I plot a figure using this code. After playing this cell the program became too much slow and it is very hard to zoom the picture. How can I speed up this code? (consider that this code is not completed and in the coming updates I will need for loop, so I can not replace it)

fig = go.Figure()

data=0
X=[]
Y=[]
for trial in range(len(frames_ts)):
    for data in range(len(frames_ts[trial])):
        new_y=original_output[trial][data];
        new_x = frames_ts[trial][data]/sampling_rate; 
        X.append(new_x);
        Y.append(new_y);
        data = data   1;
fig.add_trace(go.Scatter(x = X, y =Y, mode='markers', name='Real'))

CodePudding user response：

With go.Scatter, Plotly adds html/svg data into the document. The more data you add, the slower is going to be the rendering done by Plotly-js. More so, rendering a trace with go.Scatter(x = X, y =Y, mode='markers') is going to be significantly slower than rendering go.Scatter(x = X, y =Y, mode='lines').

As far as I can see, you have options:

1. Use Plotly and go.Scatter(x = X, y =Y, mode='lines').
2. If you really need a scatter plot, use Bokeh, which is a canvas-based and it is muuuch faster at rendering (in contrast to plotly which is html-document-based).

CodePudding user response：

• given such a large data set, use Scattergl() instead of Scatter() as trace type
• below code shows this - after plotting interaction is ok

import numpy as np
import plotly.graph_objects as go

go.Figure(
    go.Scattergl(
        x=np.linspace(0, 100, 155585),
        y=np.sin(np.linspace(0, 100, 155585)),
        mode="markers",
        name="Real",
    )
)