Call back in real time graph with multiple traces in Plotly Dash

Question

how can I make a real-time live update graph with multiple traces in it? For every time interval, i need to read line from “tmp.txt” to data[‘prof’] and data [‘pred’] and update in the line graph.

I found the solution for live-updating here (https://pythonprogramming.net/live-graphs-data-visualization-application-dash-python-tutorial/) but it does not show how to @app.callback for multiple traces. Also I found out that it uses “Event” which is outdated.

My coding is as follow, appreciate if someone can help me. Thanks.

import dash
import dash_daq as daq
import dash_core_components as dcc
import dash_html_components as html
from dash.dependencies import Input, Output
from collections import deque

def read():
    with open ("tmp.txt", "r") as f:
        for line in f:
            data=line.split(',')
    return data 

i=0

data =  { 'prof':deque(maxlen=120), 'pred': deque(maxlen=120) }

external_stylesheets = ['https://codepen.io/chriddyp/pen/bWLwgP.css']

app = dash.Dash(__name__, external_stylesheets=external_stylesheets)

app.layout = html.Div(

        children=[
            dcc.Interval(id='timer', interval=1000),
            html.Br(),
            html.Div([  
                dcc.Graph(id='graph')
                ])
            ])

@app.callback(Output('graph','figure'), [Input('timer','n_intervals')])
def update_graph():
    data=read()
    data=read()
    X.append(i)
    data['prof'].append(float(data[1]))
    data['pred'].append(float(data[2]))

    figure={
        'data':[
            {'x':X,'y':CPU['prof'],'type':'scatter','name':'Profiled'},
            {'x':X,'y':CPU['pred'],'type':'scatter','name':'Predicted'}
            ]}
    i+=1
    return figure

Edit: 'tmp.txt' is a file constantly overwritten by another program. It only has one line like below: '3.2233 4.33445', 32, 74.0, 0.13, 0.0, 0.0

Do you mind to provide a full mcve? In particular tmp.txt. — rpanai
It lloks to me that you youtr tmp.txt can be read with pandas. — rpanai
Hi, thank you for the reply. Please refer to the edited part. — Hoo

Hoo Hoo · Accepted Answer · 2020-02-26T06:19:43

After trial and error, finally I came up with a solution using plotly.graph_objs:

   ....
   dcc.Interval(id='timer', interval=1000),
   html.Div([
            dcc.Graph(id='graph', animate=True),
            ], 
    ....

as for the callback:

@app.callback(Output('graph', 'figure'),
    [Input('timer', 'n_intervals')])
def update_graph_scatter(n):
    data=read()
    X.append(X[-1]+1)
    data['prof'].append(float(data[1]))
    data['pred'].append(float(data[2]))

    data = [go.Scatter(
            x=list(X),
            y=list(CPU['prof']),
            name='Prof',
            mode= 'lines+markers'
            ),
            go.Scatter(
            x=list(X),
            y=list(data['pred']),
            name='Pred',
            mode= 'lines+markers'
            ),
            ]

    return {'data': data,'layout' : go.Layout(xaxis=dict(range=[min(X),max(X)]),
                                                yaxis=dict(range=[0,380]))}

Call back in real time graph with multiple traces in Plotly Dash

1 Answers