[p['x'] for p in selected_points],
y=[p['y'] for p in selected_points],
z=[p['z'] for p in selected_points],
marker=dict(
color='red',
size=5,
line=dict(
color='red',
width=2
)
),
showlegend=False
)
)
return f
if __name__ == '__main__':
app.run_server(debug=True)
来源：https://stackoverflow.com/questions/58310199/select-and-delete-data-points-in-plotly-dash-3d-scatter