Live Updating Candlestick chart wit Python

I’m trying to make a live chart using python and plotly, but when the update function is called, it doesn’t shows the updated data. Instead, it empties the chart. The first time it runs the function, the data is shown on the screen. The second time, it deletes the data from screen and it never gets updated again.

I’ve been searching the web for days and can’t find nothing. So, here I am.

This is the code:

from dash.dependencies import Output, Event
import dash
import dash_core_components as dcc
import dash_html_components as html
import numpy as np
import pandas as pd
import plotly
import plotly.graph_objs as go

dados = np.array([[1509608700000,0.00002246,0.00002246,0.00002246,0.00002246,100.00000000,1509609599999],
[1509609600000,0.00002800,0.00002802,0.00002800,0.00002800,6832.00000000,1509610499999],
[1509610500000,0.00002700,0.00002700,0.00002501,0.00002501,3936.00000000,1509611399999],
[1509611400000,0.00002588,0.00002678,0.00002588,0.00002614,7125.00000000,1509612299999],
[1509612300000,0.00002615,0.00002621,0.00002614,0.00002617,19318.00000000,1509613199999],
[1509613200000,0.00002627,0.00002643,0.00002625,0.00002627,109218.00000000,1509614099999],
[1509614100000,0.00002627,0.00002642,0.00002603,0.00002639,134825.00000000,1509614999999],
[1509615000000,0.00002639,0.00002655,0.00002616,0.00002618,74432.00000000,1509615899999]
])

columns = ['open_time', 'open', 'high', 'low','close', 'volume', 'close_time']

df = pd.DataFrame(data=dados, columns=columns)

last_id = 0

app = dash.Dash(__name__)

app.layout = html.Div(
    html.Div(className='container-fluid',children=
[
	html.Div(className='row', children=html.Div(dcc.Graph(id='live-graph', animate=True), className='col s12 m12 l12')),
	dcc.Interval(
		id='graph-update',
		interval=5000
	)
]),
)

@app.callback(
Output('live-graph','figure'),
events = [Event('graph-update','interval')]
)
def graph_update():
    global df

    ndf = df.iloc[0:2]
    print(ndf)

    last_id = ndf.iloc[0]['open_time']

    ndf['data'] = pd.to_datetime(ndf['open_time'], unit='ms')
    ndf.set_index('open_time', inplace=True)
    ndf.round({'close': 8, 'open': 8, 'high': 8, 'low': 8})

    data = [ dict(
        type = 'candlestick',
        open = ndf.open,
        high = ndf.high,
        low = ndf.low,
        close = ndf.close,
        x = ndf.data,
        yaxis = 'y2',
        name = 'Ripple',
    )]

    df = df.iloc[2:]
    print(df)

    return {'data': data, 'layout': {'title': str(last_id)}}

external_css = ["https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/css/materialize.min.css"]
for css in external_css:
    app.css.append_css({"external_url": css})

external_js = ['https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/js/materialize.min.js']
for js in external_js:
    app.scripts.append_script({'external_url': js})

server = app.server
dev_server = app.run_server(debug=True)

Original code performs a search in the database and returns 500 items each time, inside the updating function. I have adapted the code to be reproduceable and get 2 items each time from dataframe.

I also need the chart to be static showing days hours(0,3,6,9,12,3,6,9,0) and being updated with data from database… is it possible? How sould I adapt the code to accomplish this goals?

Thanks in advance.

This should work. Have you upgraded dash-core-components recently? See dash.plot.ly/installation. We released a couple of bug fixes to the candlestick chart recently.