Add parameter to control the lowess trendline level of adjustment

[GitHunter0]

Sometimes the lowess trendline gives a poor fit to the points.

Would you consider adding a parameter to control the level of adjustement?

What would solve the problem is the equivalent to the parameter frac of statsmodels.nonparametric.smoothers_lowess.lowess() function.

This is expressed in this Stack Overflow question:
https://stackoverflow.com/questions/67189101/how-to-adjust-plotly-express-lowess-trendline-parameters

Thank you

[nicolaskruchten]

Yep, this is a good idea and I've got a mostly-working implementation here #2997 ... just need to find time to QA and merge it :)

[GitHunter0]

Awesome, I'm glad to know, thank you

[GitHunter0]

Closing, since now that can be done as below.
OBS: trendline="lowess" requires 'date' to be in datetime format, otherwise date can be a date string.

import plotly.express as px
px.scatter(data_frame = df, 
           x = "date", 
           y = "y",
           trendline="lowess",
           trendline_options=dict(frac=0.2))